CORNELL UNIVERSITY LIBRARY

ASIA

3 1924 066 284 856

(5

JOURNAL

OF THE

ASIATIC SOCIETY OF BENGAL.

VOL. LXVIII.

PART I. (History, Antiquities, &c.)

(No. I, and Extra Nos. I and IT. — 1899.)

EDITED BY THE

^ONORARY j^HILOLOGICAL ^SECRETARY.

“ It will flourish, if naturalists, chemists, antiquaries, philologers, and men of science

in different parts of Asia, will commit their observations to writing, and send them to

the Asiatic Society of Calcutta. It will languish, if such communications shall be long

intermitted; and it will die away, if they shall entirely cease.” SiR Wm. Jones.

*

CALCUTTA :

PRINTED AT THE BAPTIST MISSION PRESS,

AND PUBLISHED BY THE

ASIATIC SOCIETY, 57, PARK STREET.

1899.

T

/

\S

A N

y*

T

J

> )

I

r "•

' )

i

y

' -»

V

CONTENTS

OP THE

JOURNAL OF THE ASIATIC SOCIETY OF BENGAL,

* Vol. LXVIII, Part I— 1899.

No. 1.

On the Kaqmiri Verb. — By G. A. Grierson, C.I.E., Ph.D., I.C.S.

On Indeclinable Particles in Kagmiri. — By G. A. Grierson, C.I.E.,

Pm « D . , I# 0 # S t ### ttt tt#

On the genuineness of the Grant of Civasimha to Vidyapati-thakkura.

— By G. A. Grierson, C.I.E., Ph.D., I.C.S.

Page.

I

93

96

Extra No. 1. (With 1 map and 19 plates, issued in a separate cover.)

A Collection of Antiquities from Central Asia , Part I. — By A. F.

Rudolf Hoernle, C.I.E., Ph.D. {Tubingen).

Introduction ... ... ... ... ... 1

Section I. — Coins and Seals ... ... ... ... 1

Section II. — Block Prints ... ... ... ... 45

Extra No. 2. (With 2 maps, issued in a separate cover.)

Memoir on Maps illustrating the Ancient Geography of KaStnir. —

By M. A. Stein, Ph.D. ... ... ... ... 1-232

( For Table of Contents, see page 223 ff.)

No Index is issued with this Volume of the Journal.

LIST OF ILLUSTRATIONS

ISSUED WITH

JOURNAL, Vol. LXVIII, Part 1—1699.

Issued with Extra No. 1 : —

Map of Eastern Turkestan, to illustrate the Report on Central

Asian Antiquities.

Plates I and II : Central Asian Antiquities : Coins.

Plate III : Central Asian Antiquities : Coins and Seals.

Plate IV : Central Asian Antiquities : Miscellaneous Objects.

Plates V to XVIII : Central Asian Antiquities : Block-Prints.

Plate XIX : Central Asian Antiquities : Miniatures.

Issued with Extra No. 2 : —

Ancient S'rinagar. Ground Map reproduced from Survey, 1859-60.

Ancient Kasmir. Do. do. do.

JOURNAL

OF THE

ASIATIC SOCIETY OF BENGAL.

Part I. — HISTORY, LITERATURE, &c.

No. I.— 1899.

On the Kagmlri Verb. — By G. A, Grierson, C.I.E., Ph.D., I.C S.

[Read January, 1899.]

The Katpniri Verb is treated in the Akliyata- and Krdanta-prakri-

yas of l^vara-ka ilia’s Kagmuagabdamrta.

Verbal roots may end either in a consonant or in a vowel. When

a root ends in a consonant, the vowel a is added to assist the pronun¬

ciation. Thus the root tear is written kara. It is still, however,

considered to be a root ending in a consonant and is referred to as such.

The final ^ a is dropped before adding the conjugational suffixes. For

this reason, throughout this series of papers, I shall so far depart from

the rule of strict transliteration, that I shall not write the final ^ a of

roots ending in consonants, although that final a will be written in

the Nagari character.

When a root ends in a vowel, that vowel is always ^ i. There are

only eight of these roots, of which five belong to the first, and three to

the third conjugation, as follows, —

First Conjugation. — f% khi, eat ; ci, drink; f^; di, give; fif niy

take ; and f% hi, take.

Third Conjugation. — % zi, be born ; fg pi, fall ; and fa yi, come

J. i. 1

2

G. A. Grierson — On the Kdr/mlrl Verb. [No. 1,

As in the above examples, all verbs will be quoted under their

root-forms.

Excepting the verbs fa ni, take, fa di , give, and fa yi, come, all

verbs whose roots end in vowels change the final ^ i to ^ yci{e) through¬

out (viii. ii. 11). Thus, the present participle of all these verbs is made by

adding lean to the root. But the present participle of fa klii , eat,

is khyawdn , not fa^T«T khiwan. On the other hand, the present

participle of far ni, take, is fa^T«T niwan, not nyawdn. This rule

is a most important one,, and will be met over and over again in the

following pages.

Conjugation.

There are three conjugations of verbs. They only differ in the

formation of the Causal Voice, and of the past participles and of the

tenses derived from them. Their consideration is postponed till these

tenses come to be dealt with.

Voice.

There are three voices : the Active, the Passive, and the Causal.

The Active Voice.

This is formed by adding the conjugation al suffixes to the root

direct according to the rules to be hereafter detailed.

The Passive Voice.

This is formed by conjugating the oblique form of the verbal noun

in un, that is to say, the shortened form of the Instrumental Singular

which is used before post-positions, and which ends in ana, with the

verb fa yi, come, which may either precede or follow. Thus, from

the verb ^ kar, make, the verbal noun is karun. The oblique

form of this is karana, and the Passive is fa karana yi, be

made, literally, come into making. Compare the Hindi ^§#1 ^ ^fTrfT ^

dekline m& atd hai, it comes into seeing, it is seen (viii. i. 49).

Examples of the use of the Passive are (viii. i. 50).

fa^T«T rcinana yiwdn chuh bata, the rice is coming

into cooking, i.e., is being cooked.

1899.]

Passive Voice.

3

tn*PT fs?37*T pdna-y chuh yiwan karan ®, lie, himself, is

being made.

Even neuter verbs can take the passive form, without, in this case,

changing their meaning (viii. i. 51). Thus, —

«ffrn*T zotdn chuh : or

5TTfT*T f^r37«T zdtana yiwan cliuh , he is shining. The latter

example is literally, he is coming into shining.

dazdn chuh , or

^^j5T f^T«T dazana yiwan chuh , he is burning.

The root boz , hear, when used in the passive means ‘ see

(viii. i. 52). Thus, —

if]aTT*r bozan chuh , he hears, but.

• ** vj n. ' '

bdzana yiwan chuh , he is being seen.

bozana dv (dv, is the irregular past tense of yi), he

was seen.

^t^nr buzana yiyi, he will be seen.

If we want to express the passive of the verb ‘ to hear/ we must

use a periphrasis. Thus, —

sftfTircr bozanas andar yiwan chuh , he is coming

into hearing, he is being heard.

The root jit gar, work metal, may mean ‘ to become hard ’ in the

Passive (viii. i. 53). Thus, —

f*TqTT«T garana yiwan chuh, it is being worked, or it is

becoming hard.

The root fir ni, take, may mean 4 to have the attention distracted ’

in the Passive (viii. i. 55). Thus, —

f*r*r f^cTT«T nina yiwan chuh, he is being distracted, or he is

being taken.

The root hi, take, may mean ‘ to be engaged in ’ in the Passive

(viii. i. 56). Thus, ^ir hyana yiwan chuh , he is engaged (in a

business), or he is being taken.

The root des, see, is irregular. Its Passive is formed thus,

misr Wl? dreth yiwan chuh , he is being seen (viii. i. 54).

4

G. A. Grierson — On the Kagmlri Verb. [No. 1,

In adding this termination ^ ana, we must remember that in the

ease of the verbs f«r ni, take, cli, give, f?T yi, come, the first ^ a is

elided (viii. ii. 11). Thus, fi?sr nina yiwan chuh , he is being

taken.

f^sr dina yiwan chuh , he is being given.

yin0 yiwan chuh , it is being come by him, i.e ., he is

coming. In the case of other verbs ending in i, that vowel becomes

y. Thus, —

khi, eat, fw*T khyana yiwan chuh , he is being eaten.

f% hi, take, hyana yiwan chuh, he is being taken

ci, drink, f^^Tl cyan ° yiivdn chuh , it is being drunk.

The Causal Voice.

A root is made causal by adding anaw (viii. iv. 2, 3). Thus,

eftT kar, make ; karanaw, cause to make. So, in the present

tense, 3&n«rT karan chuh , he makes ; ^sn^T^r karandwdn chuh ,

' vj ' 7 'VI'

he causes to make.

Exceptions —

The root wuph, fly, is regular. Thus, wuphandivan

chuh , he causes to fly. But when the causal verb means * to incite,’

the n is changed to l (viii. iv. 4). Thus, wuphalawdn

chuh, be incites.

Intransitive roots containing

anaiv in forming causals (viii.

From —

wobar, be finished.

kabsar , be tawny.

kumal, be tender.

cSkhar, be contracted.

three aksaras omit the ^?r an of

•**

iv. 7). Thus, —

^^TsTT«T ivobarawan chuh , he

vl '

finishes.

kdtsardwdn chuh , he

makes tawny.

kumaldwdn chuh, he

makes tender.

cokhardivan chuh, he

^ vj X

makes contracted.

1899.]

Causal Voice.

5

tshd bshan, be light.

mandach , be ashamed.

33^% wozal , be red.

^Tfi^ samakh , become visible.

c^qaT ivopaz , be born.

This exception does not apply 1

Thus, —

wlw^TIqT^r tsh dtshanawdn chuh ,

^ \J s

be makes light.

w ^ umndachawan chuh ,

^ VJ x

he makes ashamed.

wozalawan chuh , he

* * VJ x

makes red.

*3*r*§r^l«T Wf samakhdivdn chuh, he

- j '

makes visible.

icopazdwdn chuh, he

produces.

transitive verbs, which are regular.

kaviav, earn. 3nFR<rrT^T«T kamavandwdn chuh ,

be causes to earn.

katar, slice. kataranawan chuh ,

he causes to cut in slices.

kapat, cut (clothes). 3\^*TT^T«T kapat andwan chuh ,

be causes to cut.

The root epji wusn, be hot, also drops the an. Thus,

wusndwan chuh, he makes hot (viii. iv. 8).

vj x

The verbs f wuz, be awake ; hod , dive ; ^nr; Idr, touch ;

fw pil, arrive; 3r[«T kamp , tremble, and ranz, be pleased, add

optionally ^ av , instead of andv (viii. iv. 5, 14). Thus, ^5rqrT»r

wuzawdn chu-s, or ^5f?rreT«T ivuzanawan chu-s , he wakens him. So

ef>TWRT*r kampaivan chu-s , or efiTHTcrT<n«T kampanaiudn chu-s, he

causes him to tremble. The root w pras, be born, has three forms

(viii. iv. 5, 15) ; viz. prasawan chuh ; sr^rreTST ^ prasandwan

chuh, and r«r pinawdn chuh, he causes to bear children.

The verb f*7 yi, come, makes its causal anandw. ["This is

really the causal of an, bring]. Thus, ^*r*TT^T«r ananawdn chuh,

he causes to bring.

The root phat, be split, makes its causal phdtawdn

chuh, or ^j£*TTCT*r phdtanawan chuh (viii. iv. 17).

6

G. A. Grierson — On the Kdgmiri Verb. [No. 1,

With regard to roots ending in ^ i, the following are the forms

(viii. iv. 9-13).

Regarding fa yi , come, see supra , p. 5.

Regarding far zi , be born, see post, p. 9.

Other verbs of the third Conjugation form their causals by adding

if the final consonant of the root is <T t it becomes tI ts-

\ d „ ?! g.

*[ « or 5 ,, n (viii. iv. 25).

1899.]

Causal Voice.

7

[The only root of the third conjugation which ends in 'Sj th is

tfptf path, become, and, according to my Pandit, its causal is regular,

xrr«l5rT3ri«T spf- pathanawdn chah. ■qt^CRTT^T W^pgbsharciwan chuh , is possi¬

ble, but unusual].

kal, be dumb.

gob, be too heavy.

vyath, be fat.

^T*r kdn, be one-eyed.

thik, stand firmly.

S I

tez, be sharp,

ija- breth, be a fool.

*fT3' path, be fat.

lok, be small.

W«T chat, be white,

rfrf tat, be hot.

^ thad, be high,

pod, be pure.

ban, be.

kalarawan chuh , he

makes dumb.

33^TT«TT«T gobarawan chuh, he

makes too heavy.

3J3TT3i«T W^[ vygthardwdn chuh , he

makes fat.

*fflsPCreiir ^ kdnarawdn chuh, he

*x -x vj x • 7

makes one-eyed.

thikardwdn chuh, he

makes to stand firmly.

sft^PCTWlir tizardwdn chuh , he

x "X yj X 7

makes sharp.

irte^'RTsr brithardwdn chuh, he

X ’x \j| x 7

makes foolish.

puthardivdn chuh, he

makes fat.

^r^TTsTT«T lukardwdn chuh, he

makes small.

chatsardwdn chuh , he

makes white.

I

tatsardwdn chuh, he

makes hot.

thazarawan chuh, he

elevates.

i

^«T’a^rT*T gozarmvan chuh , he

purifies.

WTRT5?; bgharawan chuh, he

causes to be.

ris^RT^; tahHdwan chuh , he

makes thin.

<T*T tan, be thin.

8

G. A. Grierson — On the Kdgmiri Verb.

[No. ] ,

The verb qfqr tsolc, if it means ‘ be sour,' makes its causal ^3[Riqi*r

• # N "

tsokardwdn chuh ; but if it means * be angry,’ its causal is

tsukardwdn chuh. The verb khal, be loose, makes its ?■ causal

yj X * tt.

?§^Riqi«?r W\if khalardwdn chuh , or ’qaRyqioT khajardwdn chuh.

The following verbs form their causals optionally by adding either

andw, or ardw , (viii. iv. 19, 21, 22, 23).

First Conjugation : W3V chak , scatter ; wq chap, pass time ;

tsen , know by a sign ; man, confess.

Second Conjugation : tsal , flee ; ^ tshyann, be split ;

$

thak, be weary ; qre phas, be entangled ; qpr phut, be broken ; pholl,

expand (of a flower) ; qq mas , forget ; rav, be lost ; hokh, be

dry ; hobs, decay ; qq grap, be digested.

Third Conjugation : tyatli , be bitter.

Thus, tsihardivdn chuh , or %*r«nqi«FT wq tsenandzudn

* " VJ ■» ~ X

chuh: ^^TT^1*T wq tsolardwdn chuh , or ^?«n^T5f tsalanawan chuh:

X x \J ' • ** vj V '

4a»5T«n«T W^ tyatharmvdn chuh, or tyathand wan chuh.

The root chak does not alter its meaning in the causal in ardw.

Thus, W3H5T wq chak an chuh, wqRTqre chaka rowan chuh , both mean

'XyjN V yj V * /

‘ lie scatters.’ To give a causal meaning it has W3\«TRI«T chakanawan

chuh , or chakarandicdn chuh.

The root w? chap has for its causal wqTrqre chapHdwdn chuh,

WqsfRTST chapandwdn chuh, or ^xr^i«ri«T chopardwcin chuh.

The following verbs of the third conjugation form their cansals in

andw, and not in ardw (viii. iv. 17). kats , be wet; goh, shine ;

grak, boil over; after zot, shine; tak, run; qq? tel, smart; cfiq

tos, be satisfied (according to my Pandit, this verb belongs to the 2nd

conjugation); dor, run; itfq nap, shine; nil , become blue;

pis, boil over ; q^ ped, exude ; qR par, be competent ; qR prar,

wait (according to my Pandit, this verb belongs to the 2nd conjuga¬

tion) ; qq phab, be excellent ; qR phar, be stolen ; qr^T phal, become

old (of clothes); qpr phug or qpf phuh, be inwardly angry; pher,

go round ; qfR phbr, quiver (according to my Pandit, this verb belongs

1899.]

Causal Voice.

9

to the 2nd conjugation) ; bad , be powerful ; has, become clear

(according to my Pandit, this verb belongs to the 2nd conjugation) ;

^ bud, be old : braz, shine ; ^rrq- yap, pervade ; ^ ramb, be beauti¬

ful ; ^ ras, be full of juice ; rots, be preferred ; rot , be stopped ;

W rad, persistently follow; ran, be worn out; lyad , be con¬

quered ; lor, become deficient. Thus, 3i^iTrTr«r katsanawdn

chuli, not efi^T^r[«r k^tiardwdn chuh.

x x x x

All causal verbs in ardw, may optionally drop the syllable aw in

the termination, and add ar instead of ardw (viii. iv. 24). Thus, —

Instead of — We may have —

kalardwdn chuh halardn chuh

N X ^ X * V \ ^ X •

qobardivdn chuh qobaran chuh

X N X ^ ' > s > V •

tsokardwdn chuh ^f\TT»T tsokardn chuh

• x X X • • X X vj X

and so on.

The following verbs form their causals by merely lengthening

their root vowels (viii. iv. 28).

rfT tar, be crossed. Causal fTTO*T tdran chuh, he crosses.

mar, die. tRKl«T mdrdn chuh , he kills.

dal, pass over. ^T^rT«T ddlan chuh, lie causes to

pass over.

*TJT lag, be with. ^T[Jir*r lagan chuh, he unites.

When in; mar (18, 28), means ‘unite,’ and when lag (28)

means ‘suffer pain,’ or ‘fit,’ they are regular. Thus, *TT«TT3T«T

maranawdn chuh , he causes to unite ; lagandwan chuh, he

causes to suffer pain.

The following are quite irregular, —

f«T zi, be born. Causal zovardn chuh, he brings

forth (26).

khas, ascend. ^ khdrdn chuh, he causes to

ascend (29).

vas, descend. wdrdn chuh, he brings down

(29).

J. i. 2

10

G. A. Grierson — On the Kdgmirl Verb. [No. 1,

gdrjg, go to sleep. sdwdn chuh , lie puts to

sleep (30).

zdldn chuh, lie burns

(act.) (31).

hdwdn chuh, he shows (32).

' VJ »

pakandwan chuh, he

drives.

Jl^in«n*r gatshanawan chuh, he

despatches (33).

I'll us, JlT^iT ^•fT^T»T gov^n paka-

nnivdn chuh, he drives the cows ;

guTjgdya gab&hand-

wdn chu-s, he sends him to the

Ganges.

tuldn chuh, he raises, he

lifts up ; but oEpgsnsn*r

wothandwdn chuh, he causes so

and so to rise (34).

Verbal Suffixes.

Before proceeding to the consideration of the conjugation of the

verb, it is necessary to describe in detail one remarkable feature of the

K^miri language, which it shares with other languages of the North-

Western group of the Indo- Aryan Vernaculars, namely, the facility

with which the meaning of the verbal stem can be modified by the

addition of suffixes. Indeed, it may be said that, given the form of a

tense-stem, there is usually no conjugation, in the proper sense of the

word, at all. In most cases, suffixes, which may be added, or detached,

at pleasure, and most of which have an independent recognised exis¬

tence, are added, and give the various modifications of meaning which

we designate number and person, or of negation, affirmation, and so on,

by forming true compound words, and without becoming merged in the

base in the form of terminations. Thus, take the word karu. This

means ‘ made,’ and may mean, ‘ made by me,’ ‘ made by us,’ 1 made by

him’ and so on. That is to say, it means, ‘ I made,’ ‘ we made,’ ‘he

made,’ &c. If we wish to lay stress on the person who made, we may

say rrf*T tgm1 karu , ‘ by him made,’ i.e., ‘ he made.’ Instead,

however of using tam1, the instrumental singular of the third

daz, burn.

des, see.

31^ gat§h .

ivoth, rise.

1899.]

Adverbial Suffixes.

11

personal pronoun, in Ka^miri we may optionally add tlie suffix n,

which means ‘ by him,’ just as much as rrfiT tam* does, and we get

karu-n, which also means ‘made by him,’ or ‘ he made.’ Suppose

we want to express who was made by him, and that the person is the

speaker, then we can say «fT^«T boh karu-n , ‘ I was made by him,’

i-e., ‘ he made me.’ Instead, however, of boh , we may add the

suffix as, which means ‘I.’ We thus get karu-n-as , ‘ I was

made by him,’ i.e., 1 he made me.’ Again, if we wish to emphasise the

fact that I was the person made, we can add the suffix f<T ti, and we

get the form kar1l-n-as-ti, which means, ‘ I also was made by

him,’ ‘ he made me also.’ Again, if we want to make the verb interro¬

gative, we can add, after all these, the interrogative particle, a, thus,

eff^rwir karu-n-as-ty-a , ‘ was I also made by him ? ’ ‘ did he make me

also ? ’

The above examples will show the freedom with which these

suffixes are used in Kashmir!. They can be combined almost ad infini¬

tum. These suffixes may be divided into two classes, adverbial and

pronominal, and in this order, I now proceed to discuss them.

Adverbial Suffixes.

These suffixes are added to all verbs. Before all these the final

^ h of a verbal form is elided (iv. 131). The ordinary rules of sandhi

also occur. Thus i and u before & become y and w respectively, ya

(e) -f a becomes ya, and a Ad becomes a.

1. na. This negatives the verb (viii. i. 13). Thus, —

Wlf karan chuh, he

VJ n

makes.

3KT5T karan chih , they

make.

karan cheh , she

x \ \

makes.

eUTI^f karan chekli, thou

(fern.) makest.

kari, he will make.

qayov, he went.

4 eft paku, he went.

karan chuna, he does not

make.

fw*T karan china , they do not

make.

^I*T karan chena, she does not

make.

karan chekhna, thou

dost not make.

karina, he will not make.

gayovna , he did not go.

■R^cf pakvna, he did not go.

12

G. A. Grierson — On the Kacmlrl Verb.

[No. 1,

2. a. This gives an interrogative force to the verb (viii. i. 14).

Thus, —

ejpa«r kardn cliuh. he

V vj S.

makes.

karan cheli (lit.

chyah ), she makes.

kari, he will make.

WT (for ^T,W + ^T )

karan chivd, does he make ?

3>TT*T 13T karan chya , does sbe

make ?

karya , will he make ?

parav , we shall read.

karydn , he made.

kgrun , he made.

gayov, he went.

paku, he went.

In the first and third persons

a, when the person addressed is a

■qT3T parawd, shall we read P

karyon a , made he ?

kgrund, made he ?

gayovd , did he go P

pgkud , did he go ?

Feminine, ay is substituted for

woman. If a man is addressed,

^ 9 is used in the first person Singular, and d in the first person

Plural, and in the third person.

Thus, «RTT»T karan chesa , am I (fern.) making P here a man is

addressed. If a woman is addressed, the speaker would say chesay.

$jr kardn chyd ( cheh -f d) gs\ are we (fern.) making ? If a

woman is addressed, the speaker must say cMy. See No. 4.

3. °. This may be substituted for d, in the following cases.

(а) In the first person Singular Masculine (viii. i. 15)

Thus, —

kardn chtis , I make. kardn chusa (instead of

\ yj x 1 > \J '

qncr*T W^T kardn chusa), am I

making P

(б) Always in the first person Singular Feminine, when a

man is addressed (viii. i. 17). Thus, —

kardn ches, I (fern.) qPCI«T kardn chesa, am I (fern.)

make. making P Here the speaker is

addressing a man. If she was

addressing a woman she would

say kardn chesay.

(c) Honorifically in the second person Singular and Plural

(viii. i. 15).

1899.]

Adverbial Suffixes.

13

^TT*r ff’sf bar an chukh , thou

' j '

makest.

karan chiwa , you

make.

4. ^ ay or, after a vowel, y.

^CT^r karan chukh a, does Your

Honour make ?

fw*T karan chiwa , do your

Honours make ? Note here

that the vowel remains short.

Used as follows, instead of

or m a.

(a) In the first person Singular and Plural Feminine, and in

the third person Singular and Plural Feminine (viii, i.

17, 18), when a woman is addressed. Thus, —

^TT*f karan dies, I (fern.)

make.

3fTU«r karan cheh

as*, we (fern.) make.

karan cheh soli ,

slie makes.

lit fcHT karan cheh

x x

tima, they (fern.) make.

karan chesay am I (fern.)

making ? Here the speaker is

addressing a woman. If she

were addressing a man, she

would say karan chesa.

karan chey as1, are

we (fern.) making ?

karan cliey soh, is she

(fern.) making ?

karan chey tima , are

they (fern.) making ?

In the three last, the speaker is also addressing a woman. If she

were addressing a man, she would say igr chyd, instead of cliey.

(6) In the second person Feminine optionally instead of

m a, when a woman is addressed honorifically (viii. i.

16). Thus,—

karan chekh, thou ^T^F^F karan cliekhay, or

(fern.) makest.

3RTF*F karan chewa , you

(fem.) make.

karan chHkha, is Your

Honour (fem.) making? Here

the speaker is necessarily ad¬

dressing a woman.

qF’TTSf karan cheway, or ^TrT

X

^ karan chewa, are Your

Honours (fem.) making ? The

speaker is again addressing

womeu.

14

G. A. Grierson — On the Kagmiri Verb. [No. 1,

5. frj- ti (iv. 179). This suffix is used to signify ‘also,’ ‘indeed.’

Thus, —

karan chuh , he wf<T karan chuti , he makes

makes. also.

kari , he will make. kariti , he will indeed make.

6. *rr wa, way (viii. i. 14, 16). This is a compound of *r na

(No. 1) and a (No. 2) or w ° (No. 3), or of na , and ^ (No. 4).

It gives the force of an interrogative negative, and is used like the

separate parts. Thus, —

karan chuh, he ^fifT karan chuna , does he

makes. • not make ?

So ^fr karan chena, does she not make ?

cfrirfcr karyon, he made. karyonna, did he not

make ?

kari , he will make. ^f^TT karina , will he not make ?

^PQr*T karan cliekh, thou

(fern.) makest.

3TTT«T karan chekhna , dost

V X /

thou (fern.) not make ?

karan chekhnay , does

Your Honour (fern.) not

make P

karan chewa, you iprifT karan chewana , do you

(fern.) make. (fern.) not do ? or ^TT5*

karan chewanay , do Your

Honours (fem. ) not make ?

7. HIT tya or HI*! tyay. This is a combination of f?r ti, (No. 5)

and WT a (No. 2), or ^ ay (No. 4). It implies a question with

emphasis (viii. i. 14). Thus, —

WHIT karan chutya , does he make { it) ?

tfi%faHir karyontya, did he make (it) ?

^fitXHir karitya, will he make (it) ?

3>TT«T <5PTHTC karan cheivatyay, do Your Honours (fem.) make

(it)?

8. sana , ^rifT sana, ^T^«T asana , used in a question with doubt.

If there is an interrogative word also in the sentence, it is added to it.

Otherwise it is added to the verb. [ sana is not used with a verb].

15

1899.] Pronominal Suffixes.

The a of ^*rr sand and dsana , is suffix No. 2 already described

(viii. i. 25, 28). Thus, —

t^TiT fr^r«T khyawan chwasan a (chuh + dsana), is he really

eating ?

fj T kydsana ( kydh + sana ) khyawan chiva , what, is

he really eating ?

kydsana khyawdn chuln , what, is he eating ?

fw^ lata kaitydsana (kait1 -\- as ana) dsan chili

how many brahmans are there really ?

3rrf?T f^n^T karsand lag 1 yiwdn chuh , at what hour

is he coming? (&ar = when? lggl — S>\Y. Ihdga , a

portion of the day or night).

^>T^«TT karsand yiyi , when will be come ?

fjf?TW*rT kuPsand asahan , how many may there be ?

^frf^iTT katisand osu, where was he ?

'i

9. *JT sg. This is the vocative particle (vide ante , Vol. lxvii,

p. 92). It is used exactly like san *. Thus, —

SrSfT^T kaityasg (kaiP ^d-sa) lukh as *, how many

people were there ?

Pronominal Suffixes.

1. Before these as before all other suffixes, the final ^ h of a verb

is elided (iv. 131). So also, an initial ^ a of a suffix is elided when

the verb, either after the elision of h or not, ends in a vowel (viii, i.

39). Thus, ^+^*1 chuh + am becomes first chu-\- am, and

then w-f *r chu+m = W*T chum , there is to me.

\J 1 s si s

2. If the final *3 kh of a suffix is followed by another pronominal

(not an adverbial) suffix commencing with a vowel, the kh becomes

^ h (viii. i. 38). Thus, —

afr*;T«T karan chu-h-as (for chu-kli -+ as), thou makest for

vj x x '

him.

karan chu-h-akh, thou makest for them.

- "

3. The termination ^ av becomes o before suffixes (viii. ii.

18). Thus, —

sfPCm kard-th (karav + atli), we shall make thee.

16

G. A. Grierson — On the Kdgmirl Verb.

[No. 1,

4. The termination ^ iv becomes ^yu before suffixes. Thus, —

3TO1T icaryu-m (for Jcariv- f am), make ye for me, or make ye me.

0\ N

When suffixes are added to the root direct, in the second singular

imperative, a ^ u is inserted (viii. i. 16). Thus, —

mX kar , make thou. karu-m , mnke thou for me,

or make me.

When suffixes are added to the 3rd person singular of the Future,

indicative, the ^ a of the suffix is not elided. Thus, kari, he will

make, kary-as , not kari-s , he will make for him. This does

not hold with the suffixes of the second person. Thus, kari-y , he

will make for thee, kari-wa , he will make for you.

[When interrogative and other adverbial suffixes are added to the

verb, they follow the pronominal suffix. E.g., chu-m-a , is there to

me ? So W^fT^T chu-m-asana, &c.l.

VI

Some verbs are never used without pronominal suffixes of the

dative case (viii. i. 45). These are, —

tsar, be inwardly wrathful.

phog, be inwardly wrathful.

phuh, be inwardly wrathful.

marts , be inwardly wrathful.

writs , be burnt.

fhrrW phite, forget.

tyavib, look eagerly (viii. iii. 45).

Moreover these verbs are always conjugated in the feminine,

whether the subject is masculine or feminine. They are then used as

impersonal verbs. Thus, —

tsaran che-s, lit. there is inward anger to him. I.e.,

he is inwardly angry.

Similarly, phogdn che-s , phuhan che-s,

3pj martsdn che-s . Again tsaran che-m, there is inward anger

to me, I am inwardly angry, and so on. Again, wutsdn che-s,

there is burning to him, i.e., he is burning (inwardly) 5 fqPWPT

phitsan che-s, there is forgetfulness to him, he forgets : $pj

tyambdn che-s, lie looks eagerly.

1899.] Pronominal Suffixes — First Person. 17

Sometimes full pronouns are used instead of suffixes (viii. i. 46).

Thus,—

rrfW^r IT? tamis tsaran cheh , there is inward anger to

him.

The verb jtw gatsh , be proper, be desirable, is also used with the

dative (viii. i. 47). Thus, —

rrfa*3 W1^ f^T tamis gatshan chuh zi parahg , to him it is

desirable that I should read. He thinks it proper that I should read.

This is only in the third person. For the other persons always, and for

the third person optionally, suffixes are used when they exist (viii. i.

48). Thus, —

jtwi«T gatshan chu-m, it is proper for me,

3T^pr[ gatshan chuh (no suffix), it is proper for us.

WSJ gatshan chu-y, it is proper for thee.

JTwrw W*T gatshan chu-wa , it is proper for you.

JIWT«T W^; gatshan chu-s , it is proper for him.

JTWrw ws§ gatshan chu-kh , it is proper for them.

This applies only to the present tense.

First Person.

The suffix of the Nominative Singular is as , which is not used

with the Future tense (viii. i. 43). Thus, —

3STTW w^ Jcarfin chu-s , I make.

- j '

UfW pgku-s, I went.

karyo-n-as , I (as) was made ( karyo ) by him (an). I.e.,

he made me.

But ^ kara, not kara-s, I shall make. kara-s means { I

shall make for him ’ (vide post, third person).

For other cases of the singular , the suffix is am (viii, i. 24).

Thus, —

qr^l«T W*T karan chu-m , he makes me, or for me.

~ vj v

fw*l karan chi-m , they make me, or for me.

karu-m , made by me, I made.

There are no Plural Suffixes of the first person : the full pronouns

are used instead (viii. i. 44). Thus, —

W5^ ^5|^J karan chuh ase, he makes us, and so on.

J. i. 3

18

G. A. Grierson — On the Kdgmiri Verb.

[No. 1,

Second Person.

For the Nominative Singular the suffix is ^|*sr akh , which is used as

follows (viii. i. 36). Thus, —

^TCPsr karan chhu-kh , thou makest.

kara-kh, thou wilt make.

N

karahd-kh , (if) thou hadst made.

■q<pef pgku-kh , thou wentest.

kgru-m-akhf thou wast made by me, t.e., I made thee.

karyd-n-akh, thou wast made by him, i.e., lie made thee.

For the Accusative Singular , ath is used in the first person

singular and plural (viii. i. 37). Thus, —

kara-tli , I shall make thee.

X '

karo-th ( karav + ath ), we shall make thee.

^T«T crro karan chu-s-ath , I make thee.

^»TT«T karan chi-th , we make thee.

We cannot use this suffix with the third person. Thus, we cannot

say kari-th , he will make thee. We must use instead the suffix

^1*7 ay, which properly belongs to the Dative. Thus, —

*

suh kari-y , he will make thee, or for thee.

ffHf srst* tim karan-ay , they will make thee, or for thee.

3RTT«T WQ suh karan chu-y , he makes thee, or for thee.

tim karan chi-y , they make thee, or for thee.

The same suffix ( ath) is also used for the Agent Singular with

the past tenses of transitive verbs. Thus, —

karu-th , made by thee, i,e.t thou madest.

karyo-th , made by thee, i.e ., thou madest.

For the Dative Singular and also (when the verb is in the third

person) for the Accusative Singular , the suffix 5PT ay is used (viii. i. 40).

Thus, —

^1*1 karan chu-y , he makes for thee, or thee.

* ,

karan chu-s-ay , I make for thee.

fw*7 karan chi-y , we make for thee.

1899.]

Pronominal Suffixes— Third Person.

19

fire karan chi-y , they make for thee, or thee.

^ifre kari-y , he will make for thee, or thee.

This form is liable to certain changes in the Aorist and Pluperfect

tenses of verbs. These will be found duly explained in the proper

place. With the Past Conditional this suffix may have the force of the

Accusative.

For all cases of the Plural, the suffix is ^ wa. Thus, —

Nominative —

fire karan chi-wa, you make.

Accusative or Dative —

karan chu-wa , he makes you, or for you.

^Tf«T fire karan chi-wa , they make you, or for you.

3RTT«T karan chu-s-wa, I make you, or for you.

Agent —

kgru-wa , made by you, you made.

Third Person.

There is no pronominal suffix of the Nominative Singular or Plural.

The pronominal suffix of the Dative Singular is as (viii. i. 38).

Thus, — (masculine).

^T«f WW karan chu-s- as , I make for him.

karan chu-h-as (chu-kh + as, thou makest for him.

karan chu-s, he makes for him.

fire karan chi-s, we make for him.

fire^T karan chi-wa-s , you make for him.

fire karan chi-s, they make for him.

3Pre kara-s, I shall make for him.

So also for the feminine, ^CT«T karan che-s-as.

[This suffix is also used for the Accusative , when the verb is in

the third person. Thus, «re karan chu-s , he makes him,

fire karan chi-s, they make him].

For other oblique cases of the singular, ^«T an is used (viii. i. 34).

Tli us, —

Accusative —

3\TT«T karcin chu-h-an (chu-kh -f- an) , you make him.

*^T«T khyawan chu-h-an, you eat him.

20

G. A. Grierson — On the Kdgmirl Verb. [No. 1,

^ as is however, used for «Hif an with the third person. Thus,

^Ti«r kardn cha-s , not kardn chu-n, he makes him.

' s* ' - \t ~

Agent —

Jcgru-n ( ^ an), made by him, he made.

For all cases of the plural, the suffix is akh (viii. i. 35).

Thus, —

kardn chu-s-akh, I make for them, or I make them.

kardn chu-kli , he makes for them, or makes them.

karu-kh, made by them, they made.

Moods and Tenses.

The Ka^miri verb has four Moods, viz., the Indicative, the Impera¬

tive, the Benedictive, and the Conditional.

The Indicative Mood is usually credited with eight tenses, viz., —

1. The Present.

2. The Imperfect.

3. The Future.

4. The Past.

5. The Aorist.

6. The Pluperfect.

7. The Perfect.

8. The Periphrastic Pluperfect.

Of these, Nos. 2, 7, and 8, are not discussed by lyvara-kaula in his

grammar. I shall, however, give short notices of them for the sake of

completeness. Nos. 1, 2, 7, and 8, are all periphrastic tenses, made up

of Participles conjugated with auxiliary verbs. In No. 1, the Present

participle is conjugated with the Present tense of the auxiliary verb,

and in No. 2, it is conjugated with the Past tense of the same. In

No. 7, the Past participle is conjugated with the Present tense of the

same verb, and in No. 8, with its Past tense.

The Imperative Mood has three tenses, viz., —

1. The Present.

2. The Future.

3. The Past.

The Present Imperative has two forms, a Simple, and a Modified.

The Benedictive Mood has one tense, which may be called the

Future.

The Conditional Mood has two tenses, viz., —

1. A Present-Future. This is the same 2. A Past,

in form as the Future Indicative.

1899.]

Gender , Number and Person.

21

Other tenses may be manufactured on the analogy of Hindi, but

they are not in frequent use. Thus, ^ 3iTT«T suh kardn asi,

equivalent to the Hindi ^ ^fTJTT wall kar'td lioga , he will (pro¬

bably) be making. It is unnecessary to make a list of these. They

can be made up as required.

Some verbs are irregular in the use of their tenses. These are the

following.

The root sust zan, know, when it means to know how to do a thing,

uses the Future in the sense of the Present (viii. i. 58). Thus, —

^nfsr karun zani , he knows how to make a thing.

gjT5f*r parun zdnan, they know how to read.

So also, snfiT syathdh zani , he knows a great deal.

grrfa vidya zani, he is a learned man (lit. he knows

knowledge).

In the same way, in writing the ceremonial part of a letter, a past

tense is sometimes used instead of the imperative. Thus, tfiro

tamis lyukhu-th sonu namaskdr , to him was there written-

by-thee our compliment, i.e., write our compliments to him.

The root ttjvj path, be, become, has no regular Present ; and uses

the Future for that tense (viii. i. 59). Thus, —

patha, I am. xrrsjg paihaw , we are.

pathakli , thou art. Trifag pgthiv, you are.

*rrf^T pathi, he is. *TTV1«T pathan , they are.

For the Past tense, the following forms are used.

1 xnW pdthahg, I was. pathahav, we were.

2 pathahakh, thou tjf'prifa patliGilv, you were,

wast.

3 pdthihe, he was. pathahan, they were.

These forms are properly those of the Past Conditional. This verb

has no verbal nouns.

Gender, Number and Person.

Gender.

The verb has two Genders, Masculine and Feminine. The Future

Indicative, and the Imperative, Benedictive, and Conditional Moods,

do not however, make any change for Gender. Their Masculine aud

Feminine forms are identical.

22

G. A. Grierson — On the Kagmiri Verb. [No. 1,

Some verbs are conjugated only in the Feminine. They are all

impersonal. They are, —

tsar, be inwardly wrathful (viii. i. 45).

WST phog, be inwardly wrathful.

pinch, be inwardly wrathful.

marts, be inwardly wrathful.

wuts, be burnt.

phits, forget.

<*7^ tyamb, look eagerly, (viii. iii. 45).

tsuv, quarrel (viii. iii. 9). (The3e tw0 are feminine and

° < impersonal in the past

*frw morav, bear pain. ( tenses 0nly.

The peculiarities of these verbs will be found described in the

proper places. The first six form one group, which is known as the

isaradi, or ‘ tsar and the others,’ which will be frequently met

with in the course of this article.

Number. There are two numbers, — singular and plural (viii. i. 4).

There is no dual. frUT fw^f *PG«T tim chih paran, means ‘they two,’ or

4 they (many) are reading.’

Person. There are three persons, — first, second, and third (viii.

i. 3).

The first person is more worthy than the second, and the second

than the third (viii. i. 5).

Thus, suh ta tsah pariv , do thou and he read

(imperative).

fT Mh parav , let thee and me read.

fT suh ta boh parav, let him and me read.

^ <T suh ta tsah ta boh parav , let him and thee and

me read.

Auxiliary Verbs, and Verbs Substantive.

There are many verbs meaning 4 to be ’ in Kagmiri. The following

are the two commonest forms, and they are used not only as verbs sub¬

stantive, but also as auxiliary verbs.

1899.] Auxiliary Verbs , and Verbs Substantive. 23

Present. I am (viii. i. 11, 12).

Negative form chu-s-na , &c., see p. 11.

Interrogative form chu-s-d, chn-sa , &c., see p. 12.

Negative-interrogative form chu-s-na , see p. 14.

Emphatic form W^frT chu-s-ti, see p. 14.

Emphatic-interrogative form W^n*JT chu-s-ty-d, see p. 14.

Past. It was.

Negative form ^fT^*r osu-s-na , &c., see p. 11.

Interrogative form osu-s-d, &c., see p. 12.

Negative-interrogative form dsu-s-ndi &c., see p. 14.

Emphatic form dsu-s-ti, &c., see p. 14.

Emphatic-interrogative form dsu-s-ty-d , &c., see p. 14.

21

G. A. Grierson — On the Kdpniri Verb. [No. 1,

These verbs take the usual pronominal suffixes. The following are

examples.

suh chu-m , he is to me, est milii , I have him.

^ ^ > ' ' '

^ sa che-y , she is to thee, thou hast her.

frf*T f^r tim clii-s, they are to him, he has them.

suh osu-m , he was to me, I had him.

*3 sa ds^-s, she was to him, he had her.

Any other tenses required of the auxiliary verb are formed (like

the past) regularly from the root as, be. Thus, Future dsi,

he will be. It is unnecessary to give these forms.

1899. J

Verbal Nouns.

25

The

CONJUGATION

OF THE

ACTIVE VERB.

Verbal Nouns.

The Infinitive. This is formed by adding <g«T un, tm®, or ^r

*

anu to the root. It is an abstract noun. Its formation is fully described

in the chapter on Primary Suffixes. See Nos. 16, 17, 18 ( ante Vol. lxvii,

pp. 202 and ff.) Examples are karun , karunu , and karan w,

to make, making (ix. ii. 2, 3).

karun , belongs to the first declension, and the other two to

the second. All are masculine. They are declined as follows.

Singular. Nom. karun .

Acc. karanas.

X

Ag. ^K5T«T karanan.

Obi. <fp(*r karana.

Plural. Nom. karan.

Acc. 3Pjrf«T karanan.

Ag. ^ karanan.

Obi. karan au.

The various cases are used as gerunds. Thus, Dat. farj

N J

karanas kyutu , for making. The oblique form in ^[fa am, is specially

used to indicate intention (ix. i. 18). Thus, *TTfa parani

gaishan chuli , he is going to read ; T*rfa JTfa ranani gauv , he went to

cook ; *§|fa Jlfa kheni gatshi , he will go to eat.

Roots ending in vowels form the Infinitive only in anu (ix. i.

21, 24). Thus, from far khz, eat, khyanu : from fa ci, drink, ^5f

'i . , 4

cyan ®. This applies only to the nominative singular. Thus, Acc.

Sing., khyanas. As usual, fa ni, take; fa di, give: and fa yi,

J. i. 4

, i

karun u or ^fT*T karanu.

Not used.

3\Tfa karW.

x **

3Pfa karani.

karW.

Not used.

26

G. A. Grierson — On the Kagmlrl Verb.

[No. 1,

come, are exceptions. Their Infinitives are fa«T (*^j) nyunw , Acc. Sing.,

fa*I^ ninas , or fafa^ ninis , and so on for the other two.

The following verbs have feminine infinitives. They never use

the ordinary masculine forms (ix. ii. 24, 25).

^ tsar, be inwardly wrathful. Infinitive Marin, to be so.

^ tsuv, quarrel.

ift^T morav, bear pain.

marts, be impatient.

>>

tsuvin, to quarrel.

morav^n, to bear

A*"

pain.

,, marts^n, to be

impatient.

The verbs {ch°s, pluck the hair ; phits, forget ; pbog,

be inwardly angry ; phuh , be inwardly angry; ivazav , moisten ;

and <|^ wuLs, be burned, may either have a masculine infinitive in

<3«T un, etc., or a feminine one in epR (ix. ii. 25, 26). Thus, W

kasun or *3^^ kasuh, to pluck the hair ( sensu obscoeno). A great many

other verbs also optionally form abstract nouns of the feminine gender.

These are all described under the head of primary suffixes.

Nouns of Agency. There are three forms of these. The first is

made by adding ^T3«T awunu, to the root. Thus, karawunu , a

'n ^

doer (ix. i. 25-27). If the root ends in ?; i, ^ aw is inserted, and the

X i is changed to y. Thus, far khi, eat, khyawawun 11 , a doer.

Exceptions, as usual, are fa ni, take; fa di, give; and fa yi, come.

These form their nouus of agency as follows, fa^sj niwawun % a taker,

and so on for the other two. The feminine of karawunu is

karavah, and the noun is thus declined. See article on Primary Su Sixes,

No. 4, (Vol. lxvii, p. 195).

Masculine.

Singular. Norn, karawunu.

Acc. 3XXfa^ karawanis.

Ag. haraivVn \

NX

Plural. Nom. ctf^fa karaivtn *.

Acc. karawunyan.

-Ag- 3X3^ karawanyau.

Feminine.

karawah .

karawahe.

^Tsrfar karawahi.

karawahe.

karawahan.

karaivahau.

1899.]

Verbal Adjectives , or Participles.

27

The second form of the Noun of Agency is made by adding

anwol*, to the root (ix. i. 28-31). Thus, ^snri^i karanwolu , a doer.

An example of a verb ending in a vowel is khyanawolu , in which

^ a is inserted before wolu. As an example of fsr ni, di, and

yi, we may give fsjirTM ninawblu. Sometimes wdlu can be

^ st

added to a feminine abstract noun, thus, zdnwolu , a knower.

• . - j

Regarding these see the article on Primary Suffixes, Nos. 5 and 6,

(Yol. lxvii, p. 196).

The feminine of karamvolu is haranwajen , and

the noun is thus declined.

Masculine.

Singular. Nom. gfr^eft^f karanwolu .

o ~ \i

Acc. Jcaranwalis.

XX •

Ag. karanwal \

Plural. Nom. q\X*rqif% karanwal *.

x ^ •

Acc. karanwdlyan.

Ag. karanwalyau.

Feminine.

haranwajen.

karanwdjene.

karandjeni.

karanwdjene.

karanwajenan.

sRT«prr!5qi5T) karanwajenan.

The third form of the Noun of Agency is made by adding ^cfqr’sf

angrdkh to the root (ix. i. 28-31). Thus, 3PC*nu*f karangrakh, a doer.

The feminine is karangrdkan (vi. 28). It is decliued regu¬

larly. Thus, Acc. Sing., masc. eftT*nTT^W karangrdkas , fern,, qisj

karangrakane. Verbs ending in vowels have the same irregularities

as in the second form. Thus, khyanagrdkh , an eater, f*r«fqr»§

ninagrdkh , a taker.

Verbal Adjectives, or Participles.

These are Present, Future, or Past. The Past Participles are either

Verbal or Adjectival. Verbal Past Participles are the original Past

Participles of the verb, but are not now used as participles. Nowadays

they are only used as bases for the formation of Past tenses. Adjectival

Past Participles are modern formations from the Verbal Past Parti¬

ciples, and are nowadays the only forms used as participles proper.

Verbal Past Participles have three forms, viz., the Past, the Aorist,

and the Pluperfect.

28

G. A. Grierson — On the Kanmtri Verb.

[No. 1,

The Present Participle. This Participle is an active one. It

has been fully described under the head of Primary Suffixes, (No. 1 ;

Vol. lxvii, p. 193). The following resume is given for the sake of con-

venience. It is formed by adding

^fTif an to the root (viii. i. 19 ; ix. i. 2). Thus, tear, make.

Pres. Part. kardn. It does not change for gender or number.

The roots fa ni , take, fa di, give, and fa yi, come, take wan ,

thus fepri^r niwan, diivdn , fa yiwan. Other roots ending in

\ i change the ^ i to ^ ya, and add ^T«T wan. Thus, fa khi, eat, Pres.

Part. khyawdn, fa ci, drink, cyawdn (viii. i. 19, 20; ix i. 3).

When a Present Participle is repeated, it means that the thing is

done frequently (ix. i. 4). Thus fa^fT«T fa^I^ diivdn diwdn gauv,

he kept giving as he went.

The Future Participle. This is a passive Participle, equivalent

to the Latin Future Participle in - endus , or the Sanskrit Participle in

aniyah. It is the same in form as the Infinitive in ^*f un,

^•T unu or anu. Thus, iT^*f yih path chuh parun,

^ ^ l

this lesson is to be read, ayam patluih pathaniyah. qrfa

yih puth 1 cheh paran , this book is to be read, iyam pustika pathaniya.

Tn the case of Intransitive Verbs, ihe participle takes an impersonal

passive signification. Thus, tsalun , it is to be fled, calaniyam ,

rf^T '§r^ tas chuh tsalun, it is to be fled by him, he must flee.

Note that the Agent is always put in the Accusative (which is an old

Dative), and not in the Agent case, as we might expect. For further

particulars regarding the use of this Future Participle, see the article on

Primary Suffixes, Nos. J6-18, (Vol. lxvii, pp. 204 and ff.)

The Future Impersonal Participle. Another impersonal Future

Participle Passive is formed by adding ssfifl ani to the root (ix. i. 50).

It is formed with both Transitive and Intransitive verbs. Thus ?f^

x

tas karani, it is to be done by him, he must do ; ff*? tas

pakani , he must go. See Primary Suffix No. 14 (Vol. lxvii, p. 201).

The Verbal Past Participle. These will be fully dealt with

when describing the past tenses. Suffice it at present to say that the

Past Verbal Past Participle only occurs in the first and second conju¬

gations, and is formed by adding ^ u-mdtrd to the root. Thus, first

conjugation, karu, done ; second conjugation, 4^ tsalu, gone. In

the case of verbs of the first conjugation, it is a Passive' Participle, and

1899.]

Verbal Adjectives , or Participles.

29

in the case of verbs of the second conjugation, which are all intransi¬

tive, it is a Neuter Participle. See also article on Primary Suffixes,

(No. 10; Yol. lxvii, p. 197). This Past Participle refers to something

which has lately happened.

Aorist Verbal Past Participle. This will also be fully dealt with

w’hen describing the Aorist Tense. It is formed by adding yds or

yauv , to the root. Thus, karyds or Jcaryauv, done. The

Participle occurs for all conjugations, with this difference, that in the

first and second conjugations it is an Aorist or Indefinite Past, and

contains no idea of proximity or remoteness of time. In the case of

verbs of the third conjugation, which have no Verbal Past Participle,

it is used instead of that Participle, and refers to something which

has lately happened. See, also, Primary Suffix, No. 11, ( Vol. lxvii, p. 198).

Pluperfect Verbal Past Participle. This will also be fully

dealt with when describing the Pluperfect tense. It is formed by adding

yav to the root. Thus, 3v*rhr karydv, done. In the case of verbs

of the first and second conjugations it implies that the action has taken

place a long time ago. In the case of verbs of the third conjugation

it takes the place of the Aorist Participle, the proper form of which

has been used up for the Past. These verbs have a special form for

the true Pluperfect Participle, which need not be described here.

The Adjectival Past Participle. These are formed by adding the

termination mat w or mutu , to a verbal Past Participle. In the

case of verbs of the first and second conjugations, this is added to the

Past Participle in ^3 u-mdtrd , and in the case of verbs of the third

conjugation to the Aorist Participle in yov or yauv , the final

^ v of which is elided. Both members of the compound thus formed

are liable to changes of inflexion and gender (ix. i. 40). See also

article on Primary Suffixes, (No. 12 ; Vol. lxvii, p. 198). Examples are, —

First Conjugation.

karurnatu, or karnmutu , made ;

Singular.

which is thus declined.

Masc.

Nom. karumat11 or

karumutu.

Aec. karlmntis.

Ag. karumat*.

Fern.

I i T _

karumai&u .

karernatae.

karimatsi.

30

G. A. Grierson — On the Kctgmiri Verb.

[No. 1,

Plural.

Masc.

Nom. kar'mat *.

Acc. Tear' maty an.

Ag. kgrlmatyau.

Fem.

karemats °.

karemafisan.

kcirematsciu.

Second Conjugation.

bovumatu, or «*<pr<T bgvumntu, become.

Singular. Masc. bdvumatu.

x.

_ I \

Fem. bovu.mafis‘a.

=4 ^ •

Plural. Masc. bon wi*at

Fem bovematsa.

' cJ

Third Conjugation.

3Dsfr*W vyathyomat ®, vyathyanmgtu , aror^fT vyathyomut ®, or

'i ' '

vyathyaumut w, become fat, from vyath, be fat. It is thus

declined, —

Singular.

Masc.

Nom. arsfTflH vyathyomat tt, &c.

Acc. vyathyematis.

Ag. vyathyemat l.

Plural.

Nom. ajijHfrT vvyathyemat l.

Acc. ajijHcq«T vyathyematyan.

Ag. vyathyematyan .

Fem. .

vyathyemats *.

vyatliyematse.

vyathyemabsi.

vyathyemats a.

vyathyemats an.

Verbal Adverbs, or Conjunctive Participles.

There are two of these, a Present and a Past.

The Present Conjunctive Participle. This is formed from the

Present Participle by adding i-matra , and modifying the preceding

long o. Thus, from fcar, make, Present Participle, karan,

1899.]

Verbal Adverbs , or Conjunctive Participles.

31

Present Conjunctive Participle, 35Tlf«T kardn *, on making, at tlie time

of making. It is used in sentences like the following 3R;{f«r JTfa-

suh kardn * gauv , he went away as he was doing it. This form is not

mentioned by ^vara-kaula, and, according to my pandit, is only used

by rustics. It is more elegant to use the simple Present Participle.

Thus, suh kardn gauv.

The Past Conjunctive Participle. This form corresponds to the

Sanskrit Conjunctive Participle in tva, or ya tya ), and means

‘ having done so and so.’ It is fully described in the Article on

Primary Suffixes, (No. 2; Yol. lxvii, p. 193), and the more important

information there given is here repeated for the sake of convenience.

It is formed by adding itli or ith kyath , to the root. Thus,

karith, having made, parith , having read. So also

karith kyath , having made, and ^EJ^[ parith kyath , having read

(ix. i. 5, 6).

Before this suffix a radical d is modified, a radical ^ e becomes

t; l, and a radical ^Tt o becomes ^ u (ix. i. 13-15). Thus, —

From —

kdr, boil.

iTTT mar , beat.

«TTT tar, pass over.

tset, pound to powder.

lekh, write.

[So also —

byah , sit.

t£\^ khdts, fear.

bdz, hear.

rfW tol, weigh.

Roots ending in vowels,

follows (ix. i. 7, 8).

klii, eat.

ci, drink.

karith, having boiled.

mgr ith , having beaten.

fftfVvj tdrith having passed over.

tsitith , having pounded to

powder.

llkhitli, having written.

bihith, having sat].

khuisith, having feared.

*rf5T® buzith, having heard.

tulith, having weighed,

rm their Conjunctive Participle as

khycith, having eaten.

cydth, having drunk.

32

G. A. Grierson — On the Kagmirl Verb.

[No. 1,

f% hi, take,

fsr zi, l)e born,

fq pi, fall.

hydth, having taken.

zydth , having been born.

TZ}?I pydth, having fallen.

But —

f ^ di, give. ft*j dith, having given.

f*T ni , take. nith, having taken,

ft yi, cojne. ft^r yith, lia\ing come.

So also, Ichyath kydth, &c.

The following are irregular, —

wud, fly, when it means to obtain salvation ( mdJcsa ) makes

wujith. Thus, Jrft wnjith gauv, having obtained salvation,

he went, i.e., he went to heaven, but ffeiq JfiT wudith gauv , having

flown, he went, i.e., he flew away (ix. i. 9).

The causal verb flq’Cfq masardw, cause to forget, makes

mdsawith , when the object forgotten is death. Thus,

marun mdsawith, having caused to forget death. If anything else

is forgotten, it is masardivith, regularly (ix. i. 10).

When the Conjunctive Participle is repeated, to imply continuous

action, the th is dropped, and the final i becomes i-mdtrd (ix. i. 1 1, 12).

Thus, —

Wft qrf> kar' Tear*, having made, having made, i.e., having

kept making.

srftj «rftj buz', buz', having kept hearing.

kar' kar', having kept boiling.

rr!ft tar ' tar', having continued passing people over.

Verbs ending in vowels optionally retain the ^ th. Thus, ^

khe khe, or khyath khydtli , having kept eating, ft ft di di, or

ft*t ft*I dith dith, having kept giving.

Negative Conjunctive Participle. This is formed by adding

'qjcpr anay to the root (ix. i. 51). Thus, karanay, not having

done. See Primary Suffix, No. 1-1, (Vol. lxvii, p. 201).

1899.]

Present Indicative.

33

A. Indicative Mood.

1. Present Tense.

The same form is used both for the Definite, the Habitual, and the

Indefinite Present (viii. i. 7-10). Thus, VTPT ^ parcin chuh, he is

reading, or he is in the habit of reading, or he reads. So we have, —

SfJTrre TWT«T Igwar zagatas rachan chuh, God protects

the world.

TTI*r VT»P9 WT5T pdna pdnas radian chuh , by himself (i.e.,

by his own power) he protects himself. Atmana atmdnam

palayati.

vyakaran parcin chuh ldkutu, the boy is

reading (i.e., has arrived at that stage of his studies)

grammar.

jITT^r ^WT«I gaygaya gatshdn chuh prdth kumbas.

He visits (is in the habit of visiting) the Ganges at every

kumbha festival.

^PCr«T kydh chuh murbs kardn, does he make

images ? (is that his profession ?)

c»TT«T kyali chuh jdn lekhdn , is the holy man

writing (and going on writing) ?

rcrer«T kyah chuh rHu gyawdn , is the good man

singing ?

This tense is found by adding the Present Tense of the auxiliary

verb to the Present Participle of the principal verb, which does not

change for gender or number. It is therefore thus conjugated.

I make , or am making.

J. i. 5

34

G. A. Grierson — On the Kagmirl Verb. [No. 1,

Negative form, Tcarctn chusna , I do not make, &c., see

p. 11.

Interrogative form, W^JT karan chusa, 3Pn«T fg karan chusa ,

&c.j do I make ? see p. 12.

Negative Interrogative form, ^\TI*T W*7«TT karan chusna , &c., do I

not make ? see p. 14.

Emphatic form, karan chusti , &c., 1 do indeed make,

see p. 14.

Emphatic Interrogative form, ^rrr«T WWIT karan chustya , &c., do I

indeed make P see p. 14.

In this, and in other periphrastic tenses, it is elegant to put the

auxiliary before the verb (viii. i. 22, 23, 24) when standing in a sentence.

Thus, —

^ Tfj^T bata clmh suh khyawan, he is eating rice, is

more elegant than <^]«T hata suh khyawdn

chuh , though both are correct.

ff^f ifcf *jn?rr tawa pata chuli asanas

pyath bihith puza karan , after that; he is doing worship

having sat down on a seat, is more elegant than *T<T

f%fV?T ^Tr 3»Tf«T taiva pata asanas pyath

bihith puza karan chuh , though both are correct.

So the following is the most elegant order, — WT3T«T

igwar chuh asdn kagiya andar , God exists in Benares. When,

however, the verb stands by itself as in the paradigm, the auxiliary

always follows.

The following are examples of the use of pronominal suffixes,

(pp. 15 and ff ) —

«p? ^5Tf«T boh chus-ath karan , I make thee (viii. i. 37).

^*TT«T as1 chi-th karan , we make thee,

boh chus-an karan, I make him.

^ boh chus-as karan , I make for him.

^ 3T^l*T boh chus-awa karan, I make you.

boh chus-akh karan, I make them.

Similarly for the other persons, which are all regular, except

suh chu-s karan , he makes him, or for him (not W*[ chu-n) (see

p. 19).

1899.]

Imperfect Indicative.

2. Imperfect Tense.

35

This tense is not described by Igvara-kaula. It is formed exactly

like the Present, except that the Past tense of the Auxiliary Verb is

used instead of the Present. Thus, —

I was making.

Singular.

Plural.

The various adverbial suffixes are added as in the case of the

Present Tense. They will be found in detail under the paradigm

of the auxiliary verb, and need not be repeated here. One example

will suffice. kardn osusna, I was not making. As in the

case of the Present, in a formal sentence, it is more elegant to place the

auxiliary before the present participle. Thus, buta

dsu suh khyawdn , he was eating rice.

Pronominal suffixes are added as in the Present. Thus, —

boh osus-ath kardn , I was making thee.

^TT«T as* dsi-th kardn , we were making thee.

boh osus-ay kardn , I was making for thee.

boh osus-an kardn , I was making him.

W ^Ti^T boh dsus-as kardn , I was making for him.

boh osus-awa kardn , I was making you.

boh osusakh kardn , I was making them.

suh dsu-s kardn, he was making him or for him.

(not osu-n).

And so others.

36

[No. 1,

G. A. Grierson — On the Kagmirl Verb.

3. Future Tense.

This tense is conjugated as follows (viii. ii. 28).

Singular. Plural.

1 qx Tear ®, I shall make. karav , we shall make.

2 *sx^ karakh, thou wrilt make, ^fx^r kariv , you will make.

3 kari, he will make. karan, they will make.

Roots ending in vowels insert ^ m in the first person of both num¬

bers. Before X i, they insert ^ y . As usual, all roots ending in x »*,

except f*r ni, take, fx di, give, and f? yi, come, change the final x i

of the root to ya (29). We thus get for the future of khi, eat, —

Singular.

1 khyam*.

2 khyakh.

3 kheyi.

From fx di, give, we have, —

Singular.

1 dima.

2 dikh.

3 fxfa diyi.

Plural.

khyamav.

kheyiv.

khyan.

Plural,

dimciv.

diyiv .

f^iT din.

Plural.

karaw*.

kariiva.

*tX*T karcina.

This tense has a special interrogative form, which is thus conju¬

gated (30, 31).

Singular.

1 3Xr kara , shall I do or make.

2 c|n;*5r karakli °.

3 karyd .

Pronominal suffixes are added to this tense as usual, except that

the third person singular is slightly irregular, being ^4^ karyam (he

will make me), &c., not karim , &c. (viii. ii. 29). So karyas ,

not karis, and all other suffixes added to this person of this tense.

When, however, the suffixes of the second person is added there is no

irregularity. Thus, s&fx^ kariy, not karyay. Remember that final

1899.]

Future Indicative.

37

^ kh becomes ^ h before a suffix. We thus get the following forms

which will do as samples.

Added to first person, —

kara-s, I shall make for him.

kara-n , I shall make him.

ejnCW kard-s , we shall make for him.

x 7

karo-n , we shall make him.

Added to second person, —

karah-as , thou wilt make for him.

karah-an , thou wilt make him.

karyii-s, you will make for him.

Added to third person, —

kary-am , he will make me.

kari-y , he will make thee or for thee. (It should be

remembered that the suffix ath is not used with the

X

third person).

kari-w° , he "will make you or for you.

kary-as, he will make him, or for him. an is not

used with the third person).

^*4*3 kary-akhj he will make them or for them.

karan-as , they will make him or for him.

karan-akh, they will make them or for them.

Interrogative forms would be such as kara-s-0- , shall I make

for him ?

The feminine impersonal verbs tsar, to be inwardly wrathful,

&c. (see pp. 16 and 22), are thus conjugated, this tense making no

distinction between masculine and feminine.

iHni tsary-am, there will be inward anger to me. I shall be

inwardly angry.

ase tsari , we shall be inwardly angry.

tsari-y, thou wilt be inwardly angry.

isari-w0, you will be inwardly angry.

38

G. A. Grierson — On the Kagmiri Verb.

[No. 1,

tgary-as , he will be inwardly angry.

tsarycilch , they will be inwardly angry.

On the Past Tenses Generally.-

Ka^mirl has three Simple (as distinct from Periphrastic) Past

Tenses, — the Past, the Aorist, and the Pluperfect.

There are three conjugations of verbs in these tenses. The first

conjugation includes all active and impersonal verbs.

The second conjugation includes the following neuter (viii. iii.

77-97) verbs, —

1899.]

39

On the Past Tenses Generally.

belongs optionally (viii. iii.

91) to the first conjugation).

36. sor , be spent, exhausted

(of things).

37. gal , melt.

38. tsal, flee, escape.

39. ^ dal , pass over.

40 dot , be unused.

41. W phal, bear fruit, become

old (of clothes), be divided.

42. ^ pholl , expand (of a flower).

43. mel , meet, be met by a

person.

44. bov, be born, grow.

45. rdv , be lost.

46. ^ dos , trickle.

47. tfn* pos, be competent, vic¬

torious.

48. mas , forget.

49. ros , be angry.

50. as, be.

51. ^ khas, mount, ascent.

52. phas, be entangled, caught.

53. has, d^ell.

51. las, live long, live in good

health.

55. Ids, be weary.

56. was, come down, descend.

57. byah, sit.

Of the above, the verb rot, be stopped, may optionally be con¬

jugated in the third conjugation (viii. iii. 84). Thus, —

2nd Conjugation. Past Participle, ^ rut*

Aorist ,, rocyov.

Pluperfect „ rocydv.

3rd Conjugation. Aorist „ in sense of Past rotyov.

Pluperfect ,, „ Aorist rotyav.

True Pluperfect ,, rotiydv.

The verbs wup, burn inside, and ^ grap, be digested, optionally

form their Pluperfect after the manner of the third conjugation (viii.

iii. 89). Thus,—

2nd Conjugation. Pluperfect Participle gnjT* ivupydv,

’spzfl^ grapyav.

3rd Conjugation. True Pluperfect Participle wupiyav,

grapiyav.

The above 57 verbs are those given by I^vara-kaula in the Bhuta-

pada of the Akhyata-prakriyd of his Grammar (viii. iii. 77-97), but the

following are added to the second conjugation by my Pandit.

The third conjugation includes all other neuter verbs, except those

used impersonally.

I shall now describe each of the Past tenses separately.

4 (a). The Verbal Past Participle.

This Participle, from which, as will be seen, the Past Tense is

derived, is formed by adding ^ u-matrd to the root. Thus, Tear , do ;

Past Participle Tear11. Its feminine singular is formed by changing

u-matra to u-matra . Thus, kar ™. The masculine plural by changing

the u-matra to i-matra. Thus, kar'1, and the feminine plural by

changing the u-matra to *7 ya (e), thus, ^7 hare. Before the matra-

vowels, the root- vowel of the verb is liable to modifications. But (this is

important) before the *7 ya (e) of the feminine plural, there is no

modification. The modifications caused by matra- vowels are as follows, —

If the root-vowel is a, before u-matra , it becomes a (o). Thus,

kar, make, kar™ , pronounced kor, he (was) made (viii. iii. 19).

If the root-vowel is a, before u-matra , it becomes a (ii). Thus,

kar™, pronounced kiir, she (was) made (viii. iii. 19).

If the root-vowel is a, before i-matra , it becomes a (a*). Thus, *fc

kar'1, pronounced kairi, they (masc.) (were) made (viii. iii. 19).

If the root- vowel is a, before u-matra, it becomes o. Thus,

mar, kill, mor™, he (was) killed (viii. iii. 20).

If the root-vowel is a, before u-matra, it becomes a (o). Thus,

*71*1 mar™, pronounced mor , she (was) killed (viii. iii. 19).

If the root- vowel is a, before i-matra it becomes a (6*). Thus,

mar % pronounced mdlrl, they (masc.) (were) killed (viii. iii. 19).

If the root-vowel is i, before u-matrd it becomes yu. Thus, liv,

plaster, lyuv™, he (was) plastered.

If the root-vowel is i, before u-matra it becomes i (pronounced

yii). Thus, liv™, pronounced lyuv, she (was) plastered.

1899.]

The Verbal Past Participle.

41

If the root-vowel is i, before i-matra, it is unchanged. Thus, fftff

liv\ they (masc.) (were) plastered.

If the root- vowel is i, before u-matra , it becomes yu. Thus, fPtT

ctr , squeeze out, cyiir™, he (was) squeezed out.

If the root-vowel is i, before u-matra it is unchanged. Thus,

clr™, she (was) squeezed out.

If the root-vowel is i, before i-matra it is unchanged. Thus, 'fife

cir1, they (masc.) (were) squeezed out.

If the root-vowel u or u, it always remains unchanged. Thus,

from buz , parch, ^ buz™, ^5? buz™, *rfgj buz1. From but, rob,

VJ* ^ ^ C\ •’ ’

*12 lut™, ^3 lut * *rf2 lut\

If the root- vowel is e, before u-matra , it becomes yu. Thus,

pher, be turned, phyur w, he (was) turned (viii. iii. 21, 22).

If the root-vowel is e, before u-matra , it becomes i. Thus,

phir ™, she (was) turned (viii. iii. 23).

If the root- vowel is e, before i-matra , it becomes i. Thus,

phir*, they (were) turned (viii. iii. 21, 22).

If the root- vowel is o, before u-matra , it becomes u. Thus, fisr

boz , hear, buz™, he (was) heard (viii. iii. 24).

If the root-vowel is o, before u-matra , it becomes u. Thus,

buz™, she (was) heard (viii. iii. 25).

If the root-vowel is b, before i-matra , it becomes u. Thus,

buz % they (masc.) (were) heard (viii. iii. 24).

Before ^ e or ^ a of the feminine plural, the root-vowel is always

unchanged. Thus, sfvf hare , marey ffjsj live, 'ftf cire, buz9,

fif phert, bbza.

When any pronominal suffix except ^ wa is added to a form ending

in a matra-Y owel, that matra-Y owel becomes fully pronounced, but the

modification of the root-vowel remains unchanged. Thus, ejpr kar ™ (pro¬

nounced A:or) + 5T n, becomes karun , pronounced korun , and not Jcarun

or horn. So moru- f «T becomes morun , and so on. In such a

case, u-matra is pronounced as a short German u. Thus, kar™ (pro¬

nounced hiir) *T n , becomes kar™n , pronounced hiiriin. If the suffix is

J. r. 6

42

[No. 1,

G. A. Grierson — On the Kagmlvi Verb.

W wa, the mtitra- vowel remains unchanged. Thus, ^-f-«r kqru + wa, he

(was) made by you, becomes ^55 kqruwa , and is pronounced korwa , and

not koruwa. So kqr™ + ^ wa made (fem.) by you, becomes

kar^-w0, and is pronounced kiiriva. Similarly, when a matra-Y owel is

followed by any other vocalized syllable, it remains a matra-Y owel.

Thus, karu-n-as , not karunas , (was) made-by-him-I. Even

in such cases, l^vara-kaula (as in viii. iii. 3 and 4) writes kqrmva,

likhhva , &c., with the matra-Y owel apparently fully pronounced,

but in this he is, according to my Pandit, incorrect, and it is a mere

slip of the pen.

When, in the feminine plural, ya (e) follows one of the letters

Tf ts, ^ tsh, 5T z, or ST n, the ya becomes a (viii. iii. 47). Thus,

dits + n ya becomes ditsa , they (fem.) (were) given. So

hets+v ya, becomes hetea , they (fem.) (were) taken. So

37^ wats + q ya becomes wdtsa , they (fem.) arrived, ^5T daz + q

ya becomes daza, they (fem.) were burnt (viii. iii. 7, 12).

So also ^ y is often elided after ^ s (viii. iii. 13). Thus, ds-f

ya, becomes dsa, they (fem.) were. ^ ya , ^ basa,

they (fem.) dwelt. Icvara-kaula restricts this to verbs of the 2nd

conjugation, but, according to my Pandit, ?fy is also elided after all

verbs of the 1st conjugation, except in the case of teas , laugh loudly

(impersonal). Thus, from has , fry, feminine plural afi^r kasa , but,

from ^ teas, tease.

In the case of the following verbs of the 1st conjugation, the

y may, according to my Pandit, be optionally retained, arr^ thds

bury ; ^ das , beat ; «g^ mus, eat improperly ; ^ sas, break wind.

Thus, thasa, or thase.

It will be seen that the feminine of this participle ends in il-mdtrd,

in ya (e), or in a. Before these terminations, the final consonant of the

Past Participle undergoes certain changes. These changes only occur

in the first and second conjugations. They are as follows : —

Final or k, is kh , and jt g, become ^ c, w ch, and sr j respectively,

before both 37 ii-matra and q ya (e) (viii. iii. 7). Tims, —

43

1899.]

The Verbal Past Participle.

From Root thak , be weary ; thgku, he (was) weary ; but thacu ,

^ A

she (was) weary; thace , they (fem.) (were) weary.

„ ,, leJch, write; lyiikhu , he (was) written; but

^ a

lick”, she (was) written; leche , they (fem.) (were)

written.

„ „ dag, pound ; dag u, he (was) pounded; but dgj ®,

she (was) pounded; c^3£f daje, they (fem.) (were) pounded.

3 t, 3 th, and 3 d, become ^ c, w ch, and sr j respectively, before

ya ( e ), but not before u-matra (viii. iii. 70). Thus, —

From Root phat, be split ; ^3 phatu, he (was) split; <?3 phat®, she

'i A

(was) split ; but phace , they (fem.) (were) split.

„ ,, ?i3 mas, forget ; «T3 math®, he (was) forgotten; i?3 math ®,

>£ A

she (was) forgotten; but mache , they (fem.) (were)

forgotten.

i* i*

„ j> ate gand , bind ; JT3 gand ®, he (was) bound ; JT3 gand ®, she

• s» a

(was) bound; but aNj ganje , they (fem.) (were) bound.

c[ t, vi th, ^ d, and «T n, become ^ £s, if tsh, ST z, and st n

respectively, before both u-matra, and before ya ( e ), which latter then

becomes a by the rule given on the last page (viii. iii. 72, 73). Thus, —

From Root 3»?r kat, spin; ^3 katu, he (was) spun; but 35^ kats ®, she

(was) spun; katsa, they (fem.) (were) spun.

woth, arise ; wothu , he arose ; but wo^sh ®, she

^ A

>>

)»

>>

arose, and woisha, they (fem.) arose.

^ lad, build; ladw, he (was) built; but Igz ®, she

A

(was) built, and *n?T laza, they (fem.) (were) built.

ran, cook ; rgnu, he (was) cooked ; but ^ ran®, she

s» ’ A

(was) cooked, and rana, they (fem.) (were) cooked.

^ l becomes m j before both u-matra and y9 ( e ) (viii. iii. 74).

Thus, —

From Root ^ Ual , flee ; tsal®, he fled ; but Uaj®, she fled, and

'i ' A

feaje, they (fem.) fled.

44

G. A. Grierson — On the Kdgmiri Verb.

[No. t,

The verbs pih, grind ; muh , deceive ; sah, bear ;

gah, grind ; and tsah , snck, under similar circumstances change their

final ^ h to W p (viii. iii. 75). Thus, —

pyuhu, he (was) ground; but fair pi(fr she (was) ground;

and fqm pig$, they (fem.) were ground.

Combining what is said here, together with what has been said

above (pp. 40 and ff.) regarding vowel changes, we come to the three

following general rules, all of which apply only to the first and second

conjugations, and do not apply to the third.

1. In the Past Participle masculine, both singular and plural, only

the root vowel, and not the final consonant, is liable to change.

2. In the Past Participle feminine singular, both the root vowel and

the final consonant are liable to change.

3. In the Past Participle feminine plural, only the final consonant ,

and not the root vowel, is liable to change.

These three rules are most important, as the whole scheme of

conjugating the Past Tense depends upon them.

Irregular Past Participles.

The following verbs have irregular Past Participles. The irregu

larities are, of course, carried through the Past tense.

1899.]

Irregular Past Participles.

45

# las, is sometimes used regularly by the vulgar, thus, lasu, &c. (viii. iii.

96). In the fern. sing. Iqvara-kaula sometimes writes lutphu, and sometimes

lut&lx*, According to my Pandit, the latter is the correct form,

46

G. A. Grierson — On the Kagmlrl Verb.

[No. 1,

PAST PARTICIPLE.

1899.]

The Past Tense.

47

4 (h). The Past Tense.

This tense is peculiar to the first and second conjugations. It does

not occur in the third conjugation. It describes something which has

happened lately, e.g., Icarun, he has (just) made.

It is formed from the Past Participle in ^ ™. This Participle is

Passive in the case of verbs of the first conjugation. Thus, kgr™ ,

done, as™ (impersonal verb), laughed. It is Neuter in the case of

verbs of the second conjugation. Thus, bov™, become. Verbs of

the first conjugation are therefore passive in construction. Instead

of saying ‘ I made this,’ we must say ‘ this was made by me,’

2RJ yili me kgr ™, this ( yih ) by me (me) made ( kgru ). Here

the original object of the sentence has become the grammatical subject,

and the logical subject has been put into the case of the agent. If the

object (grammatical subject) is feminine, the participle must be femi¬

nine. If it is plural, the participle must be plural. Thus, —

i&j kgr ™ me kgr™, the bracelet (masc.) was made by me.

jrj pat™ me kgr ™, the tablet (fern.) was made by me.

kgr 1 me kgr % the bracelets were made by me.

xr^T JSJ 'pace me kare , the tablets were made by me.

As regards the original subject, it is put in the agent case, and

may also be added to the verb in the form of a pronominal suffix of the

ageut case. This must always be done in the second person singular

and plural. These suffixes are given on pp. 15 and ft. For the sake

of ready reference, they are here repeated.

The initial a is elided when following a vowel.

When the pronominal suffix is added to a verb (it is always added in

the case of the second person), the original subject (now agent), if a

pronoun, may be omitted. Thus, for 1 he made,’ we may say, —

48

G. A. Grierson — On the Kagmiri Verb.

[No. 1,

(a) rlffl tarn1 karu , by him made,

or ( b ) rlfifl ^«T tam1 karu-n , by him made-by-him,

or (c) karu-n , made-by-him.

Tn the second person we cannot use the first or (a) form. We can

only say,—

(b) ^ tie karu-th , by thee made-by-thee,

or (c) karu-th , made-by-thee.

So in the plural.

( b ) tohe karu-wa, by you made-by-you,

or (c) karv-wa, made-by-you.

TTT i i

We cannot say ^ tse karu or tolie karv.

This Passive construction is called in Sanskrit Grammar the kar-

mani prayoga, and is described in Ipvara-kaula’s Grammar under that

name (viii. hi. 3).

The Impersonal Construction.

Impersonal verbs are used passively with all persons. The termi¬

nations are the same as in the case of active verbs. An example is, —

(a) clfn tam1 asu, by him it was laughed,

or (b) cj'pT tarn* asu-n , by him it was laughed-by-him,

or (c) asu-n , it was laughed-by-him,

all three of which mean ‘ he laughed (viii. iii. 3). So ctsu-m , I

laughed.

The two verbs ^ tsuv , quarrel, and fmc* morav, bear pain, may

be used impersonally, and are then put in the feminine (viii. iii. 9).

Thus, tsuv^n, quarrelling was done-by-him, he quarrelled,

the verb agreeing with the nominative VC har , a quarrel, which is

feminine. So morav^n, pain was suffered-by-him, he suffered

pain, the true nominative being some feminine word like pld, pain,

understood.

In these cases we may use all the three forms, except, as before, in

the second person. Thus, —

(a) crfn ^«T tam 1 tsuv ”,

' «x<\ ' “ ’

or (6) rtffr ^3*1 tam 1 tsuv^n.

or (c) tsuv^n,

1899.]

Past Tense. Neuter Verbs.

49

These two words can even be used in the fem. plur. Thus,

1 O 9

he quarrelled (many times) ; moravyan, he suffered pains.

The following verbs (repeated from pp. 16 and 22) are also imper¬

sonal but have this peculiarity that they are conjugated with suffixes of

the dative instead of suffixes of the agent, and are also in the feminine.

tsar , be inwardly wrathful.

phop, ditto.

phuh , ditto.

marts , ditto.

wuts, be burnt.

pints , forget.

tyamb , look eagerly (conjugated in the third conjugation).

Thus, tsar^-s, not ’rW*T £sar®-w, there was inward anger to him,

he was angry.

•fhpr isar^-y, not tsar^-th, thou wast angry.

tsar^-m, I was angry.

So wuts^y, thou wast burnt. These verbs are given here,

because the first six are used in the Past Tense. tyamb , look

eagerly, belongs to the third conjugation, and does not use the Past

Tense.

This impersonal construction, whether with the Dative or with

the Agent, is called in Sanskrit, the bhcive-praydga.

Neuter Verbs.

Neuter verbs of the second conjugation, having their Past Participle

neuter, are conjugated, in the Past tenses actively. They thus take the

pronominal suffixes of the nominative, and agree with the subject of the

sentence in number and person. The addition of the pronominal suffixes

is obligatory.

They are here quoted for ready reference.

Singular. Plural

1. as. _ .... -

2. akh. aiva.

X

3. - -

There are no nominative suffixes of the first person plural or of the

third person.

J. i. 7

50

G. A. Grierson — On the Kagmiri Verb.

[No. 1,

Taking, therefore, the verb bov , become, for ‘ I became ’ we may

say,—

(b) «j^f boh bovu-s,

or (c) bovu-s.

We cannot say «T^ ^ boh bovu , any more than we can say qfa

tohV karu.

These preliminaries being understood, I shall, in the paradigms,

only give the (c) forms of the past tenses. From these the (6 ) forms

and (when they exist) the (a) forms can easily be deduced.

A. First Conjugation.

(a) Transitive Verb (viii. iii. 3-9) (c. forms only) ‘ I made,’ Lit.

‘he, she, it, &c., was (were) made by me, you, him, us, &c.’

Verbs ending in vowels are slightly irregular in the masculine

singular and plural, and in the feminine singular, which latter is the

1 E.g. The object of the sentence, which has now become the grammatical

subject, as explained above.

8 Or Icard-m, and so throughout.

51

1899.] Past Tense. First Conjugation.

same as the feminine plural (viii. iii. 6). None of these verbs belong

to the second conjugation.

As usual there are two groups of these verbs, viz., those which

change their final vowel, i, to ya, such as f% eat, and others (vide

p. 1) and those which do not, which are three in number, ni,

take, di, give, and yi, come, of which the last belongs to the

third conjugation. Taking the verb khi, eat, we get.

Similarly is conjugated, fa ci, drink.

52

G. A. Grierson — On the Kilgmtri Verb. [No. 1,

The verb fa take, is further irregular (viii. iii. 6, 34; ix. i. 38).

The verbs fa i hi, take, and di , give, are still further irregular

(viii. iii. 7, 32). Thus, —

1899.] Past Tense. Impersonal Verbs. 53

di, give, is similarly declined, but its first person is or

dyutu-m , dits^-m, diti-m, ditsa-m , and so throughout.

The verb hahar , to get a girl married, is of necessity, conju¬

gated in the past tenses only in the feminine. Moreover it forms its

feminine Past Participle, optionally, in an irregular fashion (viii. iii. 76),

Thus, ^4^ hahar^-n or haraqP-n, he got her married.

' ' ' A'

Notes. (i) Once for all. By the first person singular, is meant,

‘ I made him, her, them (masc.), or them (fern.)/ literally, 4 he, she, they

(masc. ), or they (fern.) were made by me.’ So the second person

singular means 4 thou madest him, her, them (masc.), or them (fern.),’

and so on, through the other persons.

(2) There is no suffix for the first person plural. Hence this

form is always the same as the past participle.

(0 Impersonal Verb (viii. iii. 3-9) (c. forms only), 4 1 laughed,’

literally, ‘it was laughed by me,’ &c.

Singular 1 asu-m , I laughed.

2 asu-th, thou laughedst.

3 asu-n , he laughed.

Plural 1 4^ asu , we laughed.

2 asu,-wa, you laughed.

3 4^3 asu-kh , they laughed.

54

G. A. Grierson — On the Kagmlrl Verb.

[No. 1,

When an Impersonal Verb is conjugated in the feminine (see

pp. 22 and 48), we get forms such as the following, (*^r tsuv, quarrel)

(viii. iii. 9), —

Singular 1 tsuv^-m, I quarrelled.

2 tsuv^-th, thou didst quarrel.

3 tsuv^-n, he quarrelled.

Plural 1 tsuv u, we quarrelled.

2 tsuvii-wa, you quarrelled.

3 isuv^-Jch, they quarrelled.

If such an Impersonal Verb is conjugated in the plural, we get —

Singular I isuvya-m , I quarrelled many times, and so on.

2 tsuvya-th.

3 tsuvya-n.

Plural 1 tsuve.

o

2 tsuvya-wa.

3 isuvya-kh.

So morav, bear pain.

fear and the others are thus conjugated (see pp. 16, 22,

and 49).

Singular 1 tsar^-m, I was inwardly angry, and so on.

2 isar^-y.

3 tsar^-s.

\ \ •

Plural 1 tsar™.

2 tsar^-w*1.

i

3 tsar^-kh.

XX •

In the case of Transitive verbs, when the grammatical subject

(i.e., the logical object) is a pronoun, it may be added to the verb in

the shape of an additional pronominal suffix in the Nominative case.

The following are the masculine forms which occur. The feminine

ones can easily be made on the same principle.

From karu-m , made by me, I made.

karu-m-akh, thou wast made by me, I made thee.

So kar^-m-akh, thou (fern.) wast made by. me, I made

thee (fern.).

1899.]

Past Tense. Pronominal Suffixes.

55

kqru-m-an , I made him, is not used. We always say

suh kg,ru-m. sjUiTO karu-m-as means ‘ I made

for him.’]

kar'-m-aw*1 , I made you.

kqr'-m-akh , I made them, is not used. We always

say tim kqri-m. kar'-m-akh means ‘ I

made them for them.’]

From 35^*1 kqru-th, made by thee, thou raadest.

kqr^-th-as, thou madest me (or thou madest for him).

karu-tli-an, thou madest him.

kqp-th-akh , thou madest them (or for them).

From kqru-n , made by him, he made.

efiX«T^ kqrw-n-as, he made me (or he made for him).

karu -n-akh , he made thee (or he made for them).

kar1l-n-an, he made him, is not used. We say

suh kqrun.~]

kqr'l-n-awa, he made you (or for you).

kqr’1 -n-akh, he made them, is not used. We say

tim kari-n. kqP-n-akh, means ‘ he made them

for them.’]

From karu, made by us, we made.

I

karu-kh, we made thee.

karu-n , we made him, is not used. We say

suh karu~\.

kari-wa , we made you.

|

kqri-kh , we made them, is not used. We say

tim kar *].

From kqrn-wa , made by you, you made.

karu-wa-s , you made me (or for hira)

4^«r*T karu-wa-n , you made him.

kqr'-wa-kh, you made them,

56

G. A. Grierson — On the Kdfmirl Verb.

[No. 1,

From karu-kh , made by them, they made.

kpru-h-as, they made me (or they made for him).

lcaru-h-akh , they made thee (or for them).

karu-h-an, they made him, is not used. We say,

4^*3 snh karu-kh ].

kar*-h-awa} they made you.

[^{TV5[ kar*-h-ahh, they made them, is not used. We say,

qsfc?§ tim kari-kh. qifTVa kar'-h-akh , means ‘ he

made them for them ’].

There being no suffix for the first person plural, there are no

special forms for thou, he, you, or they made us.

With regard to all these forms, the full forms of the pronouns may

also be used, with, or without the suffixes. Thus, —

We may say either

(а) jq me karu-m-akh isah, by me was-made-by-

me-thou thou, or

(б) qr^ffiq kar^-m-akh t£ahf was-made-by-me-thou thou, or

(c) jej 4^*?^ me karu-m-akh , by me was-made-by-me-thou, or

(d) qrwa karumakh , was-made-by-me-thou, or

(e) JRJ me karu-kh tsah , by me was-made-thou thou, or

(/) me karu-kh , by me was-made-thou.

We cannot, however, use the two following forms.

( g ) me karw tsah , by-me was-made thou or

( h ) qfq*j karu-m tsah, was-made-by-me thou.

In other words when the full form of the pronoun in the nominative

is used, the corresponding suffix must always accompany it.

Other pronominal suffixes can similarly be used. Thus, qUff?q

karu-m-ay , I made for thee.

1899.]

57

Past Tense , Second Conjugation.

B. Second Conjugation.

Neuter Verb (viii. iii. 77-97). (c. Forms only).

I became, &c.

Singular. I Plural.

The verb mar, die, has its past tense irregular. It is thus

conjugated (viii. iii. 26, 31, 59, 65, 92),

When mar means to unite, it is regular. Thus, marun,

he united him.

5 (a.) The Verbal Aorist Participle.

This participle occurs in all three conjugations. In the first and

second (viii. iii. 78) conjugations it is a true aorist. It expresses past

time indefinitely, with no reference to proximity or distance. E-g.,

karybv , he (was) made (by us), we made; bovyov , he became.

It is therefore the participle, and the aorist is the tense, of narration

in these two conjugations.

J. I. 8

58

G. A. Grierson — On the Kflgmiri Verb.

[No. 1,

In the third conjugation, it is used instead of the past participle ;

and describes something which has happened lately. Thus,

wuphydv, he flew (a short time ago).

This participle is formed in the first and second conjugations by

changing the final ya ( e ) of the feminine plural of the past participle

to yov. Thus, heir, do ; past participle kgru, fem. plur.

hare; aorist participle haryov.

In those cases, in which the feminine plural ends in a, instead

of in *T ya (e), (see p. 42) the aorist participle ends in or, not in

yov. Thus, daz , burn ; past participle dad w, fem. plur.

^5T daza ; aorist participle dazov. There are four exceptions

(viii. iii. 49, 81). The aorist participle of &a£s, be moist, is

katsyov ; that of vyats , pervade, is vyatsyov ; that of

rots , be preferred, roisyov ; [and that of ^ wiife, be burnt,

wutsyov~\. Note that in the fem. pi. Past, in these verbs, the

y is elided. Thus katsa, not katse ; 3}^ vyaksa, not vyatse ;

rotsa not rottsb (viii. iii. 49, 81).

In the third conjugation, the aorist participle is formed by adding

yov to the root direct. Thus, ^ ivuph , fly, aorist participle

wuphyov, but there are exceptions which will be dealt with later om

l9vara-kaula spells this participle indifferently with yov or with

yauv. Both are pronounced the same, like yov (viii. iii. 39).

[The true termination of this participle is yd, the v, as well as the y

of the plural to be noted later, are only added for the sake of euphony].

Special Buies for the First and Second Conjugations .

The base of the aorist participle of the first and second conjugations

being the same as the feminine plural of the past participle, the final

consonant of the verbal root undergoes certain changes (viii. iii. 69, 70,

71, 72, 73, 74, 75). For the same reason, in the aorist participle, the

vowel of the verbal root remains unchanged (see page 41). The

following are examples of the changes. The reader is referred to

pp. 42 and ff. for details.

1 Root ^ thak, be tired, aorist participle thacyov.

2 ,, lekh, write ,, lechyov.

3 dag, pound „ dajydv .

1899.] Aorist Participle. First and Second Conjugations,

59

Nos. 7-K) are also examples of the elision of x y . The following

are further examples, see p. 42 for details.

1 Root fx di, give aorist participle fxxfa ditsov.

2 „ fx hi, take ,, heteov.

The past participles of these two verbs are irregular, vide p. 45.

3 Root X?T daz, burn, aorist participle x?ft^ dazov.

4 ,, has, dwell

5 ,, kas , fry

But from, —

6 Root 37X thds, bury

11

XX das, beat

n

XXfa basov.

XiXfa kasov.

3TX1X thasov , or 37xflX

thasyov.

dasov, or

dasyov.

n

And from —

8 Root kats, be wet

9 „ ’XX hsas, laugh loudly „

Remember that all these changes occur only in the first and second

conjugations. The rules for the third conjugation are quite different.

only prefix kateyov (p. 50).

only XXflX teasyov (p. 42).

# My Pandit prefers muhydv.

60

G. A. Grierson — On the Kaqmiri Verb.

[No. 1,

Special Bides for the Third Conjugation.

In this conjugation, the participle (and consequently, the tense

formed from it) is not used as an aorist, but as a past ; with the same

meaning as the past participle of the first and second conjugations. It

is not the tense of narration.

The aorist participle is formed by adding *7^ yov to the root.

Thus, wuph, fly, ivuphyov. Beh re this the final consonant of

a root is not liable to change, as it is in the case of verbs of the first and

second conjugations. Thus, from 3T3’ zeth , be long, a verb of the 3rd

conjugation, the aorist participle is zethyov , and not zechydv,

as it would be, if the verb belonged to the 2nd conjugation.

Note also that in this conjugation, *7 y is not elided after ^ t$,

W tsh, si z , or sj n (viii. iii. 49). I^vara-kaula in this sutra gives the

following list of verbs, which do not elide y. It includes many of

the verbs ending in these letters which belong to the third conjugation.

gro^s, be splashed out ; isobs, have insufficient means of

livelihood ; ^ t shots , be empty ; tez, be sharp ; tf^T paz , be fit ;

braz , shine ; ^T^5T bdwaz , be preferred ; lets , be weak ;

prots, be pure. Thus, grotsyov, not TpflS' groisov.

The other verbs belonging to the third conjugation, which end in

these letters, and which are not mentioned by ^vara-kaula in the

above sutra are the following : —

ais , enter; ^ trats , fear; mots , remain over and above;

citsh , be weak ; itm gaish, go ; palaz , be useful ; ranz , be

pleased ; laz , be suitable ; wuz , be wide awake, appear.

^X7«T sapaz , become, is considered the same as sapan , or sapad ,

and belongs to the second conjugation. According to my Pandit

palaz, above recorded, belongs to the second conjugation.

According to my Pandit, of these, ^ trats and atsh, always

retain y. The others retain it optionally except ais and

gat*h, which are irregular. See below, pp. 64 and 65.

Moreover, ** y is not, in this conjugation, elided after ^ s. All the

examples of the elision of H y after this letter given by Ifvara-kaula

belong either to the first or second conjugation.

1899.]

Aorist Participle Generally.

61

The Aorist Participle Generally.

The masculine plural of this participle is formed by changing the

final yov to ffa yey. Thus, s&irfa karyovt plural karyey (viii.

iii. 15). The feminine is formed yov to yeya (viii. iii. 16).

The feminine singular and the feminine plural are the same. Thus,

karyov , fern. sing, and plur. karyeya.

When the base of the participle ends ^ W tsh, 3? 2, or n, and

elides the ^ y in sftw- yov of the masculine singular, the yey of the

masculine plural becomes ay (viii. iii. 18), and the yeya of the

feminine, becomes ay° (viii. iii. 17). Thus, —

Root kat , spin. Aorist part. kaisov ; masc. pi.

katsdy ; fern. katsdya.

„ wothy arise. Aorist part. wobshov ; masc. pi.

watsay ; fern. wotshdya.

,, ^st daz, burn. Aorist part. dazov ; masc. pi.

dazdy ; fern. dazdya,

,, ran , cook. Aorist part. ranov ; masc. pi.

randy ; fern. rahdya.

Most verbs of the third conjugation ending in these letters do not

elide the ^ y (see page 60). So also the verbs kats, vyafe,

rote, wuts (see p. 58) of the second conjugation. These verbs form

the masculine plural either in 'Sftsj ay or ey, and the feminine in

aya or ^ eya (viii. iii. 17). Thus, —

Root kaLs, (second conjugation) be wet; masc. sing,

kabsydv ; masc. plural kabsdy or katsey.

,, prots, (third conjugation) be pure ; masc. sing,

protsyov ; masc. plural grdtsdy or J5ft%*T protsey.

,, tez, (third conjugation), be sharp ; masc. sing.

tezyov ; masc. plural tezdy or tezey. >

The following verbs have irregular aorist participles. The irregu¬

larities are, of course, carried through the aorist tense.

AORIST PARTICIPLE.

G. A. Grierson — On the Kdpnirt Verb.

[No. 1,

03

P

i-9

P-

©

fl

• rH

a

• rfl

B

©

03

H

i-9

to

O

Ss

HH

02

©

P

• rH

o

00

eS

©

fl

• — H

fl

• rr

a

©

a

» rH

"fl

o

K

w

tf

>

-SI

r-<

fir'

6?

s>

•o

-Si

pS3

©

M

o3

-to

pS3

a

"fl

bo

fl

•p<

QQ

»

c8

©

s

c3

CO

A

o

>•>

r<

o'

fir

let

£

A

O

o

"3- ®

Co

io i©.

Jr SI

t -lo

9

o

>4

I©

5J»

>©

fir

%

A

IO

SI

•c*

ns

fir'

■r-

<£

a

o

55

IO

cot

-tol

p<

^3

fir'

'if

S’

55

IO

«=:

Si

«

rC

P*3

?0

•O

pS3

'ter-

n

9

%

Oh

hr'

9

W

IO

pSJ

SI

IO

CO

Cl

CO

Cl

*p

05

Cl

e\

00

1899.]

Irregular Aorist Participles

63

las is sometimes used regularly by the vulgar. Thus, lasov (viii. iii. 96).

AORIST PARTICIPLE.

G. A. Grierson — On the Kflpmiri Verb.

[No.

kj

■<

02

◄

wj

to

&

53

i— ■

m

0)

a

• r-*

a

• »H

£

a>

<D

d

03

ao

cS

O

8

a>

03

d

•1-4

d

03

oc

08

cq

AS

cq

hi

a

"d

tuO

d

• rH

00

w

ci

03

a

ed

so

id,

-21

hr'

¥

103

>3

id

41

,r

> — -

IT-

>3

•d

id

41

hr'

i

r-

<»

'd

41

£

r

SO

id

41

rcr '

r-

s>

»o

d3

id

41

hr'

t

r

id

>3

£'

5

fs

»d

d*

£'

ST

40

<sf

40

Sh

a>

-M

d

03

41

d

r

-+3

00

d

Ph

-+s

os

• pH

u

o

<1

p4

d

40

rv

oa

40

X

d

Ah

03

Xi

«<>

d

<*.

A,

•o

S>3

>d

?»3

&«

hr'

fa

w—

BT

CO

• pH

04

o

<d

5»

»d

>»

A,

hr'

hr

B

a.

a

s

1899.]

Irregular Aorist Participles ,

65

a

C 3

r— H

a

• iH

02

t»

c 3

o

a

GQ

,03

>23

a

P'

&

n3

03

a

~p

o

£

©

,03

>23

a

Sr

&

©

03

123

Sr

ftr

©

,03

IV)

-

r

&

©

553

C

03

Sr

F*

©

03

123

>3

a

Sr

rtcr

©

,03

123

a

03

Sr

©

o*

123

03

PS

V3

123

"a-

Sr

(to

©

03

123

03

PS

o

a

Sr

m

©

03

123

03

PS

123

rO

u

o

©

03

123

03

*■»

a

P$

a

p£

cr

?&

to

©

03

123

03

CP

a

*»

a

P5

Sr

to

ft9'

o

S

a

03

o

I©

03

P5

rO

(to

•

TS

03

02

a

•4-3

O

o

CO

J. i. 9

66

G. A. Grierson — On the Kagmiri Verb.

[No. 1,

5 (6). The Aorist Tense.

This is formed from the Aorist Participle, exactly as the Past tense

is formed from the Past Participle, except that a final ^ v or a final

*7 y is always elided before a pronominal suffix (viii. iii. 41). Thus,

karyov + am, made-by-me = ^%f^ karyo-m, not karyov-am.

The ^ a of am, is elided under the general rules for pronominal

suffixes (vide p. 15). So also forms like kumalyo-s (kumalydv +

as), he was tender for him ; kumalye-s , they were tender for

him.

We thus get the following forms, —

A. First Conjugation.

(a). Transitive verb (viii. iii. 14). (c. forms only).

‘I made,’ lit. ‘he, she, it, &c., was (were) made by me, you, him,

us, &c.’

67

1899.]

Aorist. First Conjugation.

As regards verbs ending in vowels, the usual rule is followed

Thus, from khi , eat, the aorist participle is kheyov ; and from

ci, drink, ceyov.

From f*r ni, take, we get, as usual niyov, not neyov.

di, give and hi, take, are, as in the Past, irregular. Of

di , the Aorist Participle is ditsov (pi. ditsgy ; fern, txnfsj

ditsgya), and of f? hi, hetsov (pi. hetsay ; fern.

hetagya ) (viii. iii. 32). See p. 61.

Double pronominal suffixes can be added, as in the case of the

Past. Thus, karyo-th-as, thou madest me ; karyd-m-ay ,

I made for thee; karyo-m-as , I made for him. Other examples

are unnecessary. The rules are exactly the same as in the case of the

past tense. Vide, however, special cases mentioned after the Paradigm

of the third conjugation (p. 70).

(6). Impersonal verb (viii. iii. 14). (c. forms only).

‘ I laughed,’ lit. ‘ it was laughed by me,’ &c.

Singular 1 aso-m , I laughed.

2 asd-th , thou laughedst.

3 aso-n, he laughed.

Plural 1 asov, we laughed.

2 aso-wa, you laughed.

3 asd-kh, they laughed.

When an impersonal verb is conjugated in the femiuine (see pp. 22

and 48) we get

Singular and Plural.

Singular 1

2

3

Plural 1

2

3

tsuvyeya-m, I quarrelled.

sgsj’q'ST tsuvyeya-th, thou didst quarrel.

tsuvyeya-n , he quarrelled.

tsuvyeya, we quarrelled.

tsuvyey<l-wa, you quarrelled.

tsuvyeya-kh, they quarrelled.

68 G. A. Grierson — On the Kagmiri Verb. [No. 1,

The feminine impersonal verbs tsar, &c., (see pp. 16, 22 and

49) have the following forms, —

tsaryeya-m , I was inwardly wrathful.

tsaryeyi-y (with inserted i; see p. 70 post), thou wast

inwardly angry.

tsaryeya-s , he was inwardly angry.

And so on.

So also (viii. iii. 45) tyamby8yci-s (3rd conj.), he glanced

eagerly.

B. Second Conjugation.

Neuter verb (viii. iii. 39). (c. forms Only).

‘ 1 became, &c.’

C. Third Conjugation.

Neuter verb wuph , fly (viii. iii. 39). (c. forms only).

‘ I flew (just now),’ used in the sense of the Past, and not as the

tense^of the Aorist.

1st Sing. Masc. wuphyd-s, &c., exactly as in the second

conjugation.

In this tense, the difference between the second and third conjuga¬

tions consists in the formation of the Aorist Participle, as already

explained, and not in the conjugation.

1 X^vara-kaula (viii. iii. 44) gives bovyeiv a, but my Pandit says this is a

mistake. The form given above is the correct one.

1899.]

Aorist. Third Conjugation.

69

The Aorist of fa piy fall, used in the sense of the Past, is (viii. iii,

48, 60, 61),—

‘ I fell, &c.’

Note the specially irregular 2nd person plur. masc.

The Aorist of verb 31^ gatsh , go, used in the sense of the past, is

(viii. iii. 48, 68, 60, 61).

‘ I went, <fec.’

When this verb is used in the meaning of £ be proper,’ it is regular,

and belongs to the 2nd conjugation. Thus, gytsh1*, it was proper

(viii. iii. 68). In this sense it is only used in the Past. The Aorist and

Pluperfect do not occur (82).

G. A. Grierson — On the Kagmiri Verb.

70

[No. 1,

There are also other irregular Aorists of this conjugation. There

are those of the roots fa yi , come ; far zi , he born ; ner, go forth ;

ats , enter ; sre pras, be born. The Aorist Participles, in the sense

of the Past will be found in the list of irregular Aorist Participles, and

no difficulty will be found in conjugating them. They are here given

for ready reference.

fa yi, come,

far zi, be born.

ner, issue.

cits, enter.

pras, be born.

Aorist Participles.

dv.

X

zdv.

x

5TcT drav.

x

isav.

pydv.

The feminine and masculine plurals are given in the list of

Aorist Participles. It will be seen that they must be conjugated as if

they were Pluperfects. Vide p. 74.

When the suffix of the dative of the second person singular is

added to a form ending in dv, d becomes 4 wa (o) (viii. iii. 43).

Thus, Tcaryov + ay, karyoy,1 we made for thee.

wuphyoy ,l he flew for thee. 3ffa gauv (root gatsh ), he went;

goy, he went for thee, taftq pyauv , he fell ; pi/oy, l1® fell for thee.

The plural of gauv, is JP7 gay, and ‘ they went for thee ’ is Ji*J gay.

So, ‘ they fell for thee ’ is xzW vyay-

[When the same form ay is added to a form in ya, x i is

inserted. Thus, karyeya, she was made by us ; karyeyi-y,

she was made by us for thee. Compare ^ffa*{ tsaryeyi-y on p. 68 ante ].

1 These are the forms according to the Sfitra. Bat my Pandit maintains that

the true forms are karyo-y, and ivwphyo-y . The exceptional forms, he

says, only occur in the case of the verbs JTS? gat&h, go, and fa pi, fall. The Sutra

while making the rule absolutely general, only gives the two last-named verbs as

examples.

1899.]

71

Pluperfect, Participle and Tense.

6 (a). The Pluperfect Participle.

This participle occurs in all three conjugations. In the first two

conjugations (viii. iii. 78), it expresses remote time. Thus,

karya-n , he made ( a long time ago) ; bovyav , he became (a long

time ago).

In the third conjugation it is used in the sense of an Aorist Parti¬

ciple, and expresses past time indefinitely, with no reference to proxi¬

mity or distance. It is hence the participle, and the Pluperfect is the

tense, of narration in this conjugation. Thus, wuphyav , he flew.

In order to supply the place of Pluperfect, a new tense is formed in

this conjugation, which may be called the True Pluperfect. Thus,

wuphiyav , he flew a long time ago.

The Pluperfect Participle is formed by changing the termination

ov of the aorist participle to WR av (viii. iii. 35). Thus Aorist

Participle, karyov ; Pluperfect Participle, karyav. In the

first conjugation, the masculine plural is formed by inserting ^ e,

before the termination of the masculine plural of the Aorist Parti¬

ciple. Thus, karyov ; masc. plur. karyey ; Pluperfect part,

masc. plur. kareyey (viii. iii. 36). When the Aorist masc.

plural ends in ay {vide, p. 61) this is changed to ayey.

Thus ; Aorist Participle hetsov, taken ; masc. plur. 'fJUT*? hetsay ;

Pluperfect Part. masc. plur. hetsayey (viii. iii. 37). The femi¬

nine singular and plural are the same as those of the Aorist Participle

(viii. iii. 38).

In the second and third conjugations, the masculine plural is

formed by changing av of the singular to ay (viii, iii, 40).

See, however, p. 74. Thus wuphyav, flown; masc. pi.,

wuphyay. The feminine (singular and plural) is formed by changing

av to eya. Thus wuphyeya (viii, iii. 44).

6 (6). The Pluperfect Tense.

This is formed from the Pluperfect Participle, exactly as the Aorist

Tense is formed from the Aorist Participle.

We thus get the following forms.

72

[No. 1,

G. A. Grierson — On the Kdgmiri Verb.

A. First Conjugation.

Transitive verb (viii. iii. 35). (c. forms only).

‘ I made,’ lit., ‘ he, she, it, etc., was (were) made by me, you, him,

us, &c.’

From khi, eat, we have kheyd-m ; plur. kheyeye-m.

So from ci, drink.

From fsr ni, take, fiRTT^ niya-m and fif%^ niyeye-m.

Similarly, from, hi, take, hetsd-m ; plur. hetsaye-m ;

and from di , give, ditsam, and ditsaye-m (see p. 71).

Impersonal verbs are similarly conjugated. Thus, ^f^TTTT asa-m ,

I laughed. Those that are conjugated in the feminine (see pp. 16, 22,

49 and 68), are, of course, the same as the Aorist.

3899.]

Pluperfect Tense.

73

.Double pronominal suffixes may be added, as in tbe case of the

Past and the Aorist. Thus, karyd-th-as , thou madest me ;

karyd-m-ay , I made for thee ; karyd-m-cts , I made for him.

Further examples are unnecessary. See, however, the special cases

mentioned after the paradigm of the third conjugation (p. 75).

B. Second Conjugation.

Neuter verb (viii. iii. 40). (c. forms only).

‘ I became a long time ago.’

C. Third Conjugation.

In this conjugation, the tense has merely the meaning of an

Aorisfc, not of a Pluperfect and is the tense used in narration. It is

conjugated as follows.

Neuter verb (viii. iii. 40, 44). (c. forms only).

‘ I flew, &c.’

1 l9vara-kaula gives ivwphye-wa, which my Pandit says is wrong.

J. i. JO

74

G. A. Grierson — On the Ka^mlri Verb.

[No. 1,

In the plural masculine I^vara-kaula gives wuphyey , but this

is directly contrary to the rule (viii. iii. 40) of which the word is given

as an example.

If it is desired to give the force of the pluperfect to a verb of the

3rd conjugation, we must insert an x i before the *TT3[ yav of the

participle. We thus get what I call the True-Pluperfect tense, which is

as follows (viii. iii. 50).

True Pluperfect (3rd conjugation only) (c. forms only).

‘ I flew (a long time ago), &c.’

For the plural masculine l9vara-kaula gives (viii. iii. 40) as ex¬

amples both mokaliyey and mokaliyay , we or they

were released. According to his own rule, of which these are examples,

the latter is the correct form. For the second person plural feminine

he gives (44) naviyew* , you became new, which, according to

my Pandit is incorrect for naviyeya-wa.

This form cannot be used after cases ending in is, w tsh , sj z, or

sr n. Thus, plup. tezydv , not teziyav (viii. iii. 51).

As an example of the True Pluperfect of the feminine impersonal

verbs (see pp. 16, 22, 49, 54, 67, and 68), we may give (viii. iii. 45)

tyambiyeya-m , I glanced eagerly.

tyambiyeyi-y (see p. 70), thou didst glance eagerly.

tyambiyeya-s, he glanced eagerly.

1899.]

75

Pluperfect Tense.

The formation of the Pluperfect and True Pluperfect participles of

the following verbs is irregular : —

It will be seen that these Pluperfect Participles (in the sense of the

Aorist) are really Aorist Participles of the second conjugation. The

feminine, and masculine plural forms, will be found in the list of

irregular Aorist Participles (pp. 62 and ff:.).

When pronominal suffixes are added to this tense, a final «r v or *7 y

is elided, as in the case of the aorist (p. 66). Thus, aydr + ^r as ,

^psTTW dyd-s, he came to him. aydy + 'QR r as, dyd-s, they

came to him. Other examples of these suffixes are dyd-s +

as, dyd-s-as, I came to him; aya-Jch + m^ as,

dyd-h-as , thou earnest to him; dyd-w® + as, ^RTTO dya-wa-s,

you came to him (viii. iii. 41).

When ay, the suffix of the dative of the second person singular

is added; ^fT«r dv becomes o, and ^TR ay, m\ a (42). Thus, ^*R oy

(^fR dr + ^R ay), he came for thee ; ^{*7 ay, they came for thee ; ^R

droy, he came out for thee ; Sfbr dray, they came out for thee ; RR tsoy, he

entered for thee ; tsay, they entered for thee. So j|fqisfR wuphiyoy,

he flew for thee (^fqRR wuphiyav + ^T ay).

76

G. A. Grierson — On the Kftgmiri Verb.

7. Perfect Tense.

[No. 1,

This tense is not described by I^vara-kaula. It is formed by

conjugating the Adjectival Past Participle with the Present tense of the

Auxiliary Verb. As in the case of the other past tenses, the construction

is passive in verbs of the first conjugation, and active in verbs of the

second and third. The pronominal suffixes are added as in the past

tenses. They are added to the Auxiliary Verb, and not to the Participle.

In the first conjugation they are suffixes of the agent case, and in the

other two of the nominative case. As in the case of the Past tense,

there are a , b, and c forms, — i.e., we may omit the suffixes at pleasure,

except in the second person. Thus, —

(a) . ?}pT ^ tarn 1 kgrumatn chub , by him has been made, —

(6). tarn * kar'umgtu cliu-n , by him has been made-

by-him, —

(c). W*r kar^mat11 chu-n , has been made-by-him, —

all meaning ‘ he has made.’

In the second person, only the b and c forms are used. Thus, —

(b) , tohe kar^mat11 chu-way by you has been made-

by-you.

(c) . ^ karnmg,t'a c7iu-wa, has been made by you, both

meaning ‘ you made.’ We cannot say w'? tohe

kg,rumgtu chuh. In the paradigms, I shall only give the

c. forms.

The Auxiliary Verb may either precede or follow the Participle.

i i

Thus, Wf{ karumgtu chu-m or WT{ chu-m kgrnmgtu , but it is

considered more elegant for it to precede, when in a sentence, and not

standing by itself. Thus, jjj JIT me chu-m gara kgrumgtui I

have built a house. When standing by itself, the Auxiliary Verb usually

follows.

The following is the conjugation of this tense. Regarding the

formation of the Adjective Past Participle, and its declension, see p. 29.

1899.]

Perfect Tense.

77

©

a

-Q

©

TJ

c3

a

d

©

r— * CD

§ ■“

&3

O

«— i

6-

•*4

O

P

•-s

S5

o

O

CH

Ui

p

00 <D

a §

d s

o *d

Pd w

o c§

' rP

rO

u

©

>

Pd .13

00

d

g

CH

©

-M

©

-4-=>

•rH

©

rd

03

©

' -<s>

53 ^

W'

©

a

©

>

ci

fti

►J

H

«

P

PH

M

◄

P

O

JZ5

rH

CQ

©

d

• rH

.a

a

©

Pd

<s

-SI

3©

d

53

Hi

IF-

ft

d

■-a «•

d Hi

O

00 frr

d <JE/

<K,

©

d

• rH

.2

a

©

is

-SI

53.

d

<3.

Hi

Pd

~ft

<V

-K>

<3.

©

d

3 53.

O He

3-r

©

=01

"K»l

53

4

d

53

Hi

©

=o|

■+■=1

e

|

d

e

n»

S

5=0

Hi

O

©

-SI

13

d

53

na

Hi

3^

Hi

©

=01

■+=>1

53

s

d

e

Hi

o

-SI

5©

d

53

Hi

t

ft

rH

O

•«»

<3.

d

53-

Hi

ft'

&

-ft

h:

-to

I

' c~>

Hi

o

•#

-to

53.

d

<3.

na

ft'

&

<&

1

-»

*<s>

Hi

©

•r.

-Kl

e.

d

J-

ft'

<£

-f

<+-/

-ft

Hi

=o

...

-+^>

i3.

d

53.

Hi

e

■

• C-i

Hi

©

e.

d

53.

Hi

V53

Hi

o

IS

=oi

-+-=>1

53.

■a

d

53.

Hi

w

5Qi

Hi

o

IS

=01

■+ot

53.

s

•s

d

53.

Hi

ft'

W

Hi

©

=01

-*ol

53.

d

53.

Hi

3

4SI

53.

•3

d

53.

HS

W‘

-ft

Vi

p

ft

4,

-F

w

-ft

4,

r//j^

-np r<

%' «

'ft

55

Hi

o

d

53.

na

g;

tc>

:i'

50

d

• rH

m

CO

d

S

CM

hs

HS

5=0

Hi

VJ

o

=01

-»l

C

S

3<U

d

53

Hi

r

•'tv

ft

Hi

Ha

i

*<S»

<o

HO

s

d

53.

na

co

78

[No. 1,

G. A. Grierson — On the Kagmiri Verb.

(b). Impersonal verb, (c. Forms only).

‘I have laughed,’ lit., 1 it has been laughed by me.’

i l

Sing. 1. asnmatu chu-m, I have laughed.

2.

3.

Plur. J.

2.

3.

i " i

to aswmatu chu-th, thou hast laughed.

VI « ’

I ' f x

asumatw chu-n, he has laughed.

i i

as1lmatu chnh, we have laughed.

I * I

’Sillier ^ Q'sUmQ't'u' chu-wa, you have laughed.

I ^ I

^5Ft«T asumatn cliu-kh, they have laughed.

Pronominal Suffixes.

When the grammatical subject (i.e., the logical object) is a pronoun,

it is frequently added in the shape of another pronominal suffix

(nominative form). The following are the masculine forms used.

Feminine forms can be easily made on the same principle : —

i i

karumatu chu-th-as, have been made-by thee-I,

thou hast made me.

\ i

So kar^maW' che-th-as, thou hast made me (fern.).

\ #Cs, n • ' ' '

I fN

fTO kar'mat * chi-th , thou hast made us (masc.).

There is no suffix for the first person plural.

kg,rumatu chu-n-as, he has made me.

i ^

kar^mat™ chu-w-as, you have made me.

I

m*5PFT«T karumat'a chu-h-as, they have made me.

i I

' M

^ chu-m-akh , I have made thee.

I ] chu-n-akh , he has made thee.

Jcarumatu d * '

i chu-kh , we have made thee.

chu-h-akh , they have made thee.

When the logical object is the third person, the forms of the first

and third persons are not used in this way. This applies also to the

plural. We thus have only the two following forms: —

kar^maV,

• • /

( W'SIrr chu-th-an, thou hast made him.

I i

( TOT chu-w-an , you have made him.

For ‘ I have made him,’ we must use the full pronoun ; thus, ^

suh chu-m karumatu , and so for the others.

> . * • /

1899.]

Perfect Tense.

79

chi-m-awa , I have made you.

Jcar'mat *, ^ chi-n-awa, he has made you.

chi-h-aioa, they have made you.

kar'mat1.

• • /

( chi-th-akh , thou hast made them,

(f*** chi-w-akh , you have made them.

B. Second Conjugation.

Neuter verb. (c. forms only).

I have become, &c.

C. Third Conjugation.

Neuter verb. ( c . forms only).

I have flown, &c.

so

[No. 1,

G. A. Grierson — On the Kagmirl Verb.

8. The Periphrastic Pluperfect Tense.

This tense, also, is not mentioned by I^ara-kaula. It is formed

exactly like the perfect, except that the past tense of the auxiliary verb

is used instead of the present. It is unnecessary to give full paradigms,

the following examples will suffice.

i i

karumatu osu-m , I had made (him).

i i

kar^mais® dsu-m. I had made (her).

' v«\s ‘ v 7

karhnat 1 dsi-m, I had made (them, masc.).

kar email a asa-m , I had made (them, fern.).

b6vumatn osu-s , I had become.

'Cfvj S» '

wnpliybmatu osu-s, I had flown.

Or, with double pronominal suffix.

^eTth^ kar^mat11 bsu-th-as , thou hadst made me.

s O s * * '

B. Imperative Mood.

1 (a). The Present Tense.

The terminations are (viii. ii. 5).

Singular. Plural.

2 is h. 2 ^ iv.

3 in. 3 ^«T in.

If the root ends in a consonant, the ^ h of the 2nd singular is

elided (6). Thus, —

^ Tear, make thou. kariv , make ye.

karin , let him make. karin , let them make.

From di, give, which does not end in a consonant, we get for

2nd singular dih, not di.

If the root of the verb contains the letter o, that ^it d becomes

^ w, in the Imperative (7). Thus, —

roz , stand, 3rd Sing. Imperat. ^f5T«T ruzin.

ffr^r tol, weigh, „ „ tulin.

xfts- poth , be fat, „ „ puthin.

So also, if the root contains ^ e, it becomes t i (7). Thus, —

jfa; ner, go forth, nirin.

ST3T plfc, fear, suffer ftk in.

ped, exude, pidin.

3899.] Present Imperative. 81

These changes, however, do not occur in the second person sin¬

gular (8). Thus, we have, roz , ?ffa tol, pbth, ner,

fffc, tfa ped.

x

We thus find the Present Imperative of Tfa roz , remain, to be con¬

jugated as follows.

Singular. Plural.

1 ^CTST roz. ruziv.

> '

2 ^fa*r ruzin. ruzin.

Every root ending in a vowel, takes the letter *r y before all

terminations, except that of the second person singular (10). Moreover

a final X i of the root is changed to ya(e), except in the case of the

verbs fa m, take, fa di , give, and fa yi , come (11). We thus get the

following conjugation of a verb whose root ends in a vowel.

(a) fa khi, eat.

Singular.

2 TcJieh.

3 kheyin .

(b) fa di , give.

Singular.

2 fa^ dih.

3 fafa«T diyin.

Plural.

^Hfa^ kheyiv.

<5ffa«T kheyin.

Plural.

fafa^ diyiv .

fafaif diy in.

The root fa yi, come, is further irregular, in that, besides being

conjugated like fa di, it also optionally takes the following form

(viii. ii. 12).

Singular. Plural.

2 wola (not wol). woliv.

3 ^fa*T wolin. ^fa«T tvolin.

The root bov , become, has the following forms (14).

Singular. Plural.

2 bov. boviv.

x x

3 bovin, «^fa^ boy in. bovin or ^fasr boy in.

J. i. 11

82

G. A. Grierson — On the Kagmiri Verb.

[No. 1,

The verbs (see pp. 16, 22, 37 and 49).

^ tsar , be inwardly wrathful.

pkog, be inwardly wrathful.

phuh, be inwardly wrathful.

marts, be inwardly wrathful.

wuts, be burnt.

yhits, forget.

tyamb, look eagerly.

All of which are impersonal, and are only used in the third person

singular, to which the appropriate pronominal suffixes of the dative are

added, (viii. ii. 9).

Thus, —

tsgir'n-ay, let there be inward anger to thee, i.e., be

thou angry.

isar'n-aw0, be ye angry.

tsaVn-as, let him be angry.

tsaVn-akh, let them be angry.

Regarding the vowel changes, see the following rule.

When a pronominal suffix is added to the Imperative third person

singular or plural, the ^ i of the imperative becomes i-matra, and the

preceding vowel is modified. Thus,

karin, let him make.

kar'n-am, let him make for me.

The second person is, —

Sing. karu-m, make thou for me.

Plur. karyu-m , make ye for me.

When a pronominal suffix is added to the second person singular

of the imperative of a root ending in a consonant, the letter ^ u is

inserted (16). Thus, karu-n , make him or it (a very common

form) ; karu-m, make for me; karu-s, make for him;

karu-kh, make for them.

As regards roots ending in a vowel, from khi, eat, we have

TjjpT khe-m , &c. From ni, take, di, give, and fcr yi , come, we

have di-m , give thou to me, &c.

1899.]

Modified Present Imperative.

83

When a pronominal suffix is added to the second plural imperative

of any verb, iv, becomes *^yu (17). Thus, haryu-m , make ye

for me ; karyu-s , make ye for him ; karyu-kh , make ye for

them. So from f^r khi, eat, j ^ khyayu-m , &c ., and from f«r ni} di,

and f?f yi , diyu-m , &c.

1 (6). The Modified Present Imperative.

This, though not a respectful imperative, is more polite than the

simple tense. It is formed by inserting the particle <T ta. It expresses

encouragement, like the Hindi rff karo to ! It also expresses permis¬

sion ; thus, 1 very well, if you wish to do it, do it.’ The terminations

are as follows (viii. ii. 14).

Singular. Plural.

2 rT t*. Hav .

x X

3 'tan. Han.

x \ ^

The terminations are all added to the root direct (15). The * being

i-matra , a preceding vowel is modified in the 2nd plural, and 3rd sing,

and plur. Thus,

Singular. Plural.

2 kart*. karHav.

3 JearHan . kar'Han.

So also from kkar , mount, *arrf^?f«T kharHan ; from wdl ,

bring down, ^rf%ri3T ivalHan ; from roz , remain, 2nd sing. rdzta ,

3rd sing. ^f5j<T«T ruzHan ; from ner , go forth, nerta ,

nirHan ; and from tsar , be inwardly wrathful, &c., tsqrHanay ,

&c., (see p. 82).

Regarding roots ending in vowels we have from f*sf khi, eat.

Singular. Plural.

2 ^jrr khet «. kheyHav.

i i

3 kheyHan. kheyHan.

84 G. A. Grierson — On the Kfigmiri Verb. [No. 1,

For ni, take, di, give, and fa yi, come, we have, however,

the following forms.

Singular. Plural.

2 fa?r dit diy'tav.

3 fafafl*! diy'tan. fafaiT^ diy'%tan.

The pronominal suffixes are added regularly, except that in the

second plural, ^ av becomes d (18). Thus, karHo-m, make

ye for me ; 'fifaifl-T karHo-n , make ye him ; karHo-s , make ye for

him ; karHo-kh , make ye for them.

2. The Future Imperative.

This is formed by adding ^far *zi if the root ends in a consonant,

and % zi if it ends in a vowel. Before *;far *zi, a preceding vowel is

modified (viii. ii. 22, 24). This tense does not change for number or

person. It means ‘you, or he, should do a thing at some future time,’

or ‘ make a practice of doing it.’ Thus

etffafar tsah kar'zi, thou shouldest do.

toh * kar'zi, you should do.

suh kar'zi , he should do.

fa^ tim kar'zi , they should do.

So also from *3TT khar , mount, kharlzi ; from wdl,

bring down, wdl'zi ; from roz , stand, ruzlzi ; and

from Sfa net, go forth, ftfrfal nir'zi. I cannot find that this form

is used with impersonal verbs like £sar etc., mentioned when dealing

with the Simple Imperative.

As regards verbs ending in a vowel, we have from far khi, eat,

khezi; so also in other cases, but from fa ni, take? fa di, give,

and fa yi9 come, we have fafar dizi, etc.

When the pronominal suffixes am and as are used with this

form, far zi becomes sq zya. (viii. ii. 25).

Thus kar'zy-am, you should make me, or for me.

So kar'zy-as , you should make for him.

1899.]

Benedictive Mood.

85

In other cases, the zi , is unchanged.

Thus j karlzi-y, he should make for thee.

[Note. My Pandit also says kar'zy-an, not karlzin ;

so also he says kar'zy-aw a, and karlzy-akJi\ .

3. The Past Imperative.

This is formed by adding % he for all persons and numbers to the

Future Imperative (viii. ii. 23). It means 4 you should have made so

and so,’ implying that he had not done it.

Thus 4f^f5T% kar'zihe , thou shouldst, you, he, or they, should have

made.

Pronominal suffixes are added regularly (25). Thus

kar'zihe-m , you should have made for me.

C. Benedictive Mood.

1. Future Tense.

This tense expresses a wish. It is formed from the Pluperfect

Indicative, by substituting the following terminations (viii. ii. 26).

%

Singular. Plural.

2 sj^sf yakh. x*

3 yan. yan.

The tense expresses a wish. Thus, lagyan , may he live long.

The following is a specimen of the conjugation of the tense of the verb

qrc kar , make ; Pluperfect karya-n he made.

1 Mayst thou make, &c.’

Singular. Plural,

2 karyakh. kariv.

3 karyan. kar yan.

So from ran, cook ; 3rd sing. Plup. ^r*j raha-n ; 3rd sing.

Bened. *csq*T rahyan. Similarly zehyan, may he conquer.

Boots ending in ^ s change the final ^ s to p. Thus, from

^ las, live long; 3rd sing. plup. ^TT^r«[ latshav ; but 3rd sing. Bened.

la?yan. So also, from as, be, 5fT*?fiT apyan.

G. A. Grierson — On the Kagmiri Verb.

86

[No. 1,

The verb chav, use, has for its second singular Bened. either

chdvyakh or WT**J*3 chdyyakh. The latter form is peculiar to the

second person singular (27).

The verb ®nr bov, be, become, changes its final ^ v to y through¬

out. Thus, boyyan (not «n?jiT bovyan), may it be; *7 m°

boyyan , may it not be, God forbid ! (26).

Pronominal suffixes are added in the usual way. Thus,

lagyan-ay, may he live for thee ! posyan-ay , may he be victorious

for thee !

This tense only occurs in the above verbs (26).

D. Conditional Mood.

1. Present Future Tense.

This is the same as Future Indicative. An example of its use is

^ JTW <|W«T bo-y gatsha , suh wucha-n. If I go, I shall see him.

^ n

«7*7 boy is contracted from boh, I, and ^5? ay, if. The object is

mentioned twice. First fully in ^ suh, and again as a pronominal

suffix («T n).

2. The Past Conditional Tense.

This tense is used if things are spoken of that might have, but

have not, happened. Thus, ^ ^*7 i27f*7% ^rf5T% rud ay peyihe,

soch sapazilie, if there had been rain, there would have been plenty.

It is conjugated as follows ( viii. ii. 32).

Singular.

1 karahd, (if) I had made.

2 karahakh,

3 karihe ,

Plural.

karahdv.

x •

kar^hlv.

s x •

karahdn.

X •

Notes. a) When the last syllable contains the vowel a, that

vowel is always modified. This is not mentioned by ^vara-kaula, but

is a fact.

(2) The short i in the second person plural is i-mdtrd , and modifies

the preceding root vowel when possible.

Verbs ending in vowels are declined as follows, inserting ^ m in

the first person, as in the Future Indicative.

1899.]

Past Conditional.

87

Singular. Plural.

1 TJSPTST khyamaha, (if) I had eaten. khyamahav.

2 khyahakh. khey'hiv.

3 kheyihe. ^?T«r khyahan.

From f*r ni, take, di , give, and fa yi, come, we have as follows.

Singular.

1 dimaha.

dihakh.

3 fafa% diyihe .

Plural.

dimahav.

diylhiv.

dihan.

Pronominal suffixes are added as follows : —

karaha-m , (if) I or we had made myself or for

myself.

Added to

1st person.

karahd-y,

<j karahd-wa ,

karaha-n ,

karaha-s ,

karaha-kh ,

>>

5)

thee, or for thee,

you, or for you.

him.

for him.

them, or for them.

Added to

2nd person «<

singular.

karaha-m , (if) thou hadst made me, or for me.

karaha-n , „ him.

karahd-s , „ for him.

karaha-kh , „ them, or for them.

f kar^hyu-m, (if) you had made me or for me.

I

Added to j kar'xhyu-n , „ him.

2nd person «{

plural. j kar'hyu-s,

kar'hyu-kh ,

»>

for him.

them, or for them.

r

Added to

3rd person *

singular.

L

karihe-tn , (if) he had made me or for me.

karihl-y , „ thee, or for thee.

karihe-wa, ,, you, or for you.

karihe-s , ,, him, or for him.

5<fa%*sJ karihe-kh ,, them, or for them.

88

[No. 1,

G. A. Grierson — On the Kdgmtri Verb.

Tcarahqn-am , if they had made me or for me.

thee or for thee,

you or for you.

him or for him.

them or for them.

Note. — All the forms added to the second person are irregular ;

and also the suffix of the second person singular, when added to the

verb iu the third person singular.

The feminine impersonal verbs a tsar , he inwardly angry, etc.,

(see pp. 16, 22, 37, 49, 54, 68 and 82) are conjugated as follows : —

tsarihe-m , (if) there had been inward anger to me; (if)

I had been inwardly angry.

*5^1 ctse tsarihe, (if) we had been inwardly angry.

tsarihi-y , if thou hadst been ,, ,,

tsarihe-wa, if you had been ,, ,,

hsarihe-s, if he had been „ ,,

tsarihe-kh , if they had been „ „

This tense may also he used in expressions like the following : —

su-y karihe , even he did it. That is to say, ‘ why did you

do it ? It was his business, and he has done it already.’ (35).

Added to karahan-ay,

3rd person karahan-aw a,

plural. karahqn-as, ,

karahan-akh ,

APPENDIX.

List of Ka^mIrI Verbs arranged according to Conjugations and

Final Letters.

As the forms taken by a Kj^miri Verb depend partly on the

Conjugation to which it belongs, and partly on the final letter of its

root, the following list of Verbal Roots is arranged under Conjugations,

and then alphabetically according to the final letter of each root.

For the meaning of these roots, and the form of the Past and

Aorist Tenses of each, the reader is referred to the List of Kapmiri

Verbs given ante , Vol. LXV, Pt. I, pp. 314 and ff.

First Conjugation.

khi , ci, di, m, hi.

• khisk, chak , chik, tak, tuk , tak , thak, thuk , dak , truk , thyak , tliok,

dr ok , phak , phuk, phuk , phrak, bak , Irak , g§k, grok, suk , hyak.

1899.]

89

Boots according to Final Letters.

jakh , lakh, lilch, lekh.

teyayg, ztig, tag , da#, p7ia#, hug , magg, ragg, lagg, lag, wugg, wolagg ,

hag.

kharac.

khach, tacli , TrncA, much , rach, wuch, Iryachh.

mtij.

fearfe , wajte, phife , marts, ivdfe, wufe.

yifeh, kafeh, pritsh, brifeh.

arz , kanz, khunz, graz, nyawtiz , ptiz , puz, prinz, haz , Zws:, 602:,

iya2, soz.

kapat, ktit. khat , chat, feat, feut, feet, fehat, d$t, pit, prat, rat,

lut , wat, wtit.

wuth.

kad, gand, fehand, fehtid, fetid, tad, mand, mad, lad, sad, hyad.

prin.

hat, khut , nyat , grut, sat.

math.

arad, gind, gund, pad, pond, mand, lad, wad, wand, vyand, sad.

an, afehan , an (=an), kan, khan, chan, fedn, feen, feliun, ztin, zen,

tasan, don, dan, pachtin , parzan, pilan, pufean, prazan, phtin, man, mun ,

men, ran , Ion, Ion, wakhan, wan, vy afehan, won.

krp, khap, gup, chap, fetip, zap, tap, tyap, tap, trap , thip, dap, nap.

chornb, fed mb, tomb, tr'ornb, lab.

kham, cum, dam, lam, warn, hum.

kay, doxy, day, pay, lay, way.

alar, tipar, tiwar, katar, kar, kar, kur, khtir, gar, gtir, gEr, gor , car, cir,

char, cher, clior,jar, jur, fear (be inwardly wrathful), fear , feur, zar,

tar, tovar, thur, darr, dtir, par, par, pyatar, pur, pair,phir, phuktir, phytir,

bagtir, bar, mathar, mar (unite), mahtir, mar, mu fear, uiusar, mutr, mur,

yar, yer, rawar, lattir, Idr, lur, wathar, war, wahtir, war, wifetir, wokhar,

wothar, wophar, wur, wusar, wor, ger, sandar, sandtir, sar, stir, sor, hahar ,

htir, hor.

gnzar, gbar, gzar , Star, ddar, kgnzar, khgndar , ggnzar, grdgar, chgkar,

chindar, chohar, chopar, feinar, feokar, feygtar, felnygtar, fehyavar, feho(ar ,

zithar , zuvar, zovar , zajar, dgkhar, tgfear, tgnar, tizar, tropar, thgzar , ddnar,

dobar , dunfehar, nanar , ndgar, nikar, nomar , nowar, nomar, pgjar , pgzar,

pgthar , pdsar, pithar, pugar, piithar, phgsar, phdsar, phutar. bgdQr, bgnzar,

balar, bdgar , bdjnr , behar , brgxnar , mgxizar , mgsar, motar, mondar, munfehar ,

ygtar, rgfehar, wadar , vezar, venar, vygthar , vygdar , vygpar, vyavar,

tvowar, gubar, sanar, sgnzar, sezar, sombar, hgtar, hokhar, hobar, homar.

ahal, khol, gal, chal, fetil, feyakhal, feyall, felial, ztil , zal, toil, dtil,

tal, tal, tul, tol, that, dal, nyaggal , pal, pal, bol, mal, 161, wal, wtil, wolal ,

sambtil.

J. 1. 12

90

G. A. Grierson — On the Kdgmiri Verb.

[No. 1,

ayav , drav , dlav, kamav, kokav , katsav , gilav, gyav, gulav, chokav,

zarav , tokav, dakliav, dnlav , £ar, trombav , £7iar, dabav , dav , dogav , danav ,

7iechav, pakav, pyav , pinav , porav, phalav, phetsav, phirav, rnanav , milav,

molar , mdrav , ranzav , rinzav, rakav, latav , lalav, lithav , lyav , wazav ,

ivanav , war, ri/ar, wohav, sagav , sokhav , surav , sulav, hasav.

abasav, arpav, alarav , aparav, dwardv , kapatdv , kamandv , ftrar

khokhaldv , guzardv , catdv , car, cukdv, chav, tsdgandv, thaguv, dulandv , £dr

tdvardv , trakardv, tragandv , £rar, £/mr, ddr, dogandv, dodarav , dasrav , nahdv

oidv , w?/ar pachatdv , patav , parkhdv, parzandv, pdv, pilandv,prazandv, prdv

pharkdv , baddv, bgrakdv, bdv, marddv, milanav , mutsarav , musardv , mfttrav

ranzandv , rdwardv, latav, laddv , watharav , wahardv, wokhardv , wuddv

wotharav, wusarav , gggandv, saganav, satdv , sandardv, sav , sokhanctv, hav

hasandv.

gdarav, anzardv , gbardv, adardv, kgnzardv , khgndardv, ggnzardv

grdgardv,chgkardv, ehatsarav, chivardv, chonardv, chgpardv , tsihardv.tsglcarav

%somardv, tshygtardv, tshygvardv, lshgtardv, zlthardv, zuvarav, zovardv

dakhardv, tgtsardv , tahardv, tizardv , trgtsardv, trgsarav, trogardv , thgzardv

dadardv, dghardv, dobardv, duntshardv, drogardv, dasrav, ngnardv, ngtardv

ndgardv nikardv, ngmardv, novardv , nomardv, pajarav, pgzardv, pgthardv

pasHav, pitardv, plthardv, pugardv, puthardv, phdsardv, phutardv, bacardv

bgdardv, bgdarav, bgnzardv, bglardv, bggardv, bajardv, benardv, brgmardv

matsarav , manzardv, masardv , mdnardv, mbtarav , mbndarav , modardv

muntshardv, ygtardv, rgtshardv, lygdardv, lukardv, wgdardv , vezavdv, vehardv

vygthardv, vyadardv, vyapardv, vyavardv, wnkardv, wugarcLv, wowardv

gubardv, sgharav, sezardv, sehardv , sogardv, sotsarav, srggardv, hatardv

handardv , kdtharav, hokhardv hotsardv , hdbardv, homardv.

tswi1, nr, ntr, Zzr, sir, swr.

fcap, kreg, pag, pheg, phug, phog, pheg, brag, gag, sorg.

des, bas, mus , was, gas.

abas, as, kas, kas, khas, tsas, zos, tas, thas, das, mus, ras, sas.

kuh , khah, gah, isah, pih, phuh, muh, sah.

Second Conjugation.

samakh, hokh.

tag, lag, gogg.

kats, khobs, pats, rots, vyats, hots.

gaish, (be proper).

daz, \_palaz], roz, wopaz, sapaz.

[palaf], phaf, phut, rot (also 3rd),

frorf.

1899.]

Boots according to Final Letters .

91

woth.

sapad.

tshyann , \_pun\ , sapan.

wup, grap. (Both these also optionally 3rd Conj. in Plup.).

pray , lay, way.

khar, tar, \_prar~\, p7iar, plier, [phor~\, mar (die), sor.

pa?, isaZ, dal, dol (or do?), plial, pholl, mel.

bov, rav.

[wap] [?os], dos, pds, mas, ros.

as, khas, pirns, has, [ bas~\ , las, Ids, was, [zco&as],

beh.

Third Conjugation.

zi, pi, yi.

grok, camak, tsdk, fink, dok, nik , lok, wok.

thag, tagg, drog , sog , srog.

bach.

mandach.

ats, grots, tsots, tshots, trats, mots, lets, grots,

atsh, gatsh(go).

tez, paz, bdwaz, braz, ranz, laz, wuz.

krt, krat, gyamat , gurat , tshyat, tshbt, tot, nat, pat , pit , prgkhat, mot,

rot (also second), wot, hat.

hath, kuth, kreth, Uamath, zeth, tyath, toth, dreth, n&th, path , poth,

breth, meth, vyath, gith , grith, h&th.

ad, ped, bad, bud, mond, rad, lyad, wud.

turan, prdn, ran, wusn.

chat, zdt, tat, mat, rat , lot, wonnat, sot.

path.

and, tund, thad, pyad, bad, brad, wud, syad, god (this is a better

spelling than goddh ).

aman, kan, kyann, krhan, gan, guman, chan, chonn, tshdtshan, tan,

tllan, nan, pakan, pran, ban, basan, bdkhan, byann, lahan, lahan, wogan,

wun, gigan , san, syann, hdn, hdman, lion, han.

kamp, kup, n<Xp, pap , pasp,ydp, wup (optionally in Pluperfect), vyap,

grap (optionally in Pluperfect), hap.

wuph.

kob, gob, tyamb, dub, tyarnb, phab, rarnb, tub, gub.

garm, tsam , tham, nam , bram, worn, gam, sam, ham.

biy, lay, wuy.

adar, yir, kahar, kdtsar, lcdyar, kaivar, kur, khbkhar, khor, glr,

gbwar, cSkhar, tsar (increase), Uhar, ziggar, zoggar, zozar, thahar, dar,

92

G. A. Grierson — On the Kdgmiri Verb.

[No. 1,

tur , trakar , tror, thar , thdthar, thar , da?*, dodar, dur , do?', ?ier, por,

phahar , baliar , bUbar , bigar , mddar , ??ior, lyadar , lor , vyadar , vyalar , vyasar ,

wukar, tvndar, wobar , sakhar , sagar , syandar, sir , sosar, hakar , hgndar,

liar , W.

aZ, A;aZ, kumal , &oZ, fc/nzZ, gdgal, gel , gragal , choked , zaZ, /aZ, df/aZ,

tambal , ZeZ, rc?Z, jedZ, pigal, pol , prazal , phdphal , 6aZ, mamal, mokal} wigal ,

wozal, wotal, wogal , w/oZ, gahal , 7m Z, hdkal , TuZ, 7<aZ.

chiv , tshyav, zuv , ??ai’, $?m».

trbg.

dwas , dos, Zras, jtds, pras, ras, Zzs, tns, wolas, was.

goh , Z°/i,

1899.] G. A. Grierson — On Indeclinable Particles in Kdgmiri.

93

On Indeclinable Particles in Kagmiri. — By G. A. Grierson,

C.I.E., Ph.D., I.C.S.

[Read January, 1899.]

I^vara-kaula does not formally deal with particles in his grammar,

but here and there he refers to them, and the following is a collection

of his scattered rules. It in no way pretends to be a complete account

of Indeclinables.

Emphatic and indefinite particles have been described by me in

Vol. LX VII, Part I, pp. 88, and following.

The following two conjunctions are also there mentioned. They

are repeated here for the sake of completeness.

?T ta , and (iv. 178). PJ.g., cT ire suh ta tsah, he and thou.

fa ti, also (iv. 179). It is also used instead of rf ta with plurals.

Thus, ^ fa ^ fa suh ti , tsah ti , he also, you also, irefafa fa fa

mahaniv * ti gujpan ti ay , both the men and the cattle came. In the

last sentence we cannot use <T ta.

The negative particle is *T w°, not (viii. ii. 19) ; but ordinarily nega¬

tive forms of the verb are used, as described under the head of adverbial

verbal suffixes. In other words, the «r na is usually compounded with

the verb as a suffix. Thus, chu-s-na , I am not.

The prohibitive particle if ma is only used with the Simple

Imperative (viii. ii. 19).

?Fr m® karin , let him not make.

** m® kar. make thou not.

\ *

m® kariv , make not ye.

With the Modified Imperative mal « is used (20). Thus,

ilcf mat a karHan , let him not make.

94 Gr. A. Grierson — On Indeclinable Particles in Kdgmiri. [No. I,

Instead of ma and *f?T mata , we may use *TT md and HfTT mats

_ ^

respectively (20). Thus, HT ma kar , ^Trfr 3icf mata karta.

Other vocative particles may also be added. Thus, mabd

kar, masa kar , matabd karta, *TfpgT matasa karta ,

and so on (20). See forms of address given in Yol. LXVII, Part I,

pp. 92 and if.

With other tenses of the imperative na is used (19). Thus,

•f kar'zi na , you should not make. «f kar^zihe na , you

should not have made.

ifT ma is used before or after a verb, to indicate a question in hesi¬

tation (viii. i. 29). Thus, —

ut w ^ kardn ma chuli , or iff md clink kardn , or

*• V* > J N

ifT ^iTT«T md kardn clink , is he making ? (J.e., see if he

is not making it, or if he is making it or not, or perhaps

he is not making it ?)

ifT karydn md, did he make ?

itT suk md kari, will he make ?

iff bok md kara, shall I make ?

X 7

The particle rfTof tail, or <TT^Tff tanat , is used in asking a question,

when the speaker is really in doubt as to whether there is anything to

ask (viii. i. 26). Thus, —

?fl«T kydk tan wanun , did he say anything ? Here

the speaker did not notice at the time what the man said,

and afterwards recalls the fact, and, being in doubt, asks

the question P

<TI«T kar tan av, did he come at any time ? If so,

when ?

fPrf cfT^T ’STfpf kutu tan dyutun , did he give anything ? If so,

how much ?

rfHT tarn, or <rm<T tdmat , may be used instead of tan, or

ffT'SjTf tanat. Thus, <TT^ 3«f«f kyak tdm wanun.

'Slcf dyatha. This added to an interrogative word converts it into

an intensive one (viii. i. 27). Thus, —

3ST kar , or ^sf kana , when ?

1899.] G. A. Grierson — On Indeclinable Particles in Kcigmiri.

95

^3: ^ dyatha Tear av, or dyatha kana av, lie

he came a long time ago.

Tcyoh , what ? «Pinr dyatha Jcyah , a great deal.

fif<T hut1, how many ? frfcT dyatlia hut % a great many,

So ^f*T dyatlia han l, for a long time.

The usual word for ‘ if ’ is ay, but, with the Past Conditional

(viii. ii. 33), ^3 hay may be used instead of *(3 ay, after the verb.

Thus, —

^53 harihe hay, if he had made.

3fTT}*T ^3 harahan hay, if they had made ;

harihe- s hay, if he had made it.

With the same tense ‘ if not ’ is represented by 33 nay (34). Thus,

33 harihe nay, if he had made it ; 33 3jf3% rud nay peyihe, if

rain had not fallen.

These particles can also be attached to the subject of the verb (35)-

Thus, ^3 3ffc% sah-ay harihe, if he had made ; f33^3 3iT^T3 tim-hay

harahan, if they had made : «g33 3»fr% su-nay harihe, if he had not made,

33 bo-y hhyamaha, if I had eaten.

96 G. A. Grierson — Grant of Qiva-simhd to Vidyapati-thakkura. [No. 1,

On the genuineness of the Grant of Civa-sintha to Vidyapati-thakkura. —

By G. A. Grierson, C.I.E., Ph.D., I.C.S.

[Read May, 1899.]

Regarding tliis grant, which is dated La-sam 292 (?), Sana 807,

Sumvata 1455, Cake 1321, see the Indian Antiquary , Yol. XI Y, p. 190,

and the Proceedings of this Society for 1895, Plate iii. The genuineness

of this plate has been doubted, but no positive proof for or against the

theory has hitherto been put forward.

Dr. Kielhorn’s Inscriptions of Northern India , No. 578, shows that

the last two dates mentioned in it both correspond to Thursday, 10th

July A.D., 1399 ; but that this day would fall in the Bengali San 806,

and in the Hijra San 801 (not 807). Unless therefore there has been

an error in the calculations of the writer of the deed, neither of these

two latter eras can be meant by the word sana. There is however

another era, also entitled san , and which is moreover the era which

in these modern days , is generally current in the part of Bihar from

which the inscription comes. It is the Fasti San , an era introduced by

the Emperor Akbar. Eor information regarding it, see Prinsep’s Useful

Tables , ed. Thomas, p. 170. The year runs exactly parallel with the

Vikrama Samvat, the only difference being that, to obtain the Fasti year,

we must subtract 648 from toe Samvat date. There are no dark and

light fortnights in the Fasti month, the days running through each

month from 1 to 30, but with this exception the Fasti day of the month

and weekday are always the same as the Samvat ones. It is thus a

very easy calculation to convert a Samvat to a Fasti date, and it will

be seen that Fasti San 807 does as a matter of fact correspond to

Y. S. 1455.

This at once stamps the grant as a very clumsy forgery, for F. S.

S07 never existed. The first year of the era, as founded by Akbar,

was, not 1, but was 963. No date purporting to be earlier than F. S.

963 is possible. It is therefore evident that the dates in this grant

must have been forged by some modern jyautisa, of whom there are

hundreds of half educated ones in Tirhut, who knew the simple equa¬

tion for converting Samvat dates to Fasti ones, but did not know the

history of the Fasti era. In his anxiety to make the grant look as

genuine as possible, he put in all the synchronous dates he knew about,

and exposed his forgery in so doing.

JOURNAL

OF THE

ASIATIC SOCIETY OF BENGAL

Part I -HISTORY, LITERATURE, &e.,

Extra-Number 1.— 1899.

A Collect ion of Antiquities from Central Asia. Part I.1 — By A, R. Rudolf

Hoernle, C.I.E., Ph.D. (Tubingen).

(With 19 Plates and a Map.)

INTRODUCTION-

On two portions of this collection I have already reported in this

Journal , Yol. LXYI, for 1897, Part I, pp. 21311, and in the Proceedings ,

for Apfil 1898, pp. 124 if. In the following pages I propose to give an

account of the entire collection ; and it will, therefore, be necessary to

briefly include the substance of those two previous papers. For the

leisure, without which I should not have been able to write it, I am

under great obligation to the Government of India who, with the con¬

currence of the Government of Bengal, placed me on special duty for

the purpose of examining and reporting on the collection.

i This is my Report to the Government of India, published, by permission of

that Government, as an Extra-Number of the Journal of the Asiatic Society of

Bengal. [Part II. of this Report will be issued subsequently. — T. B.]

J. I. I

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

• •

n

To begin with, I may briefly explain the circumstances which led

to the formation of the collection. It was the

Origin of the discovery of the Bower and Weber Manuscripts

Collection. which first drew my attention to Eastern Turk¬

estan as a promising field for epigraphical ex¬

ploration. My hopes regarding the archaeological possibilities of that

country were confirmed by what I heard about the success of the

Russians, whose Political Agents were said to actively collect manu¬

scripts and other antiquities for St. Petersburg. Accordingly on the

1st June, 1893, I wrote to Mr. (now Sir) Charles J. Lyall, who was then

the Home Secretary of the Government of India, suggesting that the

Government might send instructions to their Political Agents in differ¬

ent parts of Central Asia, to make enquiries and to secure such speci¬

mens as they may be able to obtain. My suggestion was heartily seconded

by Sir Charles Lyall, and at his instance, (in his demi-official letter,

dated the 14th June, 1893), the Foreign Secretary, Sir M. Durand, who

also fully approved of the proposal, caused the necessary instructions to

issue, on the 22nd August, 1893, to Lt.-Colonel D. W. R. Barr, Officiat¬

ing Resident in Kashmir, and through him to the Political Officers in

Gilghit, Chitral, Kashghar and Leli. Similar instructions were issued

to the British Political Agents in Khorassan, and, I believe, in Meshed.

In response to these instructions a large number of Central Asian anti¬

quities has already been secured, forming a very respectable British

Collection, to which additions are still being made. To me personally

it is source of much satisfaction to have thus been the means of initiat¬

ing the movement. It is in acknowledgment of this initiative, that all

acquisitions are transmitted to me, under the orders of the Government

of India, for examination and report, and their ultimate place of deposit,

as recommended by me, is to be the British Museum in London. The

full determination of the antiquities, especially of the manuscript portion

of them, will require more time than I have at present at my disposal.

The present report, therefore, is only o&a pre-

Scope of the liminary character, and must be limited to a

present Report. detailed* classification and description of the

antiquities, illustrated by selected specimens

and accompanied with such observations and conclusions as are obvious¬

ly suggested by them. Much of the epigraphical portion of the collec¬

tion, however, is clearly of the highest interest, whether from the palaeo-

graphic and linguistic points of view, or as bearing upon the history of

culture in Central Asia, and well deserves more elaborate treatment and

more extended publication. For this purpose my approaching retire¬

ment from India, I hope, will afford me the needful leisure, as well as

1899.]

Introduction .

in

the co-operation of fellow-labourers in the field of Central Asian research,

and such assistance as the British Government has always been prepared

to extend to endeavours for the advancement of knowledge.

Among the places above mentioned Kashghar and Leh, especially

the former, are those which are most favour-

Contributors to the ably situated for the object in view, the collec-

collection. tion antiquities from Eastern Turkestan. It

is from the Officers stationed at these two

places that nearly all the objects which at present form the collection

have been obtained. Captain Stuart H. Godfrey, who was the British

Joint-Commissioner in Leh, in 1893, when the instructions of the

Foreign Department were issued, informs me (in a letter, dated the 26th

September, 1897), that he already before that date, on his own initia¬

tive, commenced endeavouring to obtain Central Asian manuscripts.

His endeavours, though then unsuccessful, bore much fruit later on, after

he had left Leh, as may be seen from the following list of consignments.

Mr. George Macartney, who is stationed in Kashghar, as Special Assistant

for Chinese Affairs to the Resident in Kashmir, being in the most

favourable position, has also been the most successful in his contribu¬

tions to the collection. To both these officers great credit is due for the

successful part they have taken in the formation of the collection. To

Sir Adelbert C. Talbot, K.C.I.E., the British Resident in Kashmir, the

collection is under special obligation for the hearty support given to it

throughout, without which it would probably never have reached its

present dimensions.

List of

contributions.

The following is a list, in chronological

order, of the contributions to the collection,

received by me from time to time : —

(1) From Mr. G. Macartney, a collection of small fragments of

manuscripts, found near Kuchar. It was received by me

early in April 1895 from the Foreign Office, with their

D. O. letter, dated the 28th March, 1895, and reported on

in the Proceedings, As. Soc. Beng., for May 1895, and

in the Journal, As. Soc. Beng., Vol. LX VI, for 18 97,

pp. 213-224.

(2) From Captain S. H. Godfrey (G. 1), a collection of frag¬

ments of manuscripts, received by me towards the end of

November 1895, and called the “ Godfrey Manuscripts.”

They are said to have been found near Kuchar. A preli¬

minary report on them was submitted by me to the

Government of India on the -|-^th March, 1897, and a fuller

report (with facsimiles) was published by me in the Journal

iv

Dr. Hocrnle — Antiquities from Central Asia. [Extra No 1 ,

of the Asiatic Society of Bengal, Vol. LXVI, Part I, No. 4,

for 1897 (pp. 225 ff.), advance-copies of which were

transmitted to the Government of India on the 23rd

August, 1897, and to the 11th International Congress of

Orientalists in Paris in October 1897.

(3) From Mr. G. Macartney (M. 1), a collection of six sets of

manuscripts, received by me in December, 1896, from the

Foreign Office with their D. 0. letter, dated the 14th De¬

cember, 1896, and called by me the “ Macartney Manu¬

scripts.” This collection was also dealt with by me,

together with the Godfrey collection (No. 2), in the two

Reports above referred to.

(4) From Mr. G. Macartney (M. 2), a collection of miscellaneous

antiquities, found in the Takla Makan desert. This

collection, distributed in four sets, consisting of a skull,

a manuscript, pottery, bronze figures, coins, and sundry

other objects, was received by me in August 1897, with

a letter of the Revenue and Agricultural Department of

the Government of India, No. 2283-27-1, dated the 16th

August, 1897. A preliminary report on it was submitted

by me to the Government of India on the 15th September,

1897.

(5) From Captain S. H. Godfrey (G. 2) a manuscript book,

said to have been found in the Takla Makan, and received

by me early in October, 1897.

(6) From Captain S. H. Godfrey (G. 3), a collection of manu¬

script leaves, in three sets, found in the Takla Makan, and

received by me towards the end of October, 1897, with

a letter of the Revenue and Agricultural Department of

the Government of India, No. 3043-27-3, dated the 14tli

October, 1897.

(7) From Mr. G. Macartney (M. 3), a collection of miscellaneous

antiquities, procured from Khotan and the Takla Makan,

consisting of (a) thirteen books, (6) pottery, (c) coins,

(d) sundry objects. Seven books and the antiques were

purchased by Mr. Macartney in Khotan for Rs. 95 ; the

remaining six books were purchased by him from Badrud-

din. The total cost was Rs. 150. The collection was

received by me early in November, 1897, with a letter

No. 3169-27-7, from the Revenue and Agricultural

Department of the Government of India, dated the 26th

October, 1897.

1 899.] Introduction . v

(8) From Captain S. H. Godfrey (G. 4), a collection of miscella¬

neous antiquities, consisting of coins, seals and one lead

figure, obtained from a merchant Miyan Ghulam Rasul at

a cost of about Rs. 30. It was received by me about the

12lh November, 1897, and reported on to the Government

of India on the 20th of the same month.

(9) From Mr. G. Macartney (M. 4), a collection of miscella¬

neous antiquities, obtained from Khotan for Rs. 11-3, and

consisting of (a) books and detached inscribed leaves, and

(6) sundry other objects. This collection was received

by me towards the end of November 1897, and is referred

to in the letter, No. 3265-27-8, of the Revenue and

Agricultural Department of the Government of India,

dated the 3rd November 1897.

(10) From Sir Adelbert Talbot, K.C. I. E. (T. 1), a small collec¬

tion of twenty-four coins and twelve detached manuscript

leaves, and sundry other objects, obtained from Muhammad

Ghauz of Khotan, through the Wazir Wazarat of Leh,

and received by me on the 3rd December, 1897, and re¬

ferred to in the letter, No. 3480-11-8, of the Revenue

and Agricultural Department of the Government of India,

dated the 3rd December, 1897.

(11) From Mr. G. Macartney (M. 5), two books, from Khotan,

purchased for Rs. 40, and received by me on the 27th

December, 1897, with a letter, No. 3734-11-23, from the

Revenue and Agricultural Department of tbe Government

of India, dated the 27th December, 1897.

(12) From Captain S. H. Godfrey (G. 5), a miscellaneous collec¬

tion found in the Takla Makan, consisting of (a) two

books and four detached leaves of manuscript, (b) three

mud figures, and (c) eleven coins. It was received by me

on the 9th January, 1898, and reported on to the Govern¬

ment of India on the 11th of the same month.

(13) From Captain S. H. Godfrey (G. 6), a collection of pottery

and three copper figures, found in the Takla Makan, and

received by me on the 27th January, 1898.

(14) From Captain S. H. Godfrey (G. 7), a miscellaneous collec¬

tion of antiquities, found in the Takla Makan, and consist¬

ing of (a) three books and six detached inscribed leaves,

( b ) coins, (c) seals, (d) pottery, (e) metal figures, and

(/) sundry other objects. This collection was received by

me on the 25th of February, 1897, with a letter, No,

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

(15)

(16)

621-1-7, from the Revenue and Agricultural Department

of the Government of India, dated on the same day.

From Mr. G. Macartney (M. 6), a collection of miscellan¬

eous antiquities, consisting of (a) six books, (6) coins,

(c) seals, (d) pottery, (e) metal figures, (/) two stone

heads of Buddha, and (#) sundry other objects. Three of

the books were purchased from Badruddin for Rs. 40,

the other four books and all antiques from the Rev.

Hogberg for Rs. 200: total cost Rs. 240. This collection

was received by me in Simla, on the 16th June, 1898,

with a letter, No. 1297-1-14, from the Revenue and

Agricultural Department of the Government of India

dated

19th May

16th June’

1898.

From Mr. G. Macartney (M. 7), a collection of nine books,

found at Aq Tala Tiiz, and purchased for Rs. 40. It was

received by me on the 19th June, 1898, with a letter,

No. 1645-1-21, from the Revenue and Agricultural Depart¬

ment of the Government of India, dated the 17th June,

1898.

(17) From Mr. G. Macartney (M. 8), a collection of eight books,

found at Kiang Tiiz and purchased for Rs. 40. It was

received by me on the same day as the preceding con¬

signment.

(18) From Captain S. H. Godfrey (G. 8), one book, purchased

for Rs. 45. It was received by me on the 16th July, 1898,

with a letter, No. 1884-1-29, of the Revenue and Agricul¬

tural Department of the Government of India, of the

same date.

(19) From Captain S. H. Godfrey (G. 9), one book, purchased

for Rs. 40. It was received by me on the 8th August,

1898, with a letter, No. 2097-1, of the Revenue and

Agricultural Department of the Government of India,

of the same date.

(20) From Mr. G. Macartney (M. 9), a collection of miscellan¬

eous antiquities, from Khotan, purchased for Rs. 20, and

consisting of (a) two books, ( b ) eight sheets of manu¬

scripts in Brahmi characters, (c) three sheets of Chinese

manuscripts (two dated), ( d ) two painted figures in stucco

(one Buddha), (e) a painted wooden board (Buddha),

(/) two terracotta figures; (g) one metal figure, (h) a

few coins, seals, and (i) other miscellaneous objects. This

collection was received by me on the 11th October, 1898,

1899.]

Introduction.

Vll

with a letter, No. 2776-1-41, of the Revenue and Agricul¬

tural Department of the Government of India, of the

same date.

(21) From Captain S. H. Godfrey (G. 10), a collection of miscel¬

laneous antiquities, from the Takla Makan, sent, from

Leh, and consisting of (a) three blockprints, (b) a carved

wooden box, (c) seven seals, ( d ) 54 coins, (e) 56 terra¬

cottas. This collection was received by me on the 29tli

November, 1898, with his D. 0. letter, dated the 19th

November, 1898.

In the subsequent part of this report the contributions above

enumerated will be referred to by the bracketed indications “ M. 1.,

G. 1,” etc.

With regard to the contribution No. 12 (or G. 5), I should explain

that of the objects constituting the collection the two books, also two

leaves, one mud figure and three coins have been retained for the

British collection. These were replaced by objects of a similar charac¬

ter taken from other portions of the British collection ; and the whole,

thus reconstituted, Captain Godfrey was permitted by the Government

of India (see No. 159-1-4 of the Revenue and Agricultural Depart¬

ment, dated the 19th January, 1898), to give to Mr. Dauvergne for

presentation to the “ French Archaeological Society.” They are now,

as I learn from a letter received by me from M. A. Barth, in the hands

of M. E. Senart in Paris.

In addition to the antiquities, above enumerated, I may here mention

one which I have also seen and examined, but

Gold coin from which does not belong to the British collection,

Khotan. but is the private property of Captain Godfrey.

It is a gold coin regarding the acquisition of

which Captain Godfrey, in private letters, dated the 30th October and

25th November, 1897, informs me as follows : “ I had heard that gold

coins were occasionally found by native treasure-seekers in the Takla

Makan. As such coins are not liable to oxidisation, it was fairly

certain that any obtainable might be of considerable numismatic, if not

historical, interest. I accordingly requested Munshi Ahmad Din and

other native gentlemen to make enquiries as to whether a specimen

could be procured. Dr. Chiraghu-d-din, formerly Hospital Assistant at

Kasbghar, very kindly obtained one for me from a Muhammadan priest

in Khotan who had himself purchased it, together with certain seals,

from a treasure-seeker there. The coin was, it is stated, exhumed

from the soil in the desert together with those seals. For the latter

a very high price was asked, for which reason the doctor did not buy

viii Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

them.8 Little over the intrinsic value of the gold was, however, asked

for the coin itself. The latter fact seemed to lend added interest to the

scientific character of the find, since it was clear that if the coin had

been a spurious one, manufactured for sale as an antiquity, a much

higher price would have been demanded.” According to Munshi

Ahmad Din, “ two other gold coins, seemingly of a similar description,

were presented by the Russian Consul-General at Kashghar to H. I. M.

the Tsar.” Captain Godfrey’s gold coin is shown on Plate I, fig. 1.

Being but imperfectly supplied with reference-books, I have not

been able to fully identify it ; but it appears to be a Byzantine coin,

perhaps of the Emperor Constantinus V, Copronymous, who died in

795 A.D.

With the exception of the antiquities, composing the contribution

G. 4, all the others were procured from

Localities and Eastern (or Chinese) Turkestan. The antiqui-

Cireumstances of the ties G. 4 (coins and seals) come from Samar-

Finds. kand, Tashkand, and other places of Western

Turkestan. The rest of the antiquities come

from the neighbourhood of two places, Kuchar and Khotan, in

Eastern Turkestan. Their find-spots are shown in red on the accompany¬

ing Map. The town of Kuchar lies to the North, and Khotan to the

South of the Great Sandy Desert, which occupies nearly the whole of

the space intervening between the Tian Shan Mountains in the North

and Kuen Luen Range in the South. The southern portion of this

great desert which lies immediately North of Khotan, bears the name

of Takla Makan, and most of the find-places are situated within the

limits of this portion of the sandy desert. In fact, there are only two

places near Kuchar, from which, any antiquities in the British collec¬

tion have been procured. These are a mound and a “tower” (i.e., a stupa),

situated 1 mile and 16 miles respectively to the west of that town. In

the stupa the Bower and Weber Manuscripts were found. In the same

place were also found some of the Macartney Manuscripts (viz., M. 1,

Set I a and b). The fragments, composing the Godfrey Manuscripts

(G. 1), as well as some fragments, included in M. 3 and T. 1, are also said

to have been found near Kuchar, but the exact place of their discovery

is not known. As all these fragments are strikingly alike with respect

to paper, writing and general appearance, it is probable that they were

all found at the same time and in the same place. On this point the

only information available is that given by Captain Godfrey (in a private

letter to myself ; see my Report in the Journal of the Asiatic Society

8 Possibly the seals here mentioned are identical with the pieces of yellow

crystal referred to in Mr. Macartney’s Note quoted below (p. xxxii).

1899.] Introduction . i x

of Bengal, Yol. LXYI, pp. 225, 226), that he received his fragments

(G-. 1) in the autumn (September) of 1895, and that he was told that

they were dug np c< near some old buried city in the vicinity of Kuchar.”

This last statement would seem to show that they were not found on

or near the old “ tower ” of the Bower Manuscript ; for the latter locality

was not the site of a sand-buried city. The fragmentary state of the

manuscripts (specimens are shown on Plates ii-viii of my Report,

above referred to) tends to prove that they were really the proceeds of

indiscriminate digging on some ancient site, which was probably being

explored with the hope of finding treasure. A good number of such

fragments must have been carried off at that time by the diggers ; for

only a portion of them were given to Captain Godfrey in September,

1895. He received them through certain Afghan merchants trading

to Yarkand. Another (very small) portion was included (so far as I

can now recollect) in the consignment M. 3, the items of which were

purchased by Mr. Macartney “ from some treasure-seekers ” in Khotan

when he visited that town in the spring of 1897. A third (also small)

portion is included in the consignment T. 1, the items of which were

purchased in October, 1897, by Sir Adelbert Talbot from a certain

Muhammad Ghaus of Khotan through the Wazir Wazarat of Leli.

The manner in which the treasure-seekers treated their find clearly

illustrates their policy, of which more examples will be found further

on. It is to divide their spoils into small portions which they dispose

of to different purchasers at different times. On the whole I am rather

disposed to believe that all these fragments (G. 1, M. 3 and T. 1)

really come from the neighbourhood of Khotan, and were dug out

(probably in the summer of 1895) from some sand-buried place in the

Takla Makan. I suspect that in the statement made to Captain

Godfrey about the “ old buried city in the vicinity of Kuchar,” there

is some mistake, and that Khotan is really meant instead of Kuchar.

There is, however, a third collection, also of fragments, which was

really found near the latter town. They are the first consignment on

my list, above given, and were dug out (apparently in September or

October 1894) from a large mound, a little more than one mile west of

Kuchar, by the orders of the Chinese Amban of that place. See my

Report on these Fragments, printed in the Journal of the Asiatic Society

of Bengal, Yol. LXYI, pp. 213, 214. According to the Amban’s account,

that mound had already been dug into “ several years ” previously, and

on that occasion “ some ” manuscripts had been found of which no

further information could be obtained. It is quite possible that some of

the G. 1, M. 3 and T. 1 fragments, which in point of paper and script

resemble the Weber MSS., may have come., out of that find.

J. i. 2

X

Dr. HoernTe — Antiquities from Central Asia. [Extra No. I,

The Macartney MSS., as already stated, were dug out, together with

the Bower and Weber MSS., from an ancient

Diseoveiy o e jj^dhist stupa situated sixteen miles west of

Macartney MSS. 1

Kucliar, on some barren rocky hills, close to

the left hank of the river Shahyar. These manuscripts have had a

curious history of which I may give a brief account. It is mainly

based on a Note by MunshI Ahmad Din, kindly procured for me by

Captain S. H. Godfrey. In a few particulars it corrects the accounts

previously published by me in the Journal of the Asiatic Society of

Bengal, Yol. LXVI, pp. 238-210, and in the Proceedings for 1898,

pp. 63, 64. It appears that some time in 1889 some people of Kuchar

undertook to make an excavation in the stupa in question. Their object

in digging into the stupa was to find treasure, as it was well known that

in the time of Yaqub Beg much gold had been discovered in such ancient

buildings. Whether or not they found any “ treasure,” is not known,

but what they did find was a large number of manuscripts and detached

papers together with the bodies of a cow and two foxes standing. The

hole which they made into the stupa was excavated straight in, level

with the ground, and the manuscripts, accordingly, would seem to have

been found, in the centre of the stupa on the ground level, exactly in

the spot, where the original deposit of relics is usually met with in such

monuments. The manuscript books and papers were taken to the house

of the chief Qazi of the town, where a couple of days afterwards they

were seen by Haji Ghulam Qadir, heaped up in a corner, there being

“ a big sabad, or ‘ basket,’ full of them. On enquiry having been told

the whole story by the Qazi, he brought away a few of them, and later

on, early in 1890, he gave one of them, now known as the Bower

Manuscript, to Major (then Lieut.) Bower.3 The others he sent to

his younger brother Dildar Khan in Yarkand. These the latter took

with him to Leh in 1891. Here he gave one portion of it to tMunshi

A^mad Din, who in his turn presented his acquisition to Mr. Weber,

a Moravian Missionary. The latter transmitted it to me in Calcutta,

where, under the name of the Weber Manuscripts, specimens of it

were published by me in the Journal of the Asiatic Society of Bengal

(Volume LXII, for 1893). The remaining portion, Dildar Khan took

with him to India, where he left it with a friend of his at Aligarh, a

certain Faiz Muhammad Khan. On a subsequent visit to India in

1895, he re-took it from his friend, brought it back to Turkestan, and

S Major Bower calls Lima “Tnrki” merchant; but he is only such by reason

of having married a Tnrki woman, and having been settled in Kuchar for nearly

30 years. Originally he is an Afghan from Ghazni, and elder brother of Dildar

Khan.

1899.]

Introduction.

9

XI

presented it to Mr. Macartney. The latter forwarded it in 1896 to the

Foreign Office in Simla, whence it was transmitted to me, in Calcutta.

It was named by me the Macartney MSS. and specimens of it were

published by me in the Journal of the Asiatic Society of Bengal

(Volume LX VI, for 1897). What became of the rest of the

manuscripts in the Qazl’s house is not exactly known. It is

probable that Andijani merchants in Kuchar, who are Russian

subjects, must have got hold of some of them and transmitted them

to Mr. Petrovsky, the Russian Consul-General in Kashghar. The latter

forwarded them to St. Petersburg, where specimens of them were

published in 1892 by Dr. S. von Oldenburg in the Journal of the

Imperial Russian Archaeological Society, Vol. VIII. As late as 1894,

“ten manuscripts ” were reported by Dildar Khan, on the information

of his brother in Kuchar, to be in the possession of a certain Yusuf

Beg. Unfortunately the negotiations, set on foot by Mr. Marcartney for

the purchase of these manuscripts, fell through, owing to the Beg’s

denial of possession, from fear of the Chinese authorities.4 It is believed

that subsequently Mr. Petrovsky succeeded in purchasing them. If

this is correct, they should now be in St. Petersburg. The exact details

of the find are so curious that it may be best to quote Dildar Khan’s

account, kindly procured for me by Mr. Macartney in January 1898. I

translate from the original Urdu : “ I heard from my brother Ghulam

Qadir Khan that there was a dome-like tower near Kuchar at the

foot of a mountain. Some people said that there was a treasure in it ;

it must be searched out. Accordingly some people making a hole

in the tower, began to excavate it, when they found inside a spacious

room (ghar khanadar ). and in it a cow and two foxes standing. On

touching them with the hand, the cow and foxes fell to the ground as if

they were dust. In that place those two books were found packed in

wooden boards. Also there is in that place a wall made as if of stone

( diwar sang-he mudfiq ), and upon it something is written in characters

not known. It is said that a few years ago an English gentleman

( that is, Major Bower) went there, and having visited the place came

away. Nothing more is known.” With regard to the cow and the

foxes mentioned in the above- account Mr. Macartney remarks in his

covering letter : “ As far as I can make out, they must have been

found in the tower in a mummified condition. The art of stuffing

animals would not appear to have been unknown in ancient times.

M. Petrovski, the Russian Consul in Kashghar, informs me of having,

♦ This appears to be the incident, referred to in the Chinese Amban’s letter,

published by me in the Journal , Asiatic Society of Bengal , Vol. LXVII (for 1897),

. 213. The owner of the MSS. is there called Timur Beg.

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

Xll

some years ago, received from Kucliar a fisli contained in a box, found

buried in the ground.” Dildar Khan’s remark about the inscribed

stone- wall (a stone slab let into the wall?) is curious. It is, as

I learn from Mnnshi Ahmad Din, based on a statement by Qadir

Hakim Beg of Kucliar, who, passing through Yarkand in 1895 on a

pilgrimage to Mecca, was questioned on the subject of the discovery of

the manuscripts. He was requested at the time by Mr. Macartney to

procure a copy of the inscription ; but owing to his death in Mecca,

nothing more was heard of the slab. The truth of the report is

well worth further enquiry : if true, the inscription might prove to

be a most valuable record. At the same time, considering that the

“room” must have been in almost complete darkness and that the

explorers probably had no means of lighting it, it is not quite easy

to understand, how, with the exception of the manuscripts which they

brought away with them, they could identify the exact nature of what

they found inside. I may note, however, that also in the stupas of

Afghanistan occasionally similar curious deposits have been found. Thus

Masson relates (in the Ariana Antiqua , p. 110) that in “ Tope No. 11

of Hidda” there were found in “an interior cupola” “some human

bones and two or three animal teeth,” which were afterwards identified

as those “ of the ass, the goat and a species of deer.” Also with reference

to the “ spacious room ” I may note that similar large chambers, in the

form of “cupolas ” or cubical “ apartments ” have been found in

many of the “Topes and Tumuli” of Afghanistan. Thus, in “Tope

No. 2 of Kotpur there was discovered a large cupola with a diameter

of 12 feet.”6 In Buner, Dr. Stein found in the Takhtaband stupa a

cubical chamber, of 7 feet dimensions, which was lined with large and

carefully cut slabs.” This may illustrate the presence of an inscribed

slab in the Kuchar stupa.

Most of the antiquities, including all the pottery, coins and other

miscellaneous objects, as well as many manu¬

scripts and all block-print books, have come

from Khotau, from various sand-buried sites

in the Takla Makan desert. Fifteen of these sites, situated at various

distances, from 5 to 150 miles distant from Khotan, are now known,

though, only two of them, Borazan and Aq Safil, have been verified by

European visitors. Eor the remainder we have only the information

given by native treasure-seekers, principal among whom appears to be

a certain Islam Akhun of Khotan. These fifteen sites are : —

(1) Borazan ( ). This place was visited by Messrs. Hogberg

and Backlund, Swedish Missionaries in Kashghar, in 1897. It was

6 See Ariana Antiqua , pp. 65, 69 et passim.

Ancient sites near

Khotan.

1899.] Introduction. xiii

also visited by Mr. Macartney in the spring of the same year. In his

demi-official letter, No. 121, dated 21st July, 1897, he states that “It

is a largely populated village about 5 miles west of the Khotan Chinese

city. Some gold ornaments, beads, precious stones (diamonds and

hahiJc) and terracotta images have been discovered there.” In a private

letter, dated the 20th October, 1897, Mr. Macartney informs me that

when he visited Borazan, he “found a number of villagers engaged

in digging into the side of a loess cliff, the lower portion of which

visibly contained a large quantity of broken pottery, bones and decom¬

posed vegetable matter. The stratum in which the excavations were

being made was about 12 feet below the level of the present village.

The layer which lies immediately above this formation, and on which

stands the village, is also of loess ; but it is noteworthy that this

upper layer shows no streaks or stratification, a fact which points to

the conclusion that the deposit was formed during one single flood.

Geologically speaking, therefore, there appears nothing furnishing an

indication of the time during which the lower stratum had been covered.

I enquired of the villagers whether there was any tradition about

Borazan having once been destroyed by flood; but no information could

be elicited on this point. They knew nothing about it.” Mr. Backluud,

however, informs me, in a letter, dated the 10th October, 1898, that

“ Borazan is said to have been a large town with forty gates, which

was conquered by a Rustam who burnt it and led a rivulet into the

place. These things are said to have taken place before the Moslirn

time.” He adds that “ the town in the place seen by me is now buried

under the mud up to 25-30 feet, as it seemed to me. It is a find-

place for clay images. We also found there a bone, measuring 16’75"

in circumference. Whether it be a bone from a yak or an elephant, we

could not judge.” Most of the pottery, coins and other miscellaneous

objects, comprised in M. 2 and M. 3, are believed by Mr. Macartney to

come from Borazan, while the similar objects, comprised in M. 6, are

stated by Mr. Hogberg, from whom they were acquired, to have been

dug out in that place. It appears to me most probable that Borazan

marks the ancient site of the town of Khotan. At the present day

Khotan lies close to the left bank of the Yurung Qash (or ‘white

jade ’) river, and apparently about 8 or 10 miles to the east of the Qara

Qash (or ‘ black jade ’) river. In olden times, however, it seems to

have occupied a site nearly midway between those two rivers. Accord¬

ing to the Geography of the Ming dynasty (from 1368 A.D.) the Yurung

Qasli flowed 30 li, or 6 miles East of Khotan, and the Qara Qasli, 27 li

or 5| miles West of that town.6 According to other Chinese estimates

a See Abel liemusat’s Histoire de la Ville de Khotan , p. 112 ; also p. 19.

xiv Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

the distance from the Yurung Qash to Khotan was only 20 li or 4 miles.7

Anyhow, in those days Khotan appears to have stood on a site lying

about 4-6 miles West of its present one, and therefore coincident with

the site of Borazan, which is said to be “ about 5 miles west of Khotan.”

(2) Aq SapIl or Aq SafIl ( (jjl “white battlements”), an

uninhabited place in the desert, was visited by Messrs. Hogberg and

Backlund in the summer of 1897. It lies about 20 miles north-east

of Khotan. From this place was procured in the summer of

1896 a number of coins and miscellaneous objects of metal and

glass, comprised in M. 2, Set II, as well as the manuscripts, com¬

prising M. 1, Set II. A description of these manuscripts, together

with facsimile specimens, lias been published by me in the

Journal of the Asiatic Society of Bengal, vol. LXYI, pp. 237 and 251?

252. The coins include some of the curious ancient bilingual (Chinese

and Kharosthi) ones, which are referred to by me in my Presidential

Address in the Proceedings of the Asiatic Society of Bengal, for 1898,

p. 69. As these are referable to the 1st and 2nd centuries A.D. (see below

pp. 10-15 ), they indicate Aq Sapil to be a very ancient site. Turk! and

Chinese coins of the early middle ages were also found here. Among the

miscellaneous objects are several old metal seals, a small metal figure of

Buddha in meditation, and broken pieces of glass. The following notes,

made by Mr. Macartney from Mr. Hogberg’s account of his visit, are of

much interest : “ Aq Sapil is a town or rather the remains of a town in

the Takla Makan desert. It is one day’s journey north-east of Khotan,

to be reached through the oasis of Hangni.8 Before coming to Hangni

village, the remains of an old city may be seen. On leaving Hangni, the

traveller is at once in the midst of sand dunes which rise from 10 to 30

feet high. Between Hangni and Aq Sapil, the sites of two other ancient

cities may be seen, evidenced by the fragments of pottery and bones on

the ground. At this portion of the route, Mr. Hogberg picked up a few

old seals and coins. As Aq Sapil itself is reached, the remains of an

irrigation canal are noticeable, which canal must have once carried water

to the city and the surrounding country. It is from 8 to 10 feet broad

and, in places not choked up by sand, rather deep. At Aq Sapil the

1 See ibidem , pp. 21, 30.

8 The real approximate distance (about 20 miles) may be judged from the

following account of his tour by Mr. Backlund. “ Mr. Hogberg and I started from

Hangni at 5 o’clock in the morning, and arrived at Aq Sapil at 10 or 11 o’clock, having

marched slowly partly because we crossed places covered with pottery, and partly

because of the soft high sand dunes. At Aq Sapil we wandered about for a while,

took some food, and returned to Hangni. Having taken some rest there, we started

for Khotan, where we arrived a little before midnight, the same day, having then

had rather a strong march.”

1899.] Introduction. xv

remains of the city wall are still extant. The wall is from 10 to 12

feet high, and is made of square unburnt bricks (20 X 20 X 4 inches).

On most of these bricks one of the following marks is to be

found, Other bricks, again, have on them the prints of

the human foot. The section of the wall is shown in Woodcut No. 1.

Only about 25 feet of the wall can now

be seen. About the middle of it there

is an opening which must have been once

occupied by a gate. The ground out¬

side the wall is trough-shaped, and shows

that Aq Sapil was once surrounded by a

ditch. Near the gate and on the further

side of the ditch, the remains of two towers

(stupas?) are visible. They are filled with sand inside, and made of

earth on the outer side. Regarding the interior of the city the remark¬

able thing is that although no houses are left, yet the thoroughfares and

the places where the houses had once been can be easily distinguished

on whatever spot has not been overwhelmed by the sand.10 The ground

occupied by the streets is particularly hard owing to its having formerly

been constantly trodden upon, and the same observation applies to

the interior of the houses, but the comparatively soft soil on which

the house walls stood has been scooped out and forms a hollow. This

curious fact may possibly be attributed to the action of sandstorms

which have had less corroding effect on the trodden ground than on

the once wall-protected soil. The whole of the interior of the city,

where it is not covered with sand, is overstrewn with fragments of

pottery. The ground at one spot has the form of a couple of reversed

amphitheatres, there being two elevated circular stands, slightly

hollowed out like ponds, surrounded by terraces descending in widening

circles. This is illustrated by Woodcut No. 2. Treasure-seekers would

appear to have worked a great

deal amongst the debris of Aq

Sapil as evidenced by the heaps

of sifted earth, which may be

seen here and there.” The

exact spot where the manu¬

scripts M. 1, Set II were found

by Islam Akhun is not known.

It is only stated that “ the MSS.

•

9 These letters might be Kharosthi, and might mean da-di-de-sa.

10 By way of illustration I may quote Dr. Bellew’s description of the sand-

buried “ city of Nucta Rashid ” about 36 miles south-east of Yangi Hisar : “ It

xvf Dr. Hoernle — Antiquities from. Central Asia. \ Extra No. 1,

were found wrapped up in a piece of woolJen cloth and buried in about

3 feet of earth.” To judge, however, from the circumstances in which

manuscripts were found at Kok Gumbaz and at Qara Yantaq, it is not

impossible that the manuscripts M. 1, Set II, were actually dug out

from the hollow of one of the two mounds described by Mr. Hogberg.

(3) Aq Tala Tuz. The position of these

(4) Qara Qol Mazar (Kiiojam). five places may be approxi-

(5) Qara Tagh Aghaz!. mately determined from the

(6) Qara Yantaq. following itinerary of Islam

(7) Kok G-umbaz Akhun. which he gave to Mr.

Macartney. He stated that on one of his search-expeditions he started

from Guma which lies about 100 miles W. N. W. of Kliotan. Leaving

that town, “ with two other men, about the beginning of January

(apparently 1898) and travelling in a generally easterly direction, he

came to Qara Qol (Jt* 'kj* lit. ‘ black lake ’) where there is a salt water

lake covered with reeds. Qara Qol is reached from Guma in one march

(say 12 miles), the intervening ground being through cultivation.

Qara Qol itself is not inhabited. A Mazar (or ‘shrine’) may

however, be seen there. From Qara Q51, Islam Akhun went for

about 20 miles in a south-easterly direction through the sands to

Qara Tagli Aghazi ( b* or £ master of the black mountain’),

a village surrounded by sand and having about 45 houses. Water

had to be carried on a donkey from Qara Tagh Aghazi. After

three days’ march (say 21 miles) in the desert, in a generally

easterly direction, Kok Gumbaz was reached.. After another march

of about 8 miles going in the same direction, and over sandy

ground covered with withered reeds, Islam Akhun arrived at Qara

Yantaq, where the remains of an ancient town were seen. The walls

were no longer visible, but the place where they once stood was still

distinguishable. These traces extended in all directions for a long way,

and evidently Qara Yantaq had once been the site of a large town.

The remains of an old canal of about two yards wide were also noticed.

There is a tradition that Qara Yantaq was inhabited by Hindis (the

name by which Buddhists are generally called in Chinese Turkestan),

and that they were converted to Muhammadanism by one Mirza Aba

Balm. A Tazkira of this person is in the hands of an Imam named

Sadiq Akhun, now living at Qara Tagh Aghazi. From Qara Yantaq,

Islam Akhun went about 60 miles east over sand dunes, and came to

presents nothing to view bnt the outlines of the foundations of rampart walls and

bastions, now mostly buried by the drifting sand. Here and there, where the sands

have been swept away by the winds, the surface is strewed with fragments of

pottery and glass.5' ( Report of a Mission to Yarkand in 1873, p. 129).

1899.] Introduction. xvii

Aq Tala Tiiz, where a number of books were found. At Aq

Tala Tuz ( jy *13 <j>t or ‘ white salt-hill ’) the remains of mud

walls were extensively seen, whilst the ground was found to be

strewn with pieces of old iron, fragments of pottery, and bits

of wood. There was only one house which had the roof on, and

that was half buried in the sand which was heaped up against it at

one corner. As the door was not visible, a hole was made on one

of the exposed sides. This done, Takhtash, one of Islam Akhun’s

companions, crept in, and found himself in a small room of about

three yards square. This room was considerably filled with sand, so

much so that it was impossible for a person to stand up in it without

his head touching the ceiling. Takhtash found the books while digging

in the sand. There were many other books, but these were in such

a dilapidated condition that they crumbled to pieces on being handled.

Islam Akhun was too frightened to inspect the interior of the house

himself. At Aq Tala Tuz water was found by digging about a yard

into the ground.” This account, of course, must be taken quantum

valeat ; but there is nothing intrinsically improbable in the local de¬

scriptions, and the distances fairly agree with those given of the same

places at other times. Whether the discoveries of books said to have

been made in Aq Tala Tuz, were really made, is a quite different

question. The description of this place and of Qara Yantaq fairly

agrees with that given by Mr. Hogberg of Aq Sapil. The distance

between Guma and Kok Gumbaz, by this itinerary which was related

to Mr. Macartney in February, 1898, should be about 5 or 6 marches.

The same distance was mentioned to Mr. Macartney in October, 1896,

in connection with the find of M. 1, Set V, when Kok Gumbaz was

“said to be 5 days’ march east of Guma.” (See my Report in the

Journal, Asiatic Society Bengal, Vol. LXYI, p. 238.) At that time, M. 1,

Set IV, is stated to have been found in Qara Qol Mazar Rkojam, which

is said to be situated “ in the desert at 50 miles east of Guma” (see

ibidem , p. 238) . There can hardly be a doubt that this place is identi¬

cal with the Qara Qol of the itinerary, where a Mazar is stated to exist.

There, however, it is, said to be only one day’s march from Guma.

I am disposed to believe that the earlier report contains a mistake ;

for 50 miles probably 5 miles should be read, which would be about

one day’s march.11 The distance, by the itinerary, between Guma and

Aq Tala Tuz is about 119 miles. Natives of Turkestan, as Mr. Backlund

11 Can this be identical with the “ancient city” which Dr. Sven Hedin

visited from the village of Muji, at a distance of about miles north-east of the

caravanserai (see p. 736 in his book Through Asia) ? The ruins there are said to

have “ consisted principally of a number of tombs ” (mazar),

J. i. 3

xviii Dr. Hoemle — Antiquities from Central Asia. [Ex fra No. 1,

informs me, are very untrustworthy in their estimates of distances.

Allowing a discount of 25 per cent, for this uncertainty, and also for

windiugs of the march, the distance measured on the map may be

taken to be about 90 miles. This places Aq Tala Tuz not far to the

west of the Qara Qasli river, at a distance of about 30 or 40 miles

northwards of Khotan city. In favour of this determination it may

be noted that the itinerary does not mention the passage of either the

Qara Qash or the Yurung Qasli rivers. As Islam Akhhun’s expedition

took place in January, both rivers would have been, at that season, in

a very low state : still the total omission of the mention of the passage

of either river, if it did take place, would be very strange. The pro¬

bability, therefore, is that Aq Tala Tuz as well as all the other place

mentioned in the itinerary are situated in that part of the Takla Makan

desert, which lies to the west of the Qara Qasli river, and at a distance

from 10 to 20 miles north of the caravan route from Guma to Khotan.

At Aq Tala Tuz those nine block-prints are said to have been found

which are comprised in M. 7. Of their discovery an exceptionally

circumstantial account is given, which must be taken for what it may

be worth. As a rule, the only information obtainable about the block-

prints was that they were found near Khotan. The tinders or the

Khotanese merchants from whom they were obtained either could or

would give no further information.

At Qara Qol Mazar ( Jjj or ‘ black lake shrine of

my lord’), where there is said to be “an immense grave-yard in

ruins, about 10 miles long ” was found by Islam Akhun, in August,

1895, the manuscript M. 1, Set. IV, described and figured by me in

my Report in the Journal, As. Soc. Beng., Vol. LX VI, pp. 238

and 253, Plates xviii and xix. It was found simply lying on the sand,

probably uncovered by the action of the wind which had blown away the

superincumbent sand. When found, it is said to have been “ bound

between two little wooden boards” in the Indian fashion. These, having

been broken on Islam Akhuu’s journey to Kaskghar, where he sold the

manuscript to Munshl Ahmad Din for Mr. Macartney, were apparently

thrown away by him. This is a pity, as in the present state of our

knowledge of these Central Asian manuscripts every means of informa¬

tion is valuable.

At Kok Giumbaz ( or ‘blue dome’) were found by the

same Islam Akhun, in October, 1895, the two manuscripts M. 1,

Set. V and M. 1, Set. VI, also described and figured in the same

Report, pp. 238 and 253, 254, Plate xx. The latter manuscript is said

to have been simply “picked up from the ground” similarly to that

(M. 1, Set. IV) found at Qara Qol Mazar; but the other manuscript

1899.]

Introduction.

XIX

(M. 1, Set. V.) was found, enclosed in the remnants of “an iron box,”

in a hole situated apparently on the top of a circular platform.

According to Islam Akhun’s account, he “ saw a circular wall of baked

bricks three feet high, and at about J5 paces from it, there was another

wall, in which a hole plastered over with mud was discovered : in

removing this mud, the manuscript was found, contained in the

remnant of what once was an iron box.” This description reminds one

of the similar erections described by Mr. Hogberg as having been seen

by him in Aq Sapil. To judge from the latter description, which

is much more circumstantial, it would seem that what Islam Akhun

saw were two circular platforms about 3 feet high, the upper surfaces

of which were hollo wed out to hold relics.

At Kok Gumbaz were further found the manuscripts G. 3, Set III.

and the objects comprised in M. 2, Set IV. Captain Godfrey, in a

demi-official letter, No. 5208, dated the 15th September, 1897, and

addressed to Sir Adelbert Talbot, Resident in Kashmir, gives the follow¬

ing account of the discovery of the manuscripts. They were “ enclosed

in what seem to be the rotten remnauts of a cloth or cotton covering.

This I have not attempted to open, since the whole should possibly be

carefully steamed in order to prevent the brittle contents breaking up.

This work would be best performed by trained hands. One point of

interest in connection with it is the alleged fact that it was found along

with another manuscript said to have been purchased by Mr. Macartney

and transmitted to the Royal Geographical Society in Loudon. They

were both brought to Kashghar by a treasure-seeker (apparently Islam

Akhun), from whom the majority of the manuscripts have been pur¬

chased by Mr. Macartney and Munshl Ahmad Din. Both — the manus¬

cript above alluded to and that now sent — were wrapped in different

bags, and were stuck fast one upon another to a human skull. The

site of the discovery was a place called Kok Gumbaz, five days’ march

from Khotan.ia This place is seemingly an old graveyard. A small

mound of earth was seen there in the middle of the surrounding sand.

The treasure-seeker examined this. The dust crumbled away at the

touch, and two feet underneath the surface he found the manuscripts

and the skull referred to.” On receipt of the bag, it was opened by

me, and was found to contain two folded sheets, each inscribed on one

side. The manuscript, mentioned by Captain Godfrey as having been

transmitted to the Royal Geographical Society, is now deposited in the

British Museum, as will be seen from the following extract from a

12 In other accounts it is said to be five days’ march east of Gama. Both may

be correct, for it will be seen from the map, that Kok Gumbaz lies roughly midway

between Guma and Khotan, east of the former and north-west of the latter town.

XX

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

private letter to myself of Mr. Cecil Bendall, dated London, tlie 1st

October, 1897 : “ 1 tbink it may interest you to know that Mr. Macartney

sent us here two collections of fragments similar to some of those des¬

cribed by you at pp. 38 If. (of Extra-copies, corresponding to pp. 250 ff.

of my Report in the Journal, As. Soc. Beng., Vol. LX VI). We have (1)

a “book,” very similar to that described by you and figured in your

Plate xx. The peg is wood, not metal, but it comes through in about

the same part of the leaves, which are very dirty brown paper like

yours. The “ book” has blank leaves at both ends. The writing is

mostly that figured in your Plate xv; but several leaves (apparently

occurring at random) are writing in the script of your Plate xvii with

those odd ‘ ligatures’, some of which, I think, must be of Syriac (Nes-

torian) origin. (2) A few leaves, showing rulings in double lines and

folded over. The writing here is certainly of Mongol origin.” A com¬

parison of these different accounts suggests that the “mound”, in which

the skull with its pillow of manuscripts was discovered, is an

erection similar to those described by Mr. Hogberg and Islam Akhun.

The exact time when the discovery was made is no where stated. But

it is probable that it was made in October, 1896. For with regard to

the objects comprising M. 2, Set. IV, Mr. Macartney states that “ these

images and Chinese coins were found by Islam Akhun in October, 1896

along with manuscripts.” Moreover from Kok Gumbaz, Islam Akhun ap¬

pears to have gone on to Qara Yantaq where, in the November following,

he dug out the skull with its MSS. pillow which I shall next describe.

At Qara Yantaq (<3^^ or ‘ black thorn ’) was found, by Islam

Akhun in November, 1896, the skull with its MSS. -pillow just referred

to. In the same place were found two small horsemen of bronze, some

old coins and a large quantity of broken metal. The whole constitutes

M. 2, Set I. The story of the discovery, from information given by the

discoverer, is thus related by Mr. Macartney in his D. 0. letter, No. 58,

dated Kashghar, the 31st March, 1897 and addressed to the Resident in

Kashmir. “The skull with the manuscript adhering to it was found

by him in November, 1896, at Qara Yantaq, situated in the desert at

about five days’ journey east of Guma. The soil of Qara Yantaq is

described to be of loess. Here and there are to be seen, along what

must have been once the bed of a river,18 some rushes still rooted in the

ground, but withered and blackened by want of moisture and by exposure.

At Qara Yantaq, there is one solitary mound, circular in shape, and about

5 feet in diameter and 2 feet in height. The skull with the manuscript

adhering to it was discovered on this elevation, and was partially buried.

IS This, as may be seen from the Map, should be the dry bed of the river which

flows past Pialma, on the caravan route from Yarkand to Khotan.

1899.]

Introduction.

XXI

The two images of horsemen were dug up from the interior of the mound.

The other objects were picked up from the surrounding country.” The

whole ot the find was transmitted to me by Mr. Macartney, especially

the skull resting on its bag, exactly as it had been found. On opening

the bag, it was found to contain a manuscript book, similar to those

found at Kok Gumbaz and Qara Qol Mazar, but with its leaves cut in

a very peculiar shape. The skull, on examination by Dr. Alcock,

Superintendent of the Indian Museum in Calcutta, was found to be

of the Mongolian type. The mound in which it and the horsemen were

found is no doubt similar to those existing in Kok Gumbaz, Aq Sapil

and other places. The combination of the objects found in it would

seem to indicate it to be the sepulchral monument of an ancient

chieftain. This and the finding of old coins indicates Qara Yantaq to

have been a very ancient settlement.

(8) Yabu Qum or ‘load-ponies’ sands’)14 is said to be

situated 50 or 60 miles north-east of Khotan, and is the place

where the manuscripts of M. 1, Set III, are said to have been found.

These have been described and figured by me in the Journal of the

Asiatic Society of Bengal, Vol. LXVI, pp. 238 and 253, Plate xvii.

The exact time of their discovery is not known, but they were

purchased from Islam Akinin, the finder, in July 1896. Probably

they were found in the autumn or winter of 1895, about the time when

the finds at Qara Qol Mazar and Kok Gumbaz were made. Islam

Akhun stated that “ at Yabu Qum some ruins of a mud wall are still

visible,” and that “ the manuscripts were found wrapped up in a piece

of cloth and mixed up with human bones, the whole lying on some

partially exposed boards of a wooden coffin.” Putting this together

with what we know of the circumstances of the finds at Aq Sapil, Kok

Gumbaz and Qara Yantaq, it may be concluded that the “ mud wall ”

belonged to one of those circular mounds, and that the “ human bones ”

may have been the fragments of a skull, which had rested on the wrapped-

up manuscripts. As to the real nature of the boards of the so-called

“ wooden coffin,” it is premature to make any conjecture. It appears

to me probable that the manuscript sheet G. 3, Set I, was also found at

Yabu Qum. For that manuscript is said (in Captain Godfrey’s demi-

official letter, No. 5208, dated the 15th September 1897, and addressed

to the Resident in Kashmir) to have been found “ at a place 50 or 60

miles north-east of Khotan in the midst of the Takla Makan desert; ”

and Yabu Qum is also said “ to be situated at 50 or 60 miles north¬

east of Khotan in the midst of the Takla Makan desert ” (see my

14 Mr. Backlund suggests that the name marks a spot where a caravan was lost

in the sands.

xxii Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

Report in the Journal of the Asiatic Society of Bengal). It may

however, also have come from Dandan Uiliq, which, to judge from the

bearings and distance given of it (as may be seen from the Map)

cannot be very far distant from Yabu Qum.

(9) Kiang Tuz (perhaps incorrect for Kan Tuz or ‘salt

mine ’) is said to be a place situated about 150 miles east of Khotan, on

the road to Charchau. Here the eight block-print books, comprised in

M. 8, are said to have been found by Islam Akhun ; but this information

requires to be received cum grano salis .

(10) Dandan Uiliq. None of these six places are speci¬

al) Imam Aptah Mazar. lically mentioned as spots where any

(12) Khitai Uiliq. of the objects comprised in the

(13) Qotaz Langar. . British Collection have been found,

(14) Sultan Wais Qiran. though some of the objects of which

(15) Tam AghIl. the exact find-place is not stated may

have come from one or the other of them. All six are stated to be

places which are frequented by treasure-seekers from Khotan. They

are described by Mr. Macartney in a Note, attached to his demi-

official letter, No. 121, dated Kashghar, the 21st July, 1897, as follows :

Dandan Uiliq ( or ‘ Ivory House ’), at about 6 days’

journey north-east of Khotan; the remains of a Bazar, half-buried in

sand, is said to be here, the stalls, which contain piece-goods, crumbling

to dust at the touch, being still visible. The ruins of a Serai, But-khana

(or ‘idol-house’) and a Mill can also be discerned. Being situated

right in the desert, it is considered to be a difficult place to reach.

There is no water on the way ; but water may be found at the place

itself by digging at the foot of a solitary tree which is still green.

Discoveries : manuscripts, tea, weaving machines, coins, hakik and

lajivar stones, and pearls.

Imam Aptah Mazar (j\y° ^ I or ‘ Shrine of Imam Aftah or

Aptah, one of Khalif ‘ Omar’s men), about 14 miles north-west of

Khotan ; inhabited. Discoveries : seals, money, and hakik stones.

Khitai Uiliq ( &kj\ or ‘ Chinese House ’) said to be situated

H

about If march north of Khotan, near the Aksu road. It appears that

the ruins of a few houses are extant here.

Qotaz Langar (jij jb>- or * resting place of yaks’) ; on the

Khotan-Polu road, at about If march from the Khotan city. The place

is described as situated in the midst of sand-hills and inhabited by

two families. Discoveries : manuscripts and gold coins.

Sultan i Wais i Qiran ( perhaps the name of a

Governor) ; one march north-east of Khotan. Discoveries : manu¬

scripts, silver hooks, coins, and oue wooden idol.

Introduction .

xxm

1899.]

Tam AghIl (<-Ml (*G or * a walled enclosure for cattle *) ; half-march

north-east of Kliotan ; inhabited ; gold ornaments and seals are said to

have been found in this village. A human skull, with a layer of gold and

with a ring attached to the nose, appears to have been picked up here.

For convenient reference the following Table gives a summary of

all the antiquities, the localities and times

Summary. when they were found or received, and the

persons who found them or through whom they were received. By

“MSS.” are understood manuscript or xylograph books, or pothi bound

in the Indian fashion ; also detached leaves or sheets. By “ books ”

are meant volumes, either manuscript or xylograph, which are bound in

the European fashion. Dates, placed within brackets, are those of the

receipt of the antiquities, their dates of discovery being unknown.

xxiv Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

The Takla Makan desert appears to have received its name from

the large quantities of broken pottery, which

are found strewn about in many places,16 and

which show that, in ancient times, parts of

Description of the

Takla Makan Tract.16

16 Compare the account of this Tract in the Report of a Mission to Yarkand in

1873, pp. 25 ff.

18 Mr. Biickluncl informs me that “ Takla Makan is a peculiar word which the

natives apply to places covered with pottery. Such places are very numerous.

Also many skeletons can be found in those places.” Mr. Macartney also writes to

me that “ the fragments of ancient pottery, images, etc., are not always found

imbedded as at Borazan. They are often seen lying on the surface of the ground.

I was much struck, along the road between Guma and Jhanguia, with the frequent

appearance of pieces of broken earthen- ware vases (of no artistic value) covering

large areas of ground, such areas being themselves situated in the. midst of a sand-

desert, and often 5 or 6 miles from habitation.” The word ‘ Takla Makan ’ is not

as Dr. Sven Hedin appears to state (pp. 450, 457, 785, 801 in his most interesting

book Through Asia) the proper name of any one particular place.

1899.] Introduction. xxv

it must have been the seat of an advanced civilization. Among the

peculiar features of the tract of country which comprises that desert

and the adjoining belt of cultivation are the numerous rivers which

come from the valleys of the Karakorum and Kuenluen Ranges, and

traverse its surface to their final junction with the Tarim river.

In the present day it is only the two largest of these rivers, the

Yurung Qash on which Khotan lies, and the Qara Qash on which

the homonymous town lies, which, uniting north of Khotan, reach

that termination. The smaller rivers, some of which may have

been tributaries of the two large ones, after emerging from the

valleys, now soon lose themselves in the sandy desert. “ These

rivers large and small, are the seats of the fixed population and

the entire productive industry of the country. ” “Numerous canals

are drawn off from them to the lands on each side, and thus

convert considerable tracts of what would be otherwise desert-waste

into fertile and populous settlements.” At present the extent of these

settlements is very limited, but formerly — many centuries ago — they

extended much further into the interior, probably some 30 to 50 miles

beyond the present borders of the sandy desert. The climate of the

country is notable for “ the extreme dryness of its atmosphere at all times

and the trifling amount of its rainfall.” As a consequence “the soil

everywhere is characterised by its aridity and barrenness, and is more or

less highly charged with salines, which retain sufficient moisture to form,

in the desert, mud bogs and marshes on which grow coarse reeds and

dwarf tamarisks.” In these circumstances it is only by careful irriga¬

tion that the area of cultivation can be preserved and the encroach¬

ments of the moving sands of the desert prevented. The appearance

and action of these moving sands has thus been described by Dr.

Bellew : “ During the spring and summer months a north or north¬

west wind prevails, blowing with considerable force and persistence for

many days consecutively. As it sweeps over the plain, it raises the

impalpable dust on its surface, and obscures the air by a dense haze

resembling in darkness a November fog in London ; but it drives the

heavier particles of sand before it, and on the subsidence of the wind,

they are left on the plain in the form of ripples, like those on the sandy

beach washed by an ebbing current.” In course of time, there is formed

“ a perfect sea of loose sand, advancing in regular wave lines from

north-west to south-east. The sand dunes are mostly from ten to

twenty feet high, but some are seen like little hills, fully a hundred

feet high, and in some spots higher. They cover the plain, of which

the hard clay is seen between their rows, with numberless chains of

two or three or more together in a line, and follow in successive rows,

J. i. 4

XXVI

Dr. Hoerule — Antiquities from Central Asia. [Extra No. 1,

one behind the other.” When, during the earlier centuries of the

Christian era, in consequence of the troubles attending the tribal

migrations and the Muhammadan conquests, the population became

reduced and irrigation fell into neglect, the advancing sands gradually

overwhelmed one outlying settlement after the other, and narrowed

the belt of settlement and cultivation to its present limits. Many

traces of these ancient settlements and the water-courses on which they

lay are still met with in the desert ; and some of them have been named

and described above (pp. xii ff. )17 The recollection of the desert having

been once a fertile and populous country still survives in that region.

Mr. Macartney, in his demi-official letter, No. 121, dated the 21st July,

1897, reports that “ it is believed by the natives of Kashgharia that the

Takla Makan desert was once a fertile and cultivated country. There

is a tradition that before the introduction of Muhammadanism [in

the 10th and lltli centuries A. D.], forty-one cities 18 flourished in that

region under the rule of a certain Zewar Shah, king of Katak, and that

hJ reason of the disbelief of the inhabitants in the religion of the Pro¬

phet, which three Imams from Bukhara had come to preach, their

country was suddenly and miraculously destroyed by a sandstorm. This

story is told at considerable length in the Taskirali of Kamaluddin,

Zahiruddiu and Khwaja Arush.19 The natives believe that the antiques

which are constantly found in the Takla Makan desert belonged to the

cities which once formed part of the kingdom of Zewar Shah.” The

exact site of the city of Katak, here mentioned, is not known, but it prob¬

ably lay about three marches or 40 miles south-west of Lop Nor, on the

great trade-route from China to Khotan, which ran by way of C barchan.20

Its destruction by the sands occurred about 1330 A. D., and it is prob-

n A most interesting account of some others which Dr. Sven Hedin discovered

during his travels in the Takla Makan is given in his book Through Asia.

On p. 496 of Dr. Sven Hedin’s book Through Asia, another tradition con¬

cerning the former existence of “forty towns” is mentioned.

19 This Taskirah had been transmitted to me in the box which contained the

consignment M. 3. It has been returned to Mr. Macartney for the purpose of

translation and eventual publication.

20 This is the view advocated by the late N. Elias in his Translation of the

Tarikh-i-Kashidi, pp. 11, 12, footnote. Dr. Sven Hedin discovered a place, called

Katak, on the banks of the Khotan river (see p. 819 of his book Through Asia) ;

but he also met with the name further east, near the Yarkand river (see ibidem,

p. 473), and also further west, “ a long way ” south of the Achik (or old Tarim) river

(see ibid., p. 850). It would seem probable that katak should really be kottek or

dead foiest (see ibid., p. 811, 850), and that the name is shahr-i-kottek or ‘ town

in the dead foiest. Like Takla Makan it is not the proper name of a particular

place, but a general name common to a number of old sites.

1899.]

Introduction.

xxvii

ably on account of this comparatively modern date that we possess a

detailed account of the catastrophe. It may be seen in Mirza Muham¬

mad Haidar’s Tarikh-i-Rashidl (English Translation by N. Elias,

p. 10 ff.,) written between 1541 and 1546 A.D. That writer thus

describes the condition of the desert in his time ( ibidem , p. 295) : “ To

the east and south of Kashghar and Khotan are deserts which consist

of nothing but heaps of shifting sands, impenetrable jungles, waste

lands and salt-deserts. In ancient times there were large towns in

these wastes, and the names of two of them have been preserved, name¬

ly Lob and Katak; but of the rest no name or trace remains : all are

buried under the sand. Hunters who go there after wild animals,

relate that sometimes the foundations of cities are visible, and that

they have recognized noble buildings, such as castles, minarets, mosques

and colleges, but that when they returned a short time afterwards, no

trace of these was to be found ; for the sand had again overwhelmed

them.” This fact of the recurrent disappearance and reappearance of

sand-buried sites and ruins naturally follows from the action, above

.described, of the winds on the sands, and has also been noticed by

modern travellers.21 It also forms a welcome occasion for the visits of

treasure-seekers, especially in Khotan, where, as Mr. Macartney informs

us, they make a regular livelihood of that occupation, being in the

habit of visiting, after a sandstorm or a flood, such localities as seem

most promising, in the hope of picking up some objects in gold or silver

which have been laid bare by the wind or water. As an example of

such a visit the itinerary of Islam Khan has been given above.

As already stated the process of submergence of the ancient civili¬

zation of Eastern Turkestan under the advancing sands of the desert had

already commenced long before the Muhammadan period. It was already

in full operation at the time of the Chinese Buddhist pilgrim, Hiuen

Tsiang, in the seventh century A.D., when Buddhism was still the

prevailing religion and culture of Khotan and the rest of Eastern

Turkestan. On his return from India to China, in the year 644 A.D.,

he took the southern route passing through Khotan and Pima to

Charchan : tbe northern route, by which he had travelled from China

to India, passed through Kuchar and Kashghar. In those days Pima

was a comparatively new settlement, lying on the outskirts of the Takla

Makan desert, and it still existed for many centuries afterwards, for in

1274 A.D. Marco Polo saw it on his way to China. At the present day

it has disappeared in the sands, and its exact site is not known. Ac¬

cording to the data furnished by Hiuen Tsiang’s itinerary, it must have

21 See, e.g., Dr. Bellew’s observations on the subject in the Report of a Mission,

to Yarkand in 1873, pp. 28, 29, 37, 38.

XXV111

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

Conservation

of Antiquities.

lain about 100 miles East of Khotan, which would place it somewhere to

the north or north-west, of the present town of Kiria. The latter would

seem to have taken the place of Pima, when it was overwhelmed by the

sand, just as, according to Hiuen Tsiang’s account, Pima itself took

the place of the still more ancient town of Ho-lo-lo-kia, which lay

further north-west. Possibly Ho-lo-lo-kia may have occupied the site of

the present Dandan Uiliq, which is said to lie 6 marches or about 80

miles north-east of Khotan. The description of that place, given

above, would well enough suit a place such as Ho-lo-lo-kia might have

been.

The physical conditions of the Takla Makan desert, with the

extreme dryness of its atmosphere and the

trifling amount of rainfall, above referred to,

are very favourable to the conservation, for

an indefinitely long period, of everything buried under its sands. This

has been repeatedly observed by travellers ; see, e g., the remarks of

Dr. Bellew in the Report of a Mission to Yarkand in 1873, p 38, quoted

by me in the Journal of the Asiatic Society of Bengal, Yol. LXVI, p. 256.

It has been amply confirmed by Dr. Sven Hedin’s discoveries in Qara

Dung and elsewhere. That explorer more than once, in his book

Through Asia, remarks on the fact that “ the dry fine sand of the desert

unquestionably possesses the property of preserving organic matter

for a very long time ; see pp. 540, 802, 816 If. There is, therefore,

nothing intrinsically improbable in the claim of the manuscripts and

xylographs, contained in the British Collection, to be of a very great

antiquity.

One of the places were antiques have been found, Qara Qol Mazar,

near Guma, is described as “ an immense

graveyard in ruins, possibly ten miles long,”

and there is also a Mazar or Muhammadan

shrine22 there. It is possible that this place may be the site of one of

those great Muhammadan defeats which took place at the end of the

11th and beginning of the 12th century A.D. At that time the fierce

struggle for mastery took place between the Muhammadans of Kasbgliar

and the Buddhists of Khotan, which finally established the Muhamma¬

dan Faith in Eastern Turkestan. Dr. Bellew, in the Report of a Mission

to Yarkand, describes several “ vast cemeteries ” in the sandy desert

marking the sites of the slaughter of Muhammadan warriors. One is at

Ordain Padshah, about 30 miles east of Yangl Hisar, where there is a

Ancient graveyards

and stupas.

22 A mazar is a shrine and place of pilgrimage, consisting of the tomb of some

holy person with a kind of mosque built near it.

1899.]

Introduction.

XXIX

shrine of ‘All Arslan IQian and the graves of his 300 fellow martyrs.

Another is about three miles south-west of that town at Chucham Padshah

where there is a “vast cemetery consecrated to the dust of 10,000

warrior martyrs.” 23 Possibly Qara Qol Mazar may be a similar ancient

Muhammadan cemetery ; but all the available indications rather point

to the sites of the finds being ancient Buddhist “ graveyards.” Kok

Gumbaz, where the skull with the pillow of manuscripts was dug out, is

said to be “seemingly an old graveyard.” Qara Yantaq, clearly, is a

similar place, where also a skull with a bag of manuscripts was dug out

from the top of a circular mound. In Yabu Qum, the manuscripts,

M. 2, Set III, were found “mixed up with human bones, lying on some

partially exposed boards of a wooden coffin.” Mr. Hogberg says with

reference to the “ books,” purchased from him by Mr. Macartney and

comprised in M. 6, that he believes “ they were all discovered in the

sands or buried in coffins with the dead, in ancient graveyards in the

neighbourhood of Khotan, probably not more than a day or two’s

journey from the town.” One of these places, Aq Sapil, Mr. Hogberg

visited himself ; and the “ two elevated circular stands,” which he

describes as having been seen by him, curiously suggest themselves as

being the surviving bases of two stupas erected in the closest propin¬

quity: apparently twin-stupas built on slightly dififering levels. Bud¬

dhist stupas, as is well-known, used to stand on a series of circular, con¬

centric basements or terraces, decreasing upwards in diameter, the

basements thus forming steps to the uppermost platform, on which the

stupa or cylindrical dome itself was erected. These stupas were mostly

relic-towers, and the relics used to be placed in a small chamber made

in the top-most platform, immediately below the cylindrical dome This

exactly agrees with Mr. Hogberg’s description of the “ circular stands.”

The “ slight hollow ” on the top-most platform would be the remains

of the relic- chamber, from which, e.g., the skull with its bag of manu¬

scripts was dug out at Kok Gumbaz. Occasionally a Buddhist stupa

contained several deposits of relics placed at different levels, one above

the other. This would seem to have been the case at Qara Yantaq.

There, it is said, “the two images of horsemen (in M. 2, Set I) were

dug up from the interior of the mound,” on the top of which the skull

was discovered “ partially buried.” The skull, clearly, had been placed

in a chamber, near the surface of the top-most terrace, while the

horsemen had been deposited at a lower level, perhaps near the surface

of a lower terrace. It wrould seem that in all the cases reported, the

stupa proper or the cylindrical dome, has disappeared, the circular

25 See tlie Report of a Mission to Yarkand in 1873, at pp. 37, 129.

xxx Dr. Hoernle — Antiquities from Central Asia . [Extra Ko. 1,

basements only remaining.211 This cannot surprise, seeing that these

erections were made of unburnt bricks. As Dr. Sven Hedin remarks

(p. 740 of his book Through Asia ) “the natives themselves have

obseryed that the erosive action of the wind is incomparably greater

than that of water.” Buddhist stupas used to be coated with a hard,

brilliant plaster, to protect them against the erosive action of wind and

weather. This was, no doubt, also the case in Kliotan ; but when

outlying settlements were abandoned, and the plastering of the stupas

fell into disrepair, their more exposed domes naturally were corroded

and gradually swept away by the periodical sandstorms, the less exposed

and stronger basements only surviving.

It is well-known that Buddhism was introduced into Khotan from

^ North-Western India (Kashmir), including

Ancient Grseco- Bud- , . , n ,, , . ,. .

Afghanistan and the countries immediately

dhist Culture. .. c ., T ... ...

- north or it. in connection with this circum¬

stance it is curious to observe numerous points of coincidence in the

stupas of Khotan and those of Afghanistan; and these coincidences

themselves are a further argument to support the theory that the find-

places of antiquities around Khotan are the sites of groups of stupas or

tumuli, and, in that sense, of ancient places of sepulture. It was a

common practice among the Buddhists to build a stupa, or memorial

tower, over the relics of a Saint, and to group round it minor stupas

or tumuli of lesser personages, whether religious or secular. Instances

of this practice are repeatedly noted by Hiuen Tsiang in the account

of his visit to India.25 The existence of numerous such groups of

stupas and tumuli in Afghanistan is well-known. Many of them have

been opened at different times. In Wilson’s Ariana Antigua , there

is a long memoir by Masson on the “ Topes and Tumuli ” opened

by him, and the relics found in them. Among them are ornamented

funeral jars of a globular form with bones, ashes, and fragments of

charcoal ; further coins, beads, rings, seals and other trinkets, coloured

stones, pieces of crystal, etc., — all being objects which we shall see

represented in the Khotanese collection : some indeed having the very

same form. More curious still, in one tumulus which Masson opened,

belonging to the group at Passani, he found “ in the centre a human

24 it would be interesting to know why Koh Gumbaz or the “ blue or green

dome” is called so. Could it be the dome of a stupa still standing? In the Swat

country, as Dr. Stein informs us in his Report of an Archgeolog ical Tour with the

Buner Field Force, pp. 11 and 66, the word gumbaz is uniformly applied to ruined

stupas.

26 See Beal’s Buddhist Records of the Western World , Vol. J, pp. 46, 175 et

passim .

1899.]

Introduction.

xxxi

skull, and beneath it a large steatite vase,” containing ashes, coloured

stones, beads, etc., also a fragment of a birch-bark leaf inscribed with

“ Bactro-Pali ” (i.e., Kharosthi) characters.26 The similarity of this find

with that of tbe skull at Qara Yantaq is very striking. There is a

passage in the account of the mission of the Chinese Buddhist Sung-

yun to India in 518 A.D., which seems to bear on the subject of such

sepultures. Speaking of tbe customs of Khotan, the account says :

“ they burn their dead, and, collecting the ashes, erect towers over

them. When the king dies, the}r do not burn his body, but enclose it

in a coffin and carry it far off and bury it in the desert. They found a

temple to his memory, and, at proper times, paj? religious service to his

manes.”27 This would seem to suggest, that Kok Gumbaz, Qara Yantaq

and similar spots are ancient sites of the sepulture of kings and chiefs

of Khotan. The discovery of the two minature figures of horsemen,

(M. 3, Set I) in the same grave with the skull tends to corroborate

this conjecture.

The existence of early Buddhist culture in Khotan is thus amply

borne out. Much more evidence on this point is afforded by the pottery

and terracotta figures, and will be found noticed in that portion of the

report which will deal with these objects. Here I will only note that

the occurrence of the numerous figures of monkeys and elephants

clearly points to an intimate connection of the culture of Khotan with

that of India; for these animals are not found in Khotan, while they

are indigenous in India. A very early connection of Khotan with India

and China is also established by the discovery of Indo-Chinese and

Indo-Scythian coins on the one hand, and coins of the Han Dynasty

on the other. But further there are distinct traces of Grecian

and Parthian influence. For the latter, it is true, there is only one

piece of pottery (in M. 2), which bears ornamentation of a distinctly

Parthian character. For Grecian influence such as prevailed on the

western borders of India, in the earliest centuries A.D. and B.C., there

is much more evidence. The style of Graeco-Buddhist ornamentation

and sculpture is well marked on many pieces of pottery and sculptured

stones. The syrinx, or musical instrument made of a series of graduated

reeds, on which monkeys are represented as playing, is distinctly

Greek or Grecian : that kind of instrument was not known in India or

the Orient. Altogether the treatment of the monkeys, in their varied

festive or amorous postures, curiously reminding one of Satyrs and

Fauns, is instinct with the ideas of Greek or Roman culture. The

Pegasus and Centaur, which are found represented on some seals, are

See Ariana Antiqua, p. 94.

27 See Beal’s Buddhist Records of the Western World , vol. I, p. lxxxvii.

XXX11

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

also distinctly Grecian. Mr. Macartney in liis Note on the find-places

(see p. xxii) says : “ Some of the relics which have been found near

Khotau are undoubtedly of Greek origin. In May, 1897, I was shown

by a Russian merchant in that town a coin with an inscription in Greek,

and three pieces of yellow crystals of an oval shape in which there

were beautiful carvings of the classical type.”

Local tradition with regard to the makers of the habitations and

sepultures around Khotan is very uncertain, and possibly not altogether

spontaneous. With reference to Qara Yantaq, in Islam Akhun’s iti¬

nerary, already quoted above, Mr. Macartney was informed, that it was

once inhabited by “ Hindis,” a name by which Buddhists are said to be

generally called in Eastern Turkestan, and which clearly points to the

direct Indian origin of the Buddhism of that country. With regard to

Kok Gumbaz Captain Godfrey reports (in his demi-official letter,

No. 5208, dated 15th September, 1897 ) that “ local opinion seems to

incline to the belief that the cemetery was either Jew, Kalmuk or

Greek. The people to whom these graveyards are attributed are called

in the Turki language Ujat which I believe now means “strangers.”

This word is, however, I am informed, now obsolete. Dr. Bellew, in

his History of Kashghar, says that Ujat means Native Christians, and

refers I think, to Native Christians having lived near Khotan.”23 With

reference to the last observation of Captain Godfrey’s I may note that at

Aq Sapil some sheets and leaves of manuscript were found (in M. 2,

Set II) inscribed with characters in white ink, which seem to be

Uighur writing such as was once used by the Nestorian Christians. *

Considering how much we are at present dependent on native infor-

,, mation with regard to every thing connected

Need of further ex- °

with these sand- buried sites near Khotan, and

ploration. . ,. , , . £

r how cautiously such information must be re¬

ceived, it is very desirable that the localities should be visited, examined

and reported on by some European explorer with archaeological expe¬

rience. This is an undertaking well worth the support of the Indian

Government and of Learned Societies.

28 See the Report of a Mission to Yarkand , p. 127. Mr. Shaw in his Grammar

of the Language of Eastern Turkistan (in the Journal, As. Soc. Bengal, Yol XLYI for

1877), pp. 336 and 345, disputes this and says that Ujat is the name of a village

near Khotan, the inhabitants of which were “ bad Musalmans.” But the passage

from the Taskiratu-l-Bughra, which he quotes, really only proves that the people

of Ujat were considered insincere Muhammadans at the time of its composition.

At the time to which Dr. Bellew refers Khotan had not yet been converted to

Islam. It was still Buddhist ; and the people of Ujat, if they were not Buddhists,

must have been Nestorian Christians. Probably they were the latter, and being

forced to adopt Islam, did so only in outward appearance.

CENTRAL ASIAN ANTIQUITIES.

The Central Asian Antiquities which are comprised in the British

Collection distribute themselves into the following main classes : ( J )

manuscripts; (2) xylographs; (3) terra-cottas and pottery ; (4) coins

and seals; and (5) figures of stone, metal or wood, and other miscel¬

laneous objects.

As the coins present the most serviceable historical and chronolo¬

gical data, it may be best to commence with their description.

Section I.— COINS AND SEALS.

The following is a Summary of the Coins in the Collection : —

Total Coins 486

I. Indo-Chinese Coins.

There are altogether seventy-two of these coins in the Collection :

nine large and sixty -three small ones. They all come from Khotan and

J. i. 5

2

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

its neighbourhood ; and they formed part of the consignments M. 2,

M. 3, M. 6, M. 9, G. 5, G. 7, G. 10, and T. 1.

Two coins of this description, one large and one small, were first

published by Mr. Gardner in the Numismatic Chronicle , Yol. XIX

(1879), pp. 275, 276. These likewise were procured from Khotan by Sir

T. D. Forsyth. They have been republished by Dr. Terrien de

Lacouperie in the British Museum Catalogue of Chinese Coins, p. 394.

The large one is also republished in the British Museum Catalogue

of Indian (Greek and Scythic) Coins, p. 172. Both coins, especially

the small one, were in too imperfect a condition to admit of being fully

read. In the present collection there are some much better preserved

specimens.

All these coins are of copper. They are not of iron, as was at first

erroneously supposed.

Of the large coins, there are three varieties, distinguished by the

arrangement of the obverse legend. Of these varieties there are one,

three and one specimen respectively. Four specimens cannot be deter¬

mined. Of the smaller coins there are five varieties, distinguished by

differences in the reverse design, and in the arrangement of the legends.

Of these five varieties there are 17, 13, 3, 3 and 3 specimens respec¬

tively. Twenty-three specimens are too worn or corroded to admit of

being determined.

The following is a detailed list of all the coins, large and small,

with their weights and measures. Their exact find-place has also been

noted, when known : in the other cases it must be understood that the

coin came either from Khotan itself or from one of the buried sites

near it : —

1899,] Section I. — Coins and Seals. 3

( b ) Small Coins.

4

Dr. Hoevnle^- Antiquities from Ce?itral Asia. [Extra No. 1,

Total weight : 30150 Average weight : 47*857 grs.

The following is a description of the coins : —

(a) Large Coins. (Plate I, No. 6).

Obverse ; Two concentric circles, of which the outer one consists

of an ornamental band. In the small area

N o. 3 . *

within the inner circle is placed an old

form (a) of the Chinese symbol (6) for

‘money/ See Woodcut No. 3. Between

the area and the ornamental band runs a

Chinese legend, consisting of six symbols. This legend

is arranged in three different ways, making three

varieties, on which see below, p. 8.

Reverse : Two concentric linear circles ; in central area, bare horse

with stiff, upstanding mane, trotting to right. Between

the circles, an inscription in Kharosthi characters.

No. 1 of the list, shown in PI. I, 6, is nearly identical with that

figured in the British Museum Catalogue, p. 394, but the Chinese

legend, partially read by Dr. T. de Lacouperie, is far more legible.

(&) Small Coins. (Plates I, 8-18 and III, 1-4).

Obverse : Chinese legend of three symbols, in old forms; the same

on all five varieties.

1899.]

Section 1. — Coins and Seals .

5

Reverse : In first and second varieties, bare horse, standing or

walking to right ; round it a circular marginal legend

in Kharosthi characters, showing in the first variety

the letter ma (of maharaja ), in the second variety, the

letter ti (of uthahirdja ), over neck of horse.

The third variety has a Bactrian two-humped camel standing

to right, and the same Kharosthi legend as on the large coins, with ma

over head of camel.

The fourth variety has the bare horse, walking to right, within a

circular linear area, outside which is the Kharosthi legend, with maha

opposite the tail of the horse, but very incomplete.

The fifth variety has a camel walking to right, led by a man,

surrounded by a marginal legend in Kharosthi, with ma over the head

of the camel. Unfortunately both figure and legend in all three

specimens are too badly preserved to admit of being fully deciphered.

(c) The Kharosthi Legend.

The Kharosthi legend occurs in two different versions : a longer

and a shorter one. The former which consisted probably of 20 letters

is found on the large coins and on the small coins of the third (or

camel) variety. The shorter legend, comprising probably 13 letters,

is found on the small coins of the first, second and fourth (or horse)

varieties. The length of the two legends can easily be calculated from

the space distributable to the preserved and lost portions respectively.

What the legend on the small coins of the fifth variety may have been,

it is, at present, impossible to say.

The snorter legend is arranged in three different ways : in the

coins of the first variety it commences over the neck of the horse, and

in those of the fourth, behind its tail, while in those of the second

variety, it probably commenced below its feet. It is fullest preserved on

the coins Nos. 1 and 4 of the first and No. 7 of the second variety.

The best readings on coins of the first variety are the following : —

No. 1 maharajutha(bi) x ja Gugrama(d)asa (PI. I, 9).

No. 2 maharayu x X X X Gugratidasa.

No. 3 maharayu x X x x Gugra(mad)asa (PI. I, 13).

No. 4 maharayutha X ra x Gugramodasa (PI. I, 10).

No. 7 mahara(ja) x X X X ( Gu)gramo(da)sa .

No. 10 maha x x x x X X Gugradamasa (PI. I, 8).

No. 14 maha(ra) X X X X X Gugra(dama)sa (PI. I, 18).

The best readings on coins of the second variety are the following : —

No. 4 x Xrajuthubi X X XX XXX

No. 5 x xrajo x hi x x ( Gugra ) x x X (PI. I, 14).

6

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

No. 6 x x rajuthabira x x x x x x (PI. I, 15).

No. 7 x x X juthabiraja Gu(gra)x x x (PI. I, 12).

No. 8 x x xjuthubi x x x x x x x

No. 11 x x xjuthabi x x x x x x x

No. 12 x x xjuthabix x x x x x x

No. 13 x x x x x x raja Gugra x X X

Of the coins of the fourth variety, any tiling of the legend is only

visible on No. 2, where the following fragment can be read :

maharaja xxxx xxxxx

The longer legend is found on the coins of the third variety. On

these the letters are written in long, narrow shapes, closely crowded

together ; and calculating from what of the legend is preserved, it may be

seen that the face of the coin affords room for 20 letters. On the coins

of the second variety, the letters are formed large and square, taking

up much more space, so that the surface of the coin, to judge from what

remains of the legend, cannot have admitted more than 13 letters.

The same is the case with the coius of the first variety, where the

letters are formed small and square, but are set wider apart from one

another than on the coins of the third variety. On the two coins of

the latter variety, the legend reads as follows :

No. lxxxxx xxxxxx xx ta(sa) Gugra(da)max

No. 2 maharajasa ra(jati) x ( jasa ) (raa) x x (sa) Gugramadasa .

(PI. I, 16).

The same long legend, as already observed, occurs on the large

coins, with a slightly different form of the name, viz., Gugramayasa.

It reads as follows : —

(PI. T, 6).

No. 1, ( ma)harajasa ( ra ) xxxxx xxxx ( Gugramayasa )

No. 3, ( mahara ) XX XXXXXX xxxx (Gu)grax Xsa.

The bracketed letters are distinguishable ;l the others are perfectly

clear. The whole of the visible letters (eleven) occupy slightly more

than one-half of the circle; hence the total inscription must have

comprised about 20 letters.

Accordingly the complete legends, probably, stood as follows :

(1) longer legend : Maharajasa Rajatirajasa Mahatasa Gugramayasa

(or Gugramadasa or Gugradamasa ),

(2) shorter legend: Maharaj-uthabiraj a- Gugramadasa (or Gugrad-

masa or Gugraviodasa or Gugratidasa) . With variants yuthabi and

juthubi.

1 Some of these bracketed letters do not show sufficiently on the photographs

though they are quite distinguishable on the original coin.

1899.]

Section I. — Goins and Seals.

7

The letters which I read juthabi or yuthabi (or jathubi) are puzzl¬

ing. Their forms, as seen on some of the coins, are clear enough and

are shown in the subjoined Woodcut No. 4. Thus (a) is seen on

& b o cL &

Var. II, No. li, (6) on Var. II, No. 12, (c) on Var. II, No. 4, (d) on

Var. I, No. 1, (e) on Yar. I, No. 4, and (f) on Yar. II, No. 8. Of these

(a) signifies juthabi , (b, c, d) signify jutha, (e) signifies yutha , and (/)

signifies juthu. The form of the syllable bi never varies. In (a) and

(/) the vowel u is formed in an unusual way, but similar to its forma¬

tion in ( d ) of Woodcut No. 5, below. I would venture to offer the

following explanation, which must be understood to be altogether tenta¬

tive only. I would suggest that the legend might be the equivalent of

the not uncommon title Sanskrit Prthvi-rdja or Pali-Prakrit Putluivl-

rdja or Puthuvi-raja , i.e., 1 King of the earth.’2 The complete title on

the coins, accordingly, would run Sanskrit Mahdrdja-prthvirdja , or

Pali-Prakrit mahdrdja-piUhavlrdja or malidrdya-puthavlraja. In Prakrit,

as is well-known, the initial consonant of a conjunct word may be

elided, and the resultant hiatus-vowels may be contracted : in the

present case 0apu° may be changed to °aii°, and contracted to °o° or even

to °u°. We thus obtain the form of the title mahardj-uthablraja or

maharay-uthablraja , with the provincialism of hardening v. This

explanation postulates a somewhat advanced stage of Prakrit phonetic

change ; but the existence of such a stage in Khotan at the period of

these coins is rendered probable by the change of j to y in the form

maharaya.

2 I was disposed at one time to find some confirmation of my snggestion in the

Chinese Pi-qi-pi-lien, which, according to Abel Remusat’s Histoire de la Ville de

Khotan, p. 30, was the royal title of Khotan, and which I thought might represent

the Sanskrit Viqva-raya (for Viqva-raja) or ‘king of the world,’ a synonym of

Prthviraja. The context in Remusat seemed to imply that Pi-qi-pi-lien was the

title of the Khotanese Kings from ancient times up to the beginning of the 7th

century A.D., when the ’ We'i-si family ( ibidem , p. 35) succeeded the Wang family.

But from what Prof. Sylvain Levi kindly writes me (15th February, 1899) it appears

that Pi-qi-pi-lien was only the proper name of a particular king of the Wang family

which reigned in the 6th and 7th centuries, A.D. Pi-gi-pi-lien, accordingly, is more

likely to be the Chinese transliteration of some Turki name, similar to Mekelien.

8

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

The two first letters gugra of the name appear in the following

forms : —

NO. 5.

cL o cf

X

$

Perhaps the group might also be read gurga. The form (a) is the

commonest ; it occurs in Nos. 2, 3, 4 of the first variety, and can be

seen very distinctly in No. 3 (PI. I, 13) ; it is also seen in the large

coin No. 1 (PI. I, 6). The form (6) occurs in No. 1, (c) in No. 10, and

(e) in No. 14, all of the first variety. The form (/) occurs in the third

variety, and the absence of the conjunct marks at the foot of the two

letters is accounted for by the crowded state of the legend.

The final letters dasa appear in a curiously conjunct form in the

coins Nos. 2 and 4 of the first variety. They are shown as ( g ) in the

above Woodcut No. 5. The conjunction is probably merely due to the

negligence of the engraver.

There are altogether five varieties of the royal name, all commencing

with Gugra ; viz., Gugramada , Gugradama , Gugramaija , Gugramoda , and

Gugratida. Perhaps n may be read for d ( Gugramana , etc.), in every

case, or in some of them, seeing that the Kharosthi d and n are hardly

distinguishable. Seeing also that sometimes y occurs for in the title

maharaja (rnahardya) , it may be that, by a similar phonetic process,

Gugramaya is only another form of Gugramada. It is also possible

that Gugramoda is really intended for Gugramada , as what looks like

the vowel o may be a mere slip of the engraver. In any case there

still remain three names which cannot be identified with one another :

Gugramada , Gugradama and Gugratida. Accordingly these coins must

be ascribed to three, if not five different kings. As all their names

begin with Gugra (perhaps Gurga), they would all seem to have belonged

to the same family.

( d ) The Chinese Legend.

The Chinese legend, also, occurs in two different versions ; a longer

and a shorter one. The longer, consisting of six symbols, is found on

the large coins, while the shorter, consisting only of three symbols,

is seen on the smaller coins.

The longer legend is arranged in four different ways, three of which

occur in our collection. In the first variety, the legend commences op-

1899.]

Section I. — Coins and Seals.

9

posite tlie apex of the central symbol (seen at the bottom of the figure

in Plate I, 6) and then runs round from right to left. In the second

variety it also commences opposite the apex, but runs in the opposite

direction, from the left to the right. In the third variety it com¬

mences on the left of the central symbol, and runs round from the left

to the right. The British Museum Catalogue, No. 1799a, presents a

fourth variety, in which the legend runs from the right to the left, and

commences on the right side of the central symbol.

In all four varieties the legend is identical, as shown in the

subjoined Woodcut No. 6 : —

NO. 6.

1 2 3 4 5 6

A portion of this legend was read by Dr. T. de Lncouperie, in the

British Museum Catalogue, p. 394. I read the whole as follows : —

tchung (1) Hang (2) sze (3) tchn (4) thing (5) tsien (6), i.e ., “ Weight

(one) Liang (and) four Tchu (of) copper money.”

The symbol which Dr. T. de Lacouperie reads yh ‘one’ does not

occur in any of the coins of our collection, nor can I find it on the

coin figured by him in the Catalogue, No. 1799<x. The 5th and 6th

symbols were too indistinct ou his coin to be read by him. They are

clear enough on some of our coins, and are those shown in the above

Woodcut. No. 6 is the well-known sign for tsien or ‘ money ’ ( British

Museum Catalogue , p. xviii). No. 5 is a sign which I have not been

able to find in Morrison’s dictionary,3 the only one available to me; nor

is it known to any of the Chinese Literati whom I could consult. I

take it to be an old form of the symbol ^ t’ung ‘ copper ’ (see ibid.,

p. lxiv), made by omitting the long side-strokes of the upper

quadrangle of its right-hand portion. A similar modification

occurs in the old form of the symbol kuan (see ibid., p. 191),

and in the old form A A of the symbol |^| Hang (see ibid., p. 300).

The shorter legend is also identical on all the small coins, though

the symbols are drawn in rather varying forms. This is not at all

an uncommon practice, as an inspection of the British Museum Catalogue

will at once show. The legend, with the varying forms of its symbols,

8 A Dictionary of the Chinese Language in three Parts. By R. Morrison, D.D.,

1820.

J. i. 6

10

[Extra No. 1,

Dr. Hoernle — Antiquities from Central Asia.

is shown in the subjoined Woodcut No. 7, which also shows the relative

position of the three symbols in the legend.

No. 7. i •

No. I is the usual form, seen in Plate I, 11. Nos. II, III, IV may

be seen in Plate I, 8, 14, 16 respectively. No. V shows a form of the

3rd symbol which I have noticed on coins of the 3rd and 4th varieties,

shown in Plate III, 1, 3.

I read the symbols as follows : —

luh (l) tchu (2) tsien (3), i.e., “ six Tchu (of) money.”

The second and third symbols of this legend are the same as the

fourth and sixth of the longer one. The first symbol, as shown in Figure

III, is that given by Dr. T. de Lacouperie, on page xl of his Introduction

to the British Museum Catalogue, for luh 1 six.’ The corresponding

forms in fig. I, II and IV are merely ornamental modifications. A form

of lull, much like that in fig. II and IV, occurs in coin No. 453, of the Br.

Mus. Cat., p. 423. Compare also the forms of luh in coins No. 753, 816,

159-161.

The Chinese legends state the weight of the coins. According to

them the large coins should normally weigh one Hang and four tchu ,

while the small ooins should weigh six tchu. As we shall see presently,

these Indo-Chinese coins must be referred to the first and second

centuries A.D., i.e ., to the time of the Han dynasty in China. That

dynasty followed the monetary system of the preceding Tsin dynasty

which had doubled the ancient standard. According to this doubled

standard the Hang weighed about 195 grains, and the tchu, about 8T2

grains.4 Accordingly the normal weight of the large coins should be

approximately 227 ‘48 grains, and of the small coins, 48‘72 grains. A

reference to the preceding list (see pp. 2-4) will show that the actual

weights of the coins vary widely from this normal, even fully allowing

for much wear and tear. This, however, was the usual condition of

the currency in China. Dr. T. de Lacouperie in his Introduction to

the British Museum Catalogue (pp. xxiii, xxiv) shows how numerous

the variants in weight were, and how “ far they were from being

4 See Introduction to the Br. Mus. Cat., pp. xlii-xliv,

1899.]

Section J. — Coins and Seals.

11

approximate to tlie current standard.” The variations of the actual

from the normal weight appear to have been particularly great under

the Han dynasty, for the intermediate usurper Siu Wang Mang (6-25

A.D.) “ begau by annulling the decrees enacted by the Han dynasty,

as he wanted to return to the money of the Tcliou dynasty, where

‘ the mother and the child * (i.e., divisionary piece) weighed in pro¬

portion to each other, similarly to those issued by king Wang in

523 B.C.”6 In order to see how far the Indo-Chinese coins conform to

the normal weight, we must test them by their average weight. Judged

by this test they, curiously enough, very nearly agree with what should

be their normal weight. For the average weight of the nine large

coins is 213*44 grains (normal 227*48), 6 and of the 63 small coins, 47*857

(normal 48*72). The agreement in the case of the large coins would

probably be still greater, if we had a larger number of them to make

up the average.

The date of these Indo-Chinese coins can be approximately deter¬

mined by the following considerations. The fact of their superscriptions

being in Indian and Chinese characters and language shows that both

those languages must have occupied a recognised position in Khotan at

the time when the coins passed current. In the case of the bilingual

Indo-Greek coins, Indian was the language of the population of the

country, while Greek was the language of the administration or the

ruling power. Khotan, so far as known to us, never had a Chinese

population ; but it fell under the power of China at a very early dale.

In the sixth year of the Emperor Ming-ti of the Later Han dynasty,

in 73 A.D., Kuang-te, the king of Khotan, submitted to the Chinese

General Pantchao. Thenceforward the kingdom of Khotan became a

regular dependency of China, which formed that kingdom, together with

Kashghar and other Central Asian principalities, into an administrative

unit under the name of the “ Western Countries ” and under a Chinese

Governor-General,7 and placed Chinese Governors in Khotan and the

other chief towns. Shortly afterwards, King Kanishka of India (about

78-106 A.D.) is said to have held hostages from the Chinese “ tributary

Princes to the west of the Yellow River,” that is, from the princes

6 See p. 365 of the Brit. Mus. Catalogue.

6 The weight of the Brit. Mus. specimen, figured on p. 394 of the Catalogue,

would seem to be 220 grains ; for on p. xliii of the Introduction it is said “ the

Bactro-Chinese coin of 2 Hang 4 fchu = 220 grs.” There is here some confusion.

The weight inscribed on the coin is 1 Hang 4 tchu of the Han standard, which is

equal to 2 Hang and 8 tchu of the old standard ; and both alike are equal to 227*48

grains (normal).

7 See Abel Remusat’s Histoire de la Ville de Khotan, p. 3 and passim.

12 Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

included in the Chinese “Governor-Generalship” of the “Western

Countries.”8 It is true that there had been some political intercourse

between China and Khotan since the days of the Emperor Wuti

(140-87 B.C.) of the Earlier Han dynasty, but Khotan only lost

its independence in 73 A D., when it was included in the Chinese

“ Governor-Generalship” of the Western Countries. The Chinese cur¬

rency of Khotan cannot be placed earlier than that year. The native

kings continued to reign under the Chinese supremacy, and this fact

explains why the coins bear bilingual legends. It is distinctly a Chinese

currency, because the standard of the coins is Chinese, inscribed in

Chinese language and characters, and this fact clearly indicates Chinese

supremacy. On the other hand, the reverse of the coins bears the

symbols and names of the native kings, in native (Indian) characters, —

a fact which indicates both that native kings still continued to reign,

and that the language and characters, used by the native administration,

were Indian.

The first connection of India with Khotnn dntes back to the

time of king A^oka (264-233 B.C.). Ancient Khotanese chronicles,

quoted by Chinese writers, relate that the eldest son of that king, when

dwelling in Taksa^la in the Panjab, had had his eyes put out, and the

tribal chief who had been guilty of the outrage was banished, together

with his tribe, across the Himalayas. There the tribe settled and

later on chose a king from among themselves. Soon afterwards they

came into collision with another tribe settled to the east of them,

whose king had been expelled from his own country. In the result,

the western or Indian tribe was conquered, and the eastern king,

now uniting both tribes under his rule, established his capital in

the middle of the country, at Khotan.9 This must have been about

240 B.C. The eastern tribe would seem to have been the Uighurs,

of the Turk! race. They gradually occupied the whole of Eastern

Turkestan before 200 B.C., being pushed forward from the north¬

east by the Hiungnu or Huns, another Turk! tribe. The latter, in

their westward movement, displaced two Turk! tribes, the Yuechi (or

Yueti) and the Uighur ; the former migrated to the north, the latter

to the south of the Tian Shan mountains, displacing in their turn the

Saka tribe which had formerly dwelled there. The Yuechi were

gradually driven across the Ili and the Yaxartes. From 163 to

126 B.C., they occupied the country between the latter river and the

8 See Beal’s Buddhist Records of the Western World, Vol. I, pp. 57 and 173 ; also

Numismatic Chronicle, Vol. IX (1889), p. 272.

9 See Abel Remusat’s Histoire de la Ville de Khotan, pp. 37, 38, and Beal’s

Buddhist Records of the Western World, Vol, H, p. 310.

1899.]

13

Section I. — Goins and Seals.

Oxus, and by 26 B.C. they bad extended their settlements beyond the

Hindukush into Afghanistan. Here they formed a great kingdom under

the two Kndphises and under Kanerkes and Hverkes from about 25

B.C. to 180 A.D. Their rule gradually comprised the whole of North-

Western India in addition to Eastern Afghanistan. On their coins

they used both the Greek and Indian-Klmrostlri characters : the former

they retained from their Greek predecessors whose official script it had

been ; the latter was the script of secular commerce of their Indian

subjects. Co-existing with these scripts there were in use also the

Indian-Bralimi characters, favoured by the religious and learned,

especially the Buddhists.

Concurrent with the great Yuechi kingdom there was in North-

Western India a smaller one of another Turk! race under the kings

Maues, Azes, and their successors, from about 50 B.C. to 80 A.D. It did

not extend beyond the Panjab, and the Turk! invaders who founded it,

must have entered India through Kashmir and over the Karakorum

passes from the direction of Khotan. Here, we have seen, the Uighur

tribe, which still continues to form the main stock of the population of

the whole of Eastern Turkestan,10 had gradually established itself in

the second century B.C., in constant warfare with the Hiungnus and

Sakas. It was no doubt the Uighurs who, similarly to the Yuechis

further west, pressed forward aud extended their rule into India in the

first century B.C. Here they became the neighbours aud rivals of the

Yuechis, and here also they became acquainted with Greek and Indian

culture; for, like the Yuechi Indian kings, the Uighur Indian kings

Maues, Azes and their successors have both Greek and Indian-Kharosthi

legends on their coins. The Uighur kingdom, which in the South, (in

India), had to contend with the Yuechi, and in the North, (in Eastern

Turkestan), with the Hiungnu, at last declined in power. In order to

secure the assistance of the Chinese empire, its Northern portion

submitted to China and consented to pass under its administration.

This happened, as we have seen, in 73 A.D.11 About the same time its

southern portion was annexed by the Yuechi king Kanishka, who

extended his rule over Kashmir up to the Karakorum (Tsung-ling)

range, and took hostages from the remainder of the Uighur kingdom.12

Under these altered conditions, the Uighur coinage in Khotan was

conformed to the Chinese standard, and its obverse legend, which had

hitherto been Greek, was replaced by a Chinese inscription. The

reverse legend, on the other hand, continued, as hitherto, to be expressed

1° See N. Elias’ Tarikh-i-Rashidi, p. 92.

11 See Abel Remusnt’s Histoire de la Ville de Khotan , pp. 3 ff.

See Beal’s Buddhist Records of the Western World, Yol. I, pp. 56, 57.

14 Dr. Hoernle — Antiquities from Central Asia. [Extra No. I,

in the official Indian language and Indian-Kharoatln characters. This

explains the use of the latter amongst a Turk! population, such as

that of Kliotan must have been. They were the language and script

of the Uighur Government, having originally been adopted in India, and

surviving in Kliotan after the Indian portion of the kingdom had been

lost. Similarly the use of the Indian-Uighur types of the bare horse

and the Bactrian camel were continued. These types are found on

the coins of Maues, Azes, and their successors ; 13 and indeed, they

rather point to Turkestan as their home-land.

That a species of Indian script was current in Khotan, is wrell

known from Chinese writers. The case is not quite so clear with respect

to the language of the country. Hiuen Tsiang (about 645 A.D. ) relates

that “ the written characters and the mode of forming their sentences

resemble the Indian model ; the forms of the letters differ somewhat ;

the differences, however, are slight. The spoken language also differs

from that of other countries.” 14 Another account says that “ they have

chronicles, and their characters, as well as their laws and their litera¬

ture, are imitated from those of the Hindus, with some slight altera¬

tions. This imitation has diminished their barbarism, and modified

their manners and their language which (latter) differs from that of

other people.”16 These statements clearly indicate that the Uighur

population of Khotan, originally totally unlettered and uncultured,

derived the whole of their ancient culture from India ; and this fact well

agrees with, and is well explained by, the ancient extension of Uighur

rule over North-Western India. At the same time, it is not probable

that the Chinese statements about the written characters refer to the

Indian-Kharostlii script. They rather indicate a modified form of Indian-

Bi ahmi. The Kharostlii, as seen on the Indo-Chinese coins, does not

merely “ resemble the Indian model,” but is identical with that once

current in North-Western India and Eastern Afghanistan. Hiuen

Tsiang was a Buddhist monk, and on his travels he resided in Buddhist

monasteries, and came in contact almost exclusively with Buddhist

culture. The Indian-Brahml was the home-script and the peculiar script

of Buddhism, and was carried by them wherever they went. It went

13 See British Mns. Cat., pp. 72, 89, 96, 112. On their coins, as well as on the

Indo-Chinese coins, the horse is standing or walking, and is turned to the right. The

horse occurs also on the coins of other kings (Enthydemus, Heliocles, Menander, etc.),

but it is turned to the left, or is prancing. So also the camel is found on Menander’s

coins, but it is turned to the left, while on the Indo-Chinese coins it stands to the

right.

1* See Beal’s Buddhist Records of the Western World, Yol. II, p. 309.

*6 See Abel Remusat’s Histoire de la Ville de Khotan, p. 37.

1899.]

Section I. — Coins and Seals.

15

with them, as we know from the Bower and Weber Manuscripts to

Kuchar, and it is equally probable that it went with them to Khotan.

The introduction of Buddhism into both these places may be traced back

to as early a time as the first or second centuries B.C. In both places,

as the Chinese note, the Indian Brahmi developed “ slight alterations,” 16

known to us in Kuchar as the peculiar Central- Asian Brahmi.17 Hiuen

Tsiang, in the passage above quoted seems to distinguish between the

spoken and the written language of Khotan. By the latter, which he

calls “ the mode of forming their sentences,” and which he says

“ resembles the Indian model,” I presume he means Sanskrit or Pali,

such as was used in Buddhist literature, and which can have been

known only to a very limited class of people, the Religious and Learned.

The “ spoken language,” which I take to have been that of the general

population, must have been the Uighur Turk!, and this as Hiuen

Tsiang says, differed “ from that of other countries,” i.e., China

and India. This view is confirmed by a remark of Sung-yun (518

A.D.) respecting Yarkand. Of this town he says, “their customs

and spoken language are like those of the people of Khotan, but the

written character in use is that of the Brahmans,” 18 i.e , the Indian

Brahmi. Moreover, Fabian (400 A.D.) reports expressly with regard

to the whole of Eastern Turkestan, that though the people speak

different Turk! (Hu) dialects, “the professed disciples of Buddha

among them all use Indian books and the Indian ( Sanskrit) language.” 19

None of these Chinese Buddhist pilgrims appears to have noticed the

existence of the Kharosthi script, whether in Khotan or in its Indian

home-land. The only script of the Semitic class which Hiuen Tsiang

noticed, he mentions in connection with the kingdom of Kesli,20 and this

script cannot have been the Kharosthi, though it may have been allied to

it. Possibly in their time, Kharosthi had practically ceased to exist.

In Khotan, at the time of the Indo-Chinese coins, it was evidently the

secular official script of the native Government, though not quite

exclusively so, as is shown by the Kharosthi manuscript found near

that town by M. Dutreil .de Rhins and containing a portion of the

Buddhist Dhammapada.21 It does not seem probable that, after the

18 With regard to Kuchar, see Hiuen Tsiang's remark, in Beal’s Buddhist Records

of the Western World , Vol. I, p. 19.

11 See a description of it in my Report, in the Journal , ds. Soc . Beng., Vol. LXVI

(1897), p. 242, LX1I, p. 4.

18 See Beal’s Buddhist Records of the Western World, Vol. I, p. lxxxix.

' See ibidem, Vol. I, p. xxiv.

See ibidem, Vol. I, p. 38.

21 See Comptes Rendus de L' Academic des Inscriptions, Vol. XXV, (1897), pp.

251 if.

16

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

severance of the Indian connection of the Uighur kingdom of Khotan,

the use of the official Kharosthi script survived for any great length

of time. Its forms, as seen in the Dutreil de Rhins Manuscript and on

the Indo-Chinese coins, are much alike, and both are identical with that

form of it which prevailed under the Kushana (Yuechi) kings in India,

that is, in the first and second centuries A.D. Though its form remained

practically unchanged for a century or two longer in its home-land,

it is very improbable, to judge from the parallel case of the lndian-

Brahmi, that this would have been the case in a foreign country like

Khotan. It is not probable, therefore, that the Indo-Chinese coins

can be placed later than the end of the second century A.D. They

show, as already remarked, four, if not five, different regal names.

Four or five reigns, at an average of 20 or 25 years, occupy a period of

about 100 years. This brings ns to, at least, the year 173 A.D., as none

of the coins can have been struck before 73 A.D. The initial date is

certain ; the terminal date must be near the end of the second

century. The period 73-200 A.D., therefore, is a safe date to give to

the Indo-Chinese coins of Khotan.

Within that period, the Chinese records mention the names of

four or five kings of Khotan: (1) Kuang-te in 73 A.D., who first

submitted his country to the overlordship of the Chinese; (2) Tang -

t'sian in 129-131 A.D., (3) Kian , (4) ’An-kue, son of Kian , who suc¬

ceeded his father in 152 A.D., and (5) Sliansie in 220-226 A.D.82 None

of these names agrees with any of those on the coins ; but they rather

look like true Chinese names, so that it would seem that the kings bore

duplicate names, native Turki and Chinese. At that early period, as

the Chinese relate, the kings of Khotan were devoted Buddhists, and as

such, it may be surmised that they bore names which were the Uighur

equivalents of Indian Buddhistic terms. Dharma being a common

prefix of many Buddhistic names, Gugra might be its Uighur equi¬

valent. A long list of ancient Khotanese royal names, all beginning with

Vijaya , is given by Rai Sarat Chandra Das from Tibetan sources.28 If

this list can be trusted, Gugra might represent Vijaya.

II. Chinese Coins.

My knowledge of Chinese is very small, and the only numismatic

aid, available to me, is the Catalogue of Chinese Coins in the British

Museum1 by Dr. Terrien de Lacouperie, and an article on Chinese Coin¬

age in the Transactions of the China Branch of the Royal Asiatic

22 See Abel Remnsat’s Histoire de la Ville de Khotan, pp. 3, 6, 8, 15, 17.

23 See Journal, Asiat. 8oc. Beng., Vol, VI (1886), pp. 197, 198.

1899.]

Section I. — Goins and Seals.

17

Society, Part II for 1848-1850, by C. B. Hillier. I hope that this

disadvantage under which I am labouring may be accepted as a suffi¬

cient excuse for the imperfections of my descriptions of the Chinese

coins in the collection.

TVith reference to numismatic evolution, Chinese round coins fall

into three periods. In the first period, they have no inscriptions what¬

ever. In the second period, they have a legend on the obverse, consist¬

ing at first of two symbols, placed to the right and left of the central

hole, and afterwards of four, there being two additional symbols above

and below the hole : the reverse is blank. In the third period, they

have inscriptions both on the obverse and reverse sides, generally

consisting of four symbols on each side, distributed on the four sides

of the central hole. The first advance to a double-sided inscription

consisted either in repeating the obverse legend on the reverse, or in

placing on it one new symbol, in most cases a numeral indicative of the

value of the coin. There are found occasional anticipations as well as

survivals ; but roughly speaking, the course of numismatic evolution

appears to have been as above explained.24* This is amply borne out by

the coins in our collection.

In poiut of chronological sequence the coins of the collection also

happen to fall into three distinct periods : ancient coins of the 1st to

the 3rd centuries A.D., mediaeval coins of the 7th to the 13th centuries,

mostly of the two dynasties of the T’ang (618-907 A.D.) and the Sung

(960-1279 A.D.), and modern coins of the 18th and 19th centuries, issues

of the ruling Manchu dynasty. The circumstance of the two large gaps

of several centuries each is curious, but perhaps altogether accidental.

The presence of the numerous coins of the T’ang and Sung dynasties

is probably accounted for by the fact that during the periods of their

rule, as shown by the Chinese records,26 an exceptionally lively inter¬

course was kept up between China and Khotan.

The total number of Chinese coins in the collection is 148. Among

these there are 43 ancient, 77 mediaeval, and 28 modern coins. All the

ancient and many of the mediaeval coins were found in the desert

around Khotan. The modern ones came from Khotan itself. They all

formed part of M. 2, 3, 4, 6, 9 j most of the ancient ones belong to M. 2.

24 See British Museum Catalogue, pp. xxvii and 319. Examples of repetition

are ibidem , Nos. 1727-1731, 1786-1790, 1877, 1880. Examples of the addition of

numerals are Nos. 1767-1778, 1807-1814, of other symbols, Nos. 1782, 1815, 1816,

1818, 1820, 429-436. Exceptional anticipations of a double-sided legend are Nos.

1752, 1753. Examples of survivals of a two-symbol legend are Nos. 426-438,

1852-1855.

86 See Abel Remusat’s Histoire de la Ville de Khotan , pp. 67 ff.

J. i. 7

18

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

All the ancient coins are of copper, except one which appears to he of

lead. All the mediaeval and modern coins are of a species of bronze or

brass.

(a) Ancient Coins.

(1) Coins without legends. (Plate II, figs. 1 a-d.)

These coins number 27. They are of three different sizes :

large, 11 specimens, size 0*7-0*9", weight 21*5-31 grs.

middle, 11 „ , „ 066-075", „ 11*5-15*5 grs.

small, 5 ,, , „ 0*5", „ 5-8 grs.

Two good specimens are shown in the British Museum Catalogue, large,

No. 180 (p. 340), and middle, No. 407 (p. 399), weighing 38 and 19

grains respectively. Most of the specimens in our collection are not in

an equally good condition. Their weight is much lighter, their shape

is very irregular (some nearly square), and their rim in many cases is

very narrow (down to Txy of an inch). They have clearly been sub¬

jected to much clipping.

This class of coins appears to have been current under both Han

dynasties, the Former or Western as well as the Later or Eastern.

Those of our collection must belong to the later period, i.e.t 25-220

A.D., as Khotau came into closer contact with China only from about

the middle of the first century A.D.

(2) Coins with an obverse legend of two symbols. (PI. II, 2, 3).

The coins of this class number 16. They consist of the following:

(а) With the legend Wu-Tcliu or ‘Five Tchus ’ ; 9 specimens;

five well preserved (PI. II, fig. 2), weighing 33-38 grains and measuring

1 inch; four considerably rubbed and clipped, weighing 15-25*5 grains

and measuring 075-1 '0 inches. Compare British Museum Catalogue ,

Nos. 315, 316, 398-403 (pp. 363, 396).

(б) With the legend Ho-tsiuen or ‘ Spring of goods ’ ; 2 specimens,

weighing 34 and 20 grains, and measuring 0875 and 0*8 inches; too

indifferently preserved to be figured, but like British Mus. Cat , No.

365 ff. (p. 334).

(c) With unread legend, see Plate II, fig. 3 ; apparently lead; one

specimen ; weight 78 5 grs. ; size 1*0625".

The Wu-tchu currency was introduced by the Han dynasty, and the

Ho-tsiuen currency, by the usurper Sin Wang Mang (9-22 A.D.). Both

currencies continued into the period of the Later Han dynasty, 25-220

A.D., and the specimens of our collection must be ascribed to that

period.

1899.]

Section I. — Goins and Seals .

19

(6) Mediaeval Goins.

(1) Coins with an obverse legend of four symbols.

(PL II, 4-18 and Pl. III, 6, 7.)

The coins of this class number 76. They consist of the following

currencies : —

(a) With the legend K’ai-yuen-tung-pao, or ‘ Current money of the

K’ai-yuen period.’ This period comprised the years 713-741 A.D.,

under the Emperor Yuen-tsung of the T’ang dynasty.26 There are two

coins of this period ; weight 49 and 50 grs. ; size 1". Plate II, 9.

(b) With the legend K’ien-yuen-tchimg-pao , or ‘ Current money of

the K’ien-yuen period.’ This period comprised the years 758-763 A.D.,

under the Emperor Su-tsung of the T’ang dynasty.27 Of this period

there is a very large number of coins in the collection ; altogether 45.

They are of three different sizes :

large, 12 specimens; size 1*0625" ; weight 7 L— 136 grs. (PL II, 7).

middle, 3 „ ; ,, 1"; „ 48*5-49 5 grs. (Pl. II, 6).

small, 30 „ ; „ 0’875" ; „ 23-41*5 grs. (Pl. II, 5).

Many of these coins were in M. 3 ; some in M. 2.

(c) With the legend Ta-li-yuen-pao , or 1 Principal money of the

Ta-li period.’ This period comprised the years 763-780 A.D., under

the Emperor Tai-tsung II of the T’ang dynasty. There are ten coins of

this period, of three different sizes : —

large, 4 specimens; size 0*9375"; weight 45*5-59*5 grs. (Pl. II, 4).

middle, 5 ,, ; „ 0 875" ; ,, 37-51 grs.

small, 1 ,, ; ,, 0*75" ; ,, 36 grs.

These belong to M. 2, M. 4, M. 6. A Chinese manuscript petition dated

in this period is in the collection of MSS.

(d) With the legend Ghe-tao-yuen-pao or ‘ Principal money of the

Che-tao period.’ This period apparently comprised the years 984-998

A.D., under the Emperor Tai-tsung (976-998 A.D.) of the Sung dynasty.23

There is one coin of this period; legeud in “running hand”; weight

58 grs,, size 1". Figured by Hillier, Ho. 124 (p. 63). From M. 2.

(e) With the legend King-t’i-yuen-pao, or ‘ Principal money of the

King-t’i period.’ This period apparently comprised the years 998-1008

A.D., under the Emperor Chin-tsung I (998-1023 A.D.) of the Sung

dynasty.29 There is one coin of this period ; weight 36 grs., size 0'9375".

Plate II, 16; and in Hillier, No. 126 (p. 63). From M. 2.

28 See Abel Remusat’s Histoire de la Ville de Khotan , p. 70.

21 See ibidem, p. 70.

28 Compare ibidem, p. 88. The preceding period was 976 983 A.D.

2* Compare ibidem , p. 86. The following period was 1008-1116 A.D.

20

Dr. Hoernle — Antiquities from Central Asia . [Ex tra No. 1,

(/) With the legend T'ien-ging-tung-paOy or ‘ Current money of the

T’ien-ping period.’ This period comprised the years 1023-1034 A.D.,

under the Emperor Tin-tsung of the Sung dynasty.30 There is one

coin of this period ; weight 55 grs. ; size 1". Plate IT, 18.

(g) With the legend Kia-yeu-tung-pao , or ‘ Current money of the

Kia-yeu period.’ This period comprised the years 1056-1064 A.D.,

under the Emperor Jin-tsung of the Sung dynasty.31 There are two

coins of this period ; weight 47 5 and 55 grs. ; size 1". Plate II, 15.

(h) With the legend Clie-ping-yuen-pao or ‘ Principal money of the

Che-ping period.’ This period comprised the years 1064-1068 A.D.,

under the Emperor Ying-tsung of the Sung dynasty.3* There is one

coin of this period ; weight 52*5 grs., size 1". Plate II, fig. 8 ; and in

Hillier No. 138 (p. 68). From M. 2.

(i) With the legend Yuen-fung-tung-pao or ‘ Current money of the

Yuen-fung period.’ This period comprised the years 1078-1085 A.D,,

under the Emperor Chin (Shin)-tsung II of the Sung dynasty.33 There

are two coins of this period ; one with the legend in “ running hand,”

the other, in ordinary script ; weights 35 and 325 grs., size 1” and

0-9375”. Plate II, 12, and Hillier No. 140, p. 71. From M. 2.

(Jc) With the legend Yuen-yu-tung-pao or ‘ Current money of the

Yuen-yu period.’ This period comprised the years 1086-1093 A.D.,

under the Emperor Che-tsung of the Sung dynasty.34 There are three

coins of this period ; two with the legend in “ running hand,” and one

with it in “ seal-characters.” Weight of the former, 60’5 and 39 grs. ;

size 0 9375” and 0-875”; Plate II, fig. 13. Weight of the latter, 68 5

grs. ; size 0 9375” ; in Hillier, the 2nd under No. 141 (p. 71). From M. 2

and M. 6.

{ [l ) With the legend Chao-ging-yuen-pao or ‘ Principal money of

the Chao-£ing period.’ This period comprised the years 1094-1097

A.D., under the Emperor Che-tsung of the Sung dynasty.35 There are

two coins of this period, one with the legend in “ running hand,” the other

with it in “ seal- characters.” Weight, 54 and 51 grs. ; size 1” and 0 9375”

respectively. Plate II, fig. 17 and fig. 14; in Hillier, No. 142 (p. 72).

From M. 2.

(m) With the legend Tsung-ning-tchung-pao or * Weight-money

of the Tsung-ning period.’ This period comprised the years 1101-

30 See ibidem , p. 90.

See ibidem, p. 91.

32 Compare ibidem , p. 91.

3S Compare ib., pp. 92, 95, 97. The preceding period was Hi-nivg 1068-1077 A.D.

84 See ibidem , p. 97.

S5 See ibidemf p, 98.

1899.]

Section I. — Couis and Seals.

21

1106 A.D., under the Emperor Hwei-tsung (1101-1126 A.D.) of the

Sung dynasty.36 There is one coin of this period; weight 176 grs. ;

size 1*416". Plate III, 7 (inverted), and Hillier No. 145 (p. 74).

Prom M. 9.

(?i) With the legend Kuang-ting-yuen-pao or c Principal money of

the Kuang-ting period.’ This period was current under Shin-tsung in

the State Hear, apparently in the 13th century ; see Hillier, No. 194

(p. 104). There is one coin of this period; weight 58'5 grs.; size 1".

Plate II, fig. 10. From M. 2.

(o) With the legend King -king -tun g-pao or ‘ Current money of the

King-hing period.’ The exact date of this period appears to be un¬

known ; see Hillier, No. 308 (p. 154) ; it should be somewhere in the

time of the T’ang or Sang dynasties. There is one coin of this period ;

weight 41*5 grns. ; size 0 9375". Plate III, No. 6. From M. 2.

(p) With unread legends. There are three coins of this kind,

which I cannot identify in Ilillier’s article. They all have the term

tung-pao which refers them to the time of the T’ang or Sung dynasties.

One of them is shown on Plate II, fig. 11. Weights 39*5, 51*5, 58 grs. ;

size 09375-1". From M. 2.

(2) Coins with obverse and reverse legends. (PL II, fig. 19).

There is only one coin of this class, which I have not been able to

identify in Hillier’s article. I read the obverse legend Li-yung-tung-pao

or ‘ Current money of the Li-yung period.’ The reverse has only one

symbol chen or ‘ a bazar.’ In Hillier’s article I find this symbol only on

the reverses of certain coins of Yung-ming-wang and Chang-hin-chung

(Nos. 254 and 259, pp. 136 and 139), who are said to be princes at the

close of the Ming dynasty, Le., 1644 A.D. Weight 55 grs., size 09375".

(c) Modern Coins .

The total number of modern coins is 28. They fall into the

following classes :

a) Coins with Chinese legends on both sides. Of these there

are altogether 24, of the following reigns : —

(a) With the obverse legend Kang-lii-tung-pao or ‘ Current money

of the Emperor Kang-hi,’ who reigned from 1661-1722 A.D. Of his

reign there are two coins, with the same reverse legend of two words

in Manchu characters pao tsiuen or ‘ source of money,’ i.e ., mint Peking.

Weight 70 and 50 grs.; size P0625" and 0 875". Plate II, 20.

(b) With the obverse legend K' ien-lung -tung-pao, or * Current

money of the Emperor K’ien-lung,’ who reigned 1735-1796 A.D. Of

3® Compare ibidem , p. 99. The following period was Tai-Kuen 1107-1111 A.D.

22 Dr. Hoernle — Antiquities from Central Asia . [Extra No. 1,

his reign there are 14 coins. The reverse has varying Manchu legends

of two words. Weight 44-70 gr s., size 0*875-1". Plate II, 21.

(c) With the obverse legend Hien-fung-tchung-pao , i.e ., ‘ Weight-

money of the Emperor Hien-fnng,’ or Hien-fung-yuen-pao , i.e., ‘ Princi¬

pal money of Hien-fung,’ who reigned from 1850-1861 A.D. Of his

reign there are eight coins. One is of bronze, very large, and has a tri¬

lingual reverse legend, in Chinese (above and below the hole) tung 100

or ‘value 100 cash,’ in Persian (to the right) or (mint) ‘Yar¬

kand,’ and in Manchu (to the left), see Plate II, 30. The other seven

coins are of brass, of two different sizes, and with a bi-lingual reverse

legend. The larger one (Plate III, 5) has in Chinese (above and

below) tung 15 or ‘ value 15 cash,’ the smaller ones have tung 10 or

‘ value 10 cash.’ In addition all seven coins have a Manchu legend

(right and left). (Plate II, 22).

Very large, 1 specimen ; size 2*0826" ; weight 576 grs.

large, 1 „ ; „ 1*75" ; „ 351 grs.

small, 6 ,, ; „ 1" ; ,, 76-118 grs.

(2) Coins with Persian legends on both sides. There is one coin

of this kind, of copper. Plate II, No. 23. It is made in the Chinese

fashion, but is probably a coin struck during one of the more recent

periods of Muhammadan independence of Kashghar. The legend is

not fully read.

Obverse . Reverse.

1

The reverse (over saltanat ) apparently bears the date 1283 H.,

nearly obliterated, which would be 1866-1867 A.D., or the second year

of Yaqub Beg’s revolt,37 and with which the number 2 on the obverse

would agree. Kuja, which is quite distinct, may be intended for kucha

(Kuchar), but the words in brackets are uncertain; and I am unable,

with the means at my command, to identify the ruler’s name recorded

on the obverse. Weight 48 grs. ; size 1 inch. From M. 2.

(3) Coins or Tokens with a Chinese legend on one side only.

There are three of these pieces which are perhaps rather tokens than

coins. I can obtain no information on them. They are shown on

Plate II, fig. 25. They first bears the symbol for tsien or 1000 ; the

two other symbols I do not know. They are of bronze, and weigh

89*5, 80*5 and 108*5 grs. ; size 0*83". From M. 2.

87 See Dr. Bellew in Sir T. D. Forsyth’s Report of a Mission to Yarkand in 1837,

pp. 208-213.

1899.]

Section I. — Goins and Seals.

23

III. Scytho-Bactrian.

With the exception of two doubtful specimens, probably all the

coins of this class, numbering 36, have come from Western Turkestan

(Samarkand, Tashkend, etc.). They belong to G. 4, and were briefly

reported on by me in my letter to the Under-Secretary, Government

of India, Department of Revenue and Agriculture, dated the 20th

November, 1897. The two exceptions belonged either to M. 2 or M. 6,

and come from one of the buried sites lying to the north of Khotan.

(a) Imitations of Bactrian Coins .

There are seven of these ; all silver tetradrachms. They imitate

the coins of Euthydemus and Heliocles. The former reigned in Bactria

about 210-190 B.C, ; the latter, who appears to have belonged to a rival

family, about 160-120 B.C. During the reign of the former, Saka

tribes occupied the Northern provinces of the Bactrian empire between

the Oxus and Yaxartes. During the reign of the latter, the Sakas,

being driven out by Kushan (or Yue-chi) tribes, occupied Bactria

south of the Oxus.38 Their chieftains imitated the coins of their con¬

temporary Bactrian rulers. These coins can be easily recognized by

their degradation, both in point of design and of weight.

The best of the seven coins are two in imitation of Heliocles, of

his well-known type : Bust of King on obverse, and Standing Zeus on

reverse, as in the British Museum Catalogue, plate vii, fig. 2. One, which

weighs 231 grains (full weight 264), measures 1*25", and is fairly good

in design (with ringlet for omikron), though much worn, may possibly

be a genuine coin of Heliocles. It has the monogram of Brit. Mus.

Cat., No. 4 (p. 21). The other weighs only 219 grains (size 1*25"),

and, as the semi-barbarous reverse shows, is clearly a Saka imitation :

but the curiosity of it is, that while it has an imitated Heliocles

reverse, it has retained an apparently genuine Eukratides obverse;

see Plate III, 10. Eukratides (c. 190-160 B.C.) was the predecessor,

and perhaps father, of Heliocles. The imitated Heliocles reverse is

very fairly done, it has the full Greek legend, but with a dot for

omikron, and a rather rude figure of Zeus. Its monogram is W Both

this and the first-mentioned coin must be early imitations, and may be

referred to about 150 B.C.

The remaining five coins are imitations of Euthydemus, of his

w7ell-known type with Head of King on obverse, and Sitting Heracle

83 See the outlines of Bactrian history in the Introduction to the British Museum

Catalogue, pp. xviii, ff.

24

Dr. Hoernle —Antiquities from Central Asia. [Extra No. 1,

on reverse, with club resting on his knee. One of them, which is the

heaviest, weighing 170 grains and measuring 1", has the king’s portrait

as shown in Brit. Mus. Cat., pi. ii, fig. 1-4. It had also an entirely Greek

legend, which, however, is almost totally obliterated. The other four

coins, which only weigh from 155 to 144 grains, show the king’s face

as portrayed in Brit. Mus. Cat., pi. i, fig. 11, (also Ariana Antiqua ,

pi. i, figs. 2-4, and Rapson’s Indian Coins , pi. i, fig. 18, in the Indo-

Aryan Encyclopedia). Both types of face, however, are very fairly

imitated, see Plate I, Nos. 2 and 3. One of the four coins, which weighs

144 grains (size 1"), had an entirely Greek legend, now badly effaced;

but sufficient traces remain to show that it had the name of Heliokles

struck over that of Euthydemus. The two names were not struck

accurately in the same line, consequently M (of Euthydemus) is still

seen slightly projecting over the line of Heliokles, of which latter

name K is fully, and A I partially recognizable ; as No. 8.

shown in the annexed woodcut; see Plate III, 11.

The other three coins are bilingual, having the j| ty T|P

king’s name in native Bactrian letters, while the

title in Greek characters is seen in its usual place to the right,

or behind the back, of the Sitting Heracles ; see Plate I, Nos. 2-4.

Of the Greek title BAXIAEHZ only the three letters 2lI> or ZIV

(i.e., with inverted A ) together with traces of A before and E after

them are clearly legible.39 Coins of this description, that is, with

the title to the right and the name in Bactrian letters to the

left of Heracles, appear to have been found previously. Two such

coins, from the collection of General Fox (if I understand the

account correctly) are described by Mr. Thomas in his edition of

Prinsep’s Indian Antiquities , vol. I, p. 32. But, so far as I know,

none of them has ever been figured. Similar coins, but with the

Greek and Bactrian legends transposed, that is, the title in Bactrian

and the name (Euthydemus) in Greek, have been published. One, in

rather good preservation, has been figured by Sir A. Cunningham in

the Numismatic Chronicle , vol. IX (1889), pi. xiii, (also in Rapson’s

Indian Coins , pi. i, 19). Another series of similar coins has the whole

legend in Bactrian characters, see Numismatic Chronicle, vol. IX.,

pi. xiii, 6, also Ariana Antiqua, pi. i, 9, 10, Indian Antiquities , pi. ii, 6.

It is probable that, as Sir A. Cunningham says ( Num . Chron., vol. IX,

p. 307), the oldest imitations are those with Greek legends only, next

come those with mixed legends of rude Greek and Bactrian letters,

the latest are those with Bactrian characters only. In the second class,

On the* photographic plates they are not so distinct as on the original.

1899.]

Section I. — Corns and Seals.

25

I suppose, those coins which preserve the Greek fashion of arranging

the legends, and show the title on the right in Greek, and the name on

the left in Bactrian, may be considered to be older than those which

show the mixed legends in the reversed position, i.e ., the name in Greek

on the left, and a Bactrian legend on the right, the latter also being

a name. Accordingly the bilingual coins of the present series may be

referred to about 130 B.C. It would also seem, if Dr. Gardner’s theory

of the change of standard is correct (see Brit. Mus. Cat., Introd.,

pp. lxvii, lxviii), that these coins are didrachms of the Persian standard

(full weight 160-170 grains), such as began to be minted in Heliocles’

reign.

Seeing that the Bactrian legend 4,0 on our coins takes the place of the

Greek name, it seems reasonable to assume that, like the latter, it runs

parallel to the Greek title and must be read from the outside of the

coin. This assumption is certainly supported by the general appearance

of the characters, which, after the Semitic fashion must be read from

the right to the left. They are shown in the subjoined woodcut.

No. 9.

Plate I, No. 2. W. 155 grs.

do. No. 3. W. 148 grs.

do. No. 4. W. 145 grs.

The third, fourth and fifth letters of No 1 legend have a distinct

resemblance to the Kharosthi letters ja , a and ha ; and at first I was

disposed to take the second letter as a crude Kharosthi ra , and to read

the whole as a mutilation of (ati)raja Aha(thuhleyasa). But the

In order to prevent any misunderstanding I may explain that I use the term

Bactrian in the definite sense of referring to Bactria proper, and the immediately

adjacent northern provinces of what was once the Bactrian Kingdom. What I

wish to suggest (the suggestion only to be taken for what it may be worth) is that

corresponding to the modified Aramaean script • current to the South of the Pnro-

pamisus and known as Kharosthi, there may have been another modified and allied

Aramaean script current to the north of that range, of which the letters on the

coins in question may be witnesses. This suggestion refers only to the script

whether the language hidden in the legends of the coins was a species of old

Turki or old Iranian is a point on which I hazard no opinion. For a similar

suggestion, if I understand it rightly, see Isaac Taylor’s The Alphabet, Vol II,

pp. 232, 233.

J. i. 8

26

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

remaining signs do not suggest Kharosthi letters. The fifth letter of

Nos. 2 and 3 suggests the Kharosthi e ; but on the whole the three

legends suggest themselves as identical ; for the first three letters in

all are clearly the same ; so are most probably the sixth and seventh ;

and the fifth letter of Nos. 2 and 3 may be only a badly drawn form

of the corresponding letter in No. 1. The only apparent difference

between the three legends is, that the fourth letter of No. 1 is wanting

in Nos. 2 and 3. I am not able to decipher the legend ; but considering

the juxtaposition with the other coins of Euthydemus and Enkratides

which bear the name of Heliocles, I would like to suggest that the

Bactrian legend might also contain that name. The alphabet current in

Bactria must have been one of the very early modifications of the

Aramaean, similar to the ancient Pahlavl and Kharosthi. The first and

fifth letters are very like the Pahlavi h and the Kharosthi k respectively.

The second letter resembles the Kharosthi l. The third and fourth

• •

letters resemble the Pahlavi aleph and vau respectively, and together

might have been used to express the vowel o. In Nos. 2 and 3 the

fourth character is omitted ; and the third might also be taken to

represent the Aramaean ‘ay in and to express the vowel o. Anyhow

the initial four or five characters may easily be interpreted to represent

h-l-o-k , the initial portion of the name Heliok(les). It is more difficult

to fit-in the remainder, unless we may assume that the name was

pronounced with r instead of l , as in its Indian form Heliyakreya. In

that case the sixth letter is r, in its form closely resembling the corres¬

ponding Pahlavi and Kharosthi characters. The seventh letter appears

to be mutilated, and there may have been an eighth ; but I do not know

what the genitive inflection of the local Bactrian or Scythian dialect may

have been in those days. Thus the characters may represent the letters

h-l-o-k-r , which would well enough make up the name of Heliokles.

m Coins of Hyrkodes.

There are twenty-six coins of Hyrkodes, about 110 B.O., silver obols ;

mostly of the two well-known types, with Head of King on obverse, and

either Standing Figure (17 specimens), or Head of Horse (7 spec.) on

reverse, as shown in Brit. Mus. Oat., pi. xxiv, 10 (10 spec.), ibidem ,

pi. xxiv, 11 (7 spec.), and ibid., pi. xxiv, 12 (7 spec.). But there are

two obols, one being a new variety of the well-known type, the other

an entirely new type. The new variety (see Plate II l, No. 8) shows on

the reverse the Standing Figure holding a spear in his left hand, while

the usual variety shows the spear in his right hand. Weight 13 grs. ;

size 0‘5". The new type (see Plate III, No. 9) shows the usual Head

of King on the obverse, but the reverse has a standing figure to the

1899.] Section I. — Coins and Seals. 27

right, apparently Nike standing on a scroll (cloud P) with traces of a

Greek legend. The King’s head is distinctive for this coin. Size

0*5625". Weight 17 grs.

(c) Coin of Azes.

There is one coin of Azes, c. 30 B.C., silver; nearly the entire

legends of both sides clipped away ; of the well-known type with

mounted King on obverse, and Zeus holding Nike on reverse ; ap¬

parently in every respect (inch of monograms) the same as Brit. Mus.

Oat., No. 32, p. 75. Weight 36 grs., size 0-5625".

( d ) Uncertain Coins.

These are two copper coins, from the neighbourhood of Khotan ;

apparently Indo-Bactrian, but too much worn to permit of identification.

One is a small round coin, measuring | inch, weighing 18'5 grs., and

showing on one side traces of a bull’s head facing (?), within an irregular

square, enclosed within a marginal circle of dots, without any legend :

the other side is entirely indistinguishable. The only, hitherto known,

Indo-Bactrian coins with a bull’s head facing, so far as I know, are two

square copper coins of Menander, in Brit. Mus. Cat., No. 66, p. 49 and

No. 4, p. 169 (pis. xii, 5 and xxxi, 10). The other is a small, apparently

square coin, measuring f of an inch, weighing 11 grs., and showing on

one side traces of a conventional stupa ( P) surrounded by an illegible

legend : the other side is quite indistinguishable. The only, hitherto

known, coin with a stupa, I believe, is a square copper one of Agathocles,

in Brit. Mus. Cat., No. 15, p. 12 (pi. iv, 10).

IV. Indo-Scythian Coins.

The coins of this class number 10, and belong to two distinct

periods, an earlier from about 50-130 A.D., and a later from about

490-570 A.D.

(a) Early.

These coins, numbering 9, were found in the collections M. 2, M. 3,

M. 6, G. 10, and T. 1. They came from the Khotan country, and their

condition shows that they have been dug out from ancient sites.

(1) Kadphises II, c. 50 A.D., two coins, copper ; obv. and rev.

designs just discernible, legends quite obliterated; type (obv. King

Standing; rev. Qiva and Bull) as shown in British Museum Catalogue,

pi. xxv, No. 12. Size TO". Weight 240, 5 and 181*5 grs.

(2) Kanerkes, c. 78-110 A.D., six coins, copper, of twc different

sizes ; all in very poor condition.

28

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

( a ) four coins; obv. King standing to right, rev. figure standing

to right, its posture resembling MIOPO or MAO or AGPO ; there

are only faint traces recognizable ; on one obverse also traces of the

Greek legend. Two weigh 54 grs , one 64, and one 46 grs., but

a piece of the last is broken off its edge. Size of all, 0‘7".

( b ) two coins; size 0 5/r ; weight 315 and 20*5 ; one reverse shows

figure and legend MAO ; the other shows traces, apparently of OADO

(figure stepping to right, with both arms uplifted).

(3) Hoerkes, c. 110-130 A.D., one coin, copper, in poor condition ;

sizeO’875"; weight 78*0 grs., obv. King standing to right; rev. faint

traces of fiva and Bull to left.

(b) Late.

(1) Toramana, c. 495-510 A. D., one coin, copper, indifferent con¬

dition. See Cunningham’s Coins of Mediaeval India , p. 42, pi. iii, 1, 2.

Size 10". Weight 83*0 grs.

Y. Sassanian Coins.

(c. 458-484 A.D.).

There are seven (or six) of these, all apparently of Firuz II (458-484

A.D.) Plate I, Nos. 5 and 19. They belong to M. 2. They are of

some mixed metal, and inextricably baked together in two clumps, one

consisting of three and the other, apparently of four coins, weighing

192-5 and 205'6 grs. respectively, and measuring P125".

YI. Medieval Hindu Coins.

(c. 900-1100 A.D.).

These coins number 8, and belong to the following two classes,

(a) Maharajas of Kashmir.

The Kashmir coins number 6. They belong to M. 6, and were pro¬

cured from Khotan, probably found in its Bazars, and not in sand-buried

sites. They are similarly still found in Kashmir and India. There

has always been commercial intercourse between Khotan and Kashmir.

(1) A very early coin, but unknown. No legend on obverse, one

aksara, illegible, on reverse.

(2) Sugandha, c. 924-926 A.D., one coin, copper. As in Journal ,

As. Soc. Beng., vol. XLVII1 (1879), p. 281, pi. xi, No. 4.

(3) Diksema Gupta, c. 971-979 A.D., one coin, copper. As in

ibid., pi. xi, No. 6.

(4) Didda, c. 1001-1024 A.D., one coin, copper. As in ibid.,

pi. xi, No. 11.

1899.] Section I. — Goins and Seals. 29

(5) Harsa, c. 1062-1072 A.D., two coins, copper. As in ibid.,

pi. xii, No. 15.

(b) Brahman Kings of Kabul.

Samanta Deva, about 926-940 A.D. ; 2 coins, silver ; of the so-

called “ Bull and Horseman ” type, as in Prinsep’s Indian Antiquities

(ed. Thomas), Yol. I, Plate XXV, 3, 4, 5 ; weight 46 and 44 grs. ; size

0 7 and 0625". From G. 4.

VII. Medieval Muhammadan Coins.

(c. 800-1585 A.D.).

The total of these coins is 127. Many of them, as will he noticed

under the several coins, belong to G. 4, and were obtained in Western

Turkestan. Of the others, belonging to M. 2, M. 6, many were found

in the Takla Makan desert ; but it is probable that the more modern

ones were procured in Khotan itself and its bazars.

These coins belong to very different classes and ages. In the

following list they are arranged in chronological order.

(a) ‘ Abbasi Khalifahs.

Ar- Rashid ; 1 coin, silver, like British Museum Catalogue, Vol. I,

Plate V, No. 224 (p. 83) ; with a loop for suspension; mint Madina-

tu-l-Islam, date 192 H. ( = 807 A.D.) : weight 47*5 grs., size 0*83".

Belongs to G. 4.

(b) Khans of TurJcistdn.

(1) Yilik Khan; 3 coins, silver, like Br. Mus. Cat., Vol. II, Plate V,

No. 433 (p. 121); two of mint Samarqand,41 dates 397 and 39[8] H,

(=1006 and 1007 A.D.), weight 35*5 and 42 grs., size 0 9375" and

1*03125"; one of mint Sarraqustah, date 394 H. ( = 1003 A.D.) weight

38 grs., size 0*9375", see Plate T, fig. 21. The latter as well as one of

the Samarqand coins have on the reverse area <*!( above and Jtgj below

the central legend, but nothing corresponding on the obverse, while

the other Samarqand coin has All and on the reverse and or>B

on the obverse. From M. 2.

Yilik Khan, a chief of the Uighurs, is also known as Satuq Bughra

Khan. He lived from 333-429 H. ( =944-1037 A.D. ), to the age of

96 years. He was the founder of a very extensive, but short-lived

empire of the Uighurs, with a capital at Kashghar. See Dr. Bellew

in Sir T. D. Forsyth’s Report of a Mission to Yarkand in 1873 , pp. 125,

*1 On one of them apparently spelled Samarkand.

30

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

126 (also 121, 130), and Shaw’s Grammar of the Language of Eastern

TurJcistan , in the Journal of the Asiatic Society of Bengal for 1877,

p. 334.

(2) Muhammad Arslan Klian ; 14 coins ; all copper ; not in the

British Museum Catalogue ; date and perhaps mint were in the marginal

legend, which is almost entirely clipped off in all specimens. They are

from M. 2, M. 3, M. 9 and Gr. 10. There are three varieties, as

follows : —

First Variety. (Plates I, 22 and III, 15).

Six coins. Weight 105*5-62*5 ; size 0*9".

♦3 in most cases spelled &U|h.

Section 1. — Coins and Seals.

31

1899.]

Muhammad Arslan Khan was a son of ‘Ala Nur Khanim. a

daughter of Satuq Bughra (or Yilik) Khan and wife of Toq Buba

Khan ; he succeeded in or about 411 H. ( = 1049 A.D.) his uncles

Hasan Bughra, Husain Bughra, and Yusuf Qadir who reigned, in the

aggregate, twelve years. Both Hasan and Husain, as well as Muham¬

mad Arslan’s half-brother Sayyid ‘Ali Arslan Khan were “ martyred ”

in battle against the “ infidel ” Buddhists of Khotan. See Bellew,

ibidem , pp. 126-129 and Shaw, ibidem , pp. 334, 335, 339. The term

Al-mustanjid-billah means ‘ Seeker of help from God.’

(c) Ghori Sultans.

Muhammad bin Sam ; 2 coins ; mixed silver ; like British Museum

Catalogue (Sultans of Delhi), Plate I, No. 12, (p. 9), also No. 10 in

Thomas’ Chronicles of the Bathan Kings of Delhi ; of the “ Bull and

Horseman” type; date circa 1195 A.D. ; weight 45*5 and 53 grs., size

0*625". From G. 4.

(d) JJrtuqis of Mdridin.

‘Alau-d-din Kaiqobad ; 1 coin, copper ; like British Museum Cata¬

logue , Vol. Ill, Plate IX, No. 474 (p. 169, under Nasiru d-din’s coins);

mint lost, date 630 H. ( = 1232 A.D.) with on the left of the

central legend of reverse; weight 135*5 grs., size 1*16". Plate I, fig-

20. M. 2 (?).

This species of coin seems to have been imitated in making orna¬

mental plaques, and guards for binding block-prints ; see below General

Remarks on Block-prints and Plate IV, figs. 2 and 3.

(e) Khans of the Golden Horde.

(1) Jani Beg Khan; 2 coins, silver; like Br. Mus. Cat., Yol. VI,

Plate VII, No. 388 (p. 133); mint Khwarizm, date 743 H. ( = 1342

A.D.) ; weight 28*5 grs., size 0*6". From G. 4.

(2) Bardi Beg Khan ; 1 coin, silver; like ibidem , No. 413 (p. 146)

mint Khwarizm. date 760 H. ( = 1358 A.D.) ; weight 30*5 grs., size

0‘6". Plate I, fig. 29.

(/) House of Timiir.

(1) Shah Rukh; 1 coin, silver; similar to Br. Mus. Cat., Vol. VII,

Plate II, No. 61 (p. 24), but adds at beginning of third line, as

in No. 80 (p. 29) ; mint and date lost (reign 1404-1447 A D.) ; weight

78 grs.; size 0*875". Plate II, fig. 27. From G. 4.

(2) Sultan A^mad (1467-1493 A.D.) ; 1 coin, silver; the same

as ibidem , No. 61 (p. 24) of Shah Rukh, but counter-struck with a

32

Dr. Hoerule — Antiquities from Central Asia. [Extra No. 1,

six-foil die of Sultan Abroad, as in ibidem , No. 117 (p. 42) ; mint

obliterated by counter-struck die, date 828 H. ( = 1424 A.D.) ; weight

77'5 grs., size 0*875". Plate I, fig. 25. From G. 4.

(g) Sultans of Kashmir.

a) Zaiuu-l-‘Abidm, 1417-1467 A.D., one coin, copper. Like

Journal, As. Soc. Beng., Vol. XL VIII (1879), p. 284, pi. xiii, No. 2.

(2) Haidar Shah, 1467-1489 A.D., one coin, copper; date illegible.

Like ibidem , pi. xiii, No. 5.

(3) Hasan Shah, 1469-1481 A.D., one coin, copper; double-struck

on a coin of Haidar Shah ; date illegible.

(4) Muhammad Shah, 1481-1537 A.D. ; two coins, copper; dates

illegible. Like ibid., pi. xiii, No. 7.

(5) Fath Shah, 1483-1520 A.D., two coins, copper; dates ille¬

gible; one shows 90 (between 919 and 926 H. or 1513 and 1520

A.D.) Like ibid., pi. xiii, No. 8.

( h ) Unidentified.

(1) Seven coins, copper; notin British Museum Catalogue ; mint

Kashghar, date 950 H. in words ( = 1543 A. D.) ; weight 80*5-123 grs. ;

size 0*9375-1*2". From M. 2. Plate II, fig. 2 6a-d.

Obverse.

Reverse.

Single-lined large circular

area, within circle of

dots.

Small central square, with

semicircles upon the four

sides.

In square either alone

(2 6d) or with another

unread word (26e).

In semicircles (?) - (?) -

Two of the coins show a square counter-struck die (fig. 26d), with

^*£1? on it.

The two words after ‘adl probably contain a name which I am

unable to identify. At the time of the date of the coin several

dynasties were contending for the possession of Kashghar. e.g ., the

Bukhara rulers, the Mongol Khans and the Doughlat Amirs.

(2) Sulaiman Khaqan (?) ; 52 coins, copper; not in the British

Museum Catalogue ; with neither mint nor date; weight 26-116 grs.,

average 60 grs. ; size 1-1*4". From M. 2, M. 3, M. 4, M. 6, G. 5, G. 7,

G. 10. Two varieties, as follows :

1899.]

Section I. — Coins and Seals.

33

First Variety.

37 coins. Plate I, figs. 32-35.

Obverse. Reverse.

Within two concentric circles

with dots between them.

Octagonal area within a

circle, with jleur de lys

Inscribed margin nearly

obliterated, only ...^*1

visible.

Margin, inscribed with date,

nearly obliterated, apparent¬

ly 977 in words.

(4) Six coins; copper; similar to No. 2. Notread.

(5) Ten coins ; copper, of 8 different kinds ; unread ; five of them

shown in Plates I, 7 and 23, II, 24, III, 13 and 14. Two from G. 4.

(6) Two coins; silver; in very indifferent condition; unread;

apparently Mongol issues. Weight 30 and 21 grs., size 0 9" and

From M. 2.

(7) One coin ; gold ; mint and date lost with margin ; weight 75

grs. ; size 0’83". Plate III, 16.

J. i. 9

34

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

Obverse.

Circular area.

ill aJJ| if

*JJ|

s) y

Inscribed margin, almost

entirely obliterated and

illegible.

Reverse.

Circular area.

ill aJJ| il

o+zz* <*P|

Margin inscribed with date,

almost'entirely obliterated,

only visible.

(i) Line of Shaihan.

Iskandar; 1 coin, silver, with a loop for suspension; apparently

similar to British Museum Catalogue , Vol. VII, Nos. 135 and 145; mint

and date obliterated (reign 1560-1583 A.D.) ; weight 62 grs., size \2b".

Plate III, 12.

Obverse.

Square area. *

*U| ill il

Reverse.

Eight-foil area

with two five-rayed stars.

t>_ 4 - S?°

all) Jj**)

Margin nearly obliterated, only

^Lo [^J| ] at bottom visible.

'ft

...'ft

Margin entirely ’obliterated.

VIII. Modern Turk! Coins.

(18th and 19th centuries).

There are altogether 18 of these coins which fall into the following

three classes.

(a) Khans of Khoqand.

Apparently Shall Rukh II ; 1 coin, copper ; mint Khoqand ; date

lost, but should be 1184 H. ( = 1770 A.D.) ; not in British Museum Cata¬

logue ; weight 58 grs. ; size 0*66". Plate I, fig. 28.

Obverse.

Double-lined square area

across another double-

lined square, surround¬

ed by arabesques.

V

Reverse.

Single-lined small

cirular area within

broad ornamental

margin.

• —

V/®

1899.]

Section I. — Coins and Seals.

35

( b ) Ataliq of Kashahar.

Yaqub Beg; 3 coins, copper; not in British Museum Catalogue ;

mint Kashghar ; date 1293 H. (1876 A.D.) ; weight 51-58 grs., size

Both areas within a double-lined circle, surrounded by a circle of dots.

(c) Coins of Yarqand .

There are 14 of these coins. They are heart-shaped, and their

legends are counter-sunk. There are three varieties, differing by the

obverse legends ; the reverse legend of all (see PI. II, 28, first figure on

the left) is or mint Yarqand (or possibly Yarkand).

There is no date. Weight 107-132 grs. ; size 0*75 xO‘5625". The

obverse legends have not been read ; two of them are shown in Plate II,

28; the third in Plate III, 17. One of them (PI. II, 28, on extreme

right) is, by MunshI Ahmad Din, doubtfully read uk*. This

particular variety is the commonest. From M. 2, 3, 6, 9 ; Gr. 7, 10.

IX. Modern Indun Coins.

The total of these coins is 59. They are of a very great variety,

as follows.

(a) British Coins.

(1) East India Company; Sumatra; 1 coin, copper; obv. Cock,

rev. native legend, with date 1247 H. ( =1831 A.D.). See Indian Museum

Catalogue , No. 12083, in Vol. II, p. 133.

(2) One TV Anna piece of 1889. Copper. Obverse: Victoria

Empress.

( b ) Native States.

(1) Bhuj ; 1 coin, copper; weight 52 grs., size 0'6//. Obv. in NagarT,

round margin, Cri Khegar-ji Mirjd Mahdrdo , in centre trident, with 193S

( Samvat) below it ( =1882 A.D.) . Rev., in Persian, zarb Bhuj , Victoria

Qaisar-i- Hind, in centre sanah 1882 ( A.D.), with Rajput dagger below

it. Plate II, fig. 31.

(2) Bhopal; 1 coin, copper, J anna; mint Bhopal ; date 1277 H.

( = 1860 A.D.). See Journal , Asiatic Society of Bengal} Vol. LX VI,

p. 270, PI. XXXIII, fig. 57.

36 Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

(3) Orclia ; 1 coin; silver; 1 Rupee; mintUrcha; date 1211 H.

( «=1790 A.D.). See Journal , Asiatic Society of Bengal, Vol. LXYI,

p, 267, Plate XXXII, fig. 29.

(4) Sikh State; 10 coins, of different varieties, all copper:

(a) 2 coins; obv., star; rev., legend, date 1894 Samvat

( = 1838 A.D.) ; mint-mark, leaf, Plate II, fig. 29.

As in Journal, Asiatic Society of Bengal, Yol. L, Plate IX,

68 (p. 90).

( b ) 2 coins ; obv., mint Amritsar, with mark of leaf ; rev.,

GurmukhI legend, and mark of sword.

(c) 1 coin ; with mint-mark of leaf on reverse ; as in ibidem,

Plate Y1II, 53, but in very indifferent condition ; weight

122 grs. ; size O SS''.

(d) 2 coins ; legends in Persian characters ; obv., mint-mark,

leaf within of jaliis; rev., zarb\ but both mint and date

lost ; weight 188 and 86*5 grs. ; size 0*916" and 0*75".

(e) 1 coin, with GurmukhI legends, similar to ibidem, Plate

YII, 52. Weight 159*5 grs. ; size 0'916".

(/) 2 coins ; obv., mint-mark leaf, and mint on one Jchitah

Kashmir , on the other Jammu ; rev., legend Nanak Shah

with a leaf-ornament; weight 209*5 and 121*5 grs.;

size 0 9" and O S'7.

(5) Nepal; 1 coin; a modern paisa of the present reign; date

illegible.

(6) Jammu; 1 coin; weight 90 grs., size 0*7"; obv., Nagari

legend; rev., Persian legend with date 1937 Sam. ( = 1880 A.D.).

(7) Kashmir; 6 coins of two sizes, all extensively clipped, with

Nagari on obverse and Persian on reverse, and leaf and sword as

mint-marks :

(a) large, l coin, size 01 75" ; weight 84 5 grs. ; date 194*

Sam. ( = 188* A.D.).

(b) small, 5 coins; size 0*6-0*7"; weight 47-48 grs.; dates

1938, 1939, 1941 Sam. ( = 1882, 1883, 1885 A.D.).

(8) Afghan ; 6 coins, of different varieties ; all copper :

(a) 3 coins, of Timur Shah Durrani, ordinary Persian

legends on both sides ; obv., name and regnal year 12;

rev., mint Kashmir, date 118* H. ( = 177* A.D.) ; weight

230-278 grs. ; size 0*8-0*9375".*

(5) 2 coins ; apparently also of Timur Shah, but name lost,

date of one 1165 H ; rev., of one zarb Kashmir, of the

other, Jchitah Kashmir, date lost; weight 119 and 130

grs. ; size 0 75" and 0 625", but much clipped.

1899.]

Section I. — Coins arid Seals.

37

(c) 1 coin, ascription uncertain ; obv., curved sword in orna¬

mental ellipse; rev., zarb Kashmir' sanah 2 (?) ; weight

86 grs. ; size 09".

(9) Unidentified coins :

(а) large; 6 copper; 3 square, 3 round; with most fragment¬

ary legends; weight 169*5-83 grs. ; size 07-O66".

(б) small ; 24 copper, all round, with most fragmentary

legends ; weight 46*5-4 grs. ; size 0333-0,5625".

X. Modern European.

A Russian copper coin ; got from the house of a Khotan merchant ;

obv., Imperial arms, rev., Imperial monogram within laurel crown, and

date 1758 A.D. ; weight 802 grs., size T66". Plate II, fig. 32.

X. Seals, Intaglios, Cameos, Etc.

(Plate III.)

The total number of these objects in the collection is sixty-five.

They are of a great variety, in point of material, size, shape, form, and

engravement. There are also ten unfinished pieces (all of stone) which

had clearly been cut to make seals, but which bear no engraving of any

kind. Most of these objects come from the Takla Makan desert, and

belong to M. 2, M. 3, M. 4, M. 6, M. 9, G. 7 and G. 10. A few, belong¬

ing to G. 4, come from Western Turkestan.

With regard to material, there are of

brass or copper, bronze ... ... 37 specimens,

stone or glass, or ivory ... ... 28 ,,

Total 65 ,,

In size, they vary very widely ; from 1* 25" to 0'25" (PI. Ill, fig. 21)

in length.

In shape, there are the following varieties : —

Square or nearly so

Oblong

Triangular

Quaterfoil

Round or nearly so

Elliptical

Fanciful

Total

• t

38

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

With regard to form, among the seals there are three made in the

form of the ordinary signet-ring, to be worn on the finger. One of

them is complete (PI. Ill, 27 and PI. XIX, 3) ; in the case of the other,

two portions of the ring are missing. All three are of brass, and belong

to M. 2. There are also five pieces, all of stone, which are cut in the

form of a signet-ring, but their ring-portion is solid, with a hole drilled

through it, in order to be worn on a string (Plate III, 62 and

Plate XIX, 2). All these solid rings belong to G. 4, and probably came

from Western Turkestan. The usual form of the seal, however, is a

thin, flat plate, to the back of which is attached a small perforated peg

for the passage of the string on which it is worn (Plate XIX, 4). All

the seals of this kind are from the Takla Makan desert. They are

also all of brass or copper, except one (PI. Ill, fig. 29) which is made of

steatite. One seal, of bronze, is furnished with two flaps (PI. Ill, fig.

22 and PI. XIX, 6).

Among the intaglios, there are two with a high, conical back

(Plate III, 63 and Plate XIX, 7). They are both of glass, and came

from the Takla Makan desert. All the rest are flat pieces of stone or

glass, of the usual form, with a rounded or planed back.

Two of the objects (PI. Ill, 65 and 73 and PI. XIX, 8) appear to

have merely served as ornaments, as they are provided with the remains

of what seems to have been a nail. These are from the Takla Makan.

One object (Plate III, 23 and Plate XIX, 5) seems to be one of a

set of dice used in divination. This, also, comes from the Takla Makan

desert.

With reference to the matter engraved, it consists either of some

writing, or some figure, mostly human or animal, or some ornamental

design. Particulars, so far as definable, will be found in the following

detailed description. The figures on the Plate are all full size.

Plate III, No. 18. Round intaglio of black glass f13 engraved with

Arabic characters, apparently ‘ ajiyat Nadir or “ Success of Nadir.”

From G. 4.

No. 19. Oblong flat brass seal, with perforated peg ; face engraved

with what looks like writing in Persian characters ( .... ££*»). From M. 2.

No. 20. Obverse and reverse of a square amulet, of ivory;*13

thickness yth of an inch, through which runs a fine string-hole ; both

4>3 For the identifications marked * and f 1 am indebted to the kindness of

Mr. C. L. Griesbach, C.I.E., and Dr. Fritz von Notling respectively, of the Geological

Survey of India. Mr. Griesbach informs me (19th January, 1899), that “ the stones

might all have come from Badakshan or Bokhara. Lapis lazuli conies certainly

from Badakshan, and spinel is found in Shignan and other places on the upper

Oxus, also at Tagdallak east of Kabul,”

39

1899.] Section I. — Coins and Seals.

faces engraved with what looks like writing. From the Takla Makan ;

consignment unknown.

No. 21. Minute lozenge-shaped intaglio of bronze, engraved with

minute writing, apparently Arabic, unread. From G. 10.

No. 22. Square bronze seal ; back furnished with two flaps, ap¬

parently for fixing to a handle (PI. XIX, 6) ; face engraved with what

seems to be writing. From M. 3.

No. 23. (PI. XIX, 5). An oblong piece of brass, with four equal

sides measuring T% x T0g inches, the ends accordingly being •§ th of an

inch square ; perforated right along its long axis, and bearing engrave-

ments on all its four long sides. One side shows a dot ; the side next

to it has two strokes ; these would seem to be intended for the numbers

1 and 2. On the other two sides one would expect the numeral signs

for 3 and 4 ; and so they may be ; but they are very indistinct, and,

curiously enough, seem to suggest human standing figures : that on

the third side being a long-robed woman with shield and spear ; that

on the fourth side, a man standing to the left before a fire altar, as on

Kanishka’s coins. But these resemblances may be deceptive, the

engravements being excessively corroded. The piece is probably one

of a set of dice, such as are still used in the present day for purposes

of divination. See the remarks below.

No. 24. Elliptical intaglio ; Grecian ; engraved with nude figure

of a man standing to front, with his wrists behind his back, bound to a

pillar (Prometheus ?).

No. 25. Elliptical intaglio ; Indian ; engraved with bust of a

woman, showing ancient Indian fashion of hair-dress.

No. 26. Elliptical intaglio ; Grecian ; engraved with seated

draped figure of woman to left ; perhaps Tyclie with cornucopice in left

and sceptre in right hand.

No. 24 to 26 were found “ near Khotan,” and are those “ three

pieces of yellow crystal of an oval shape ” mentioned on p. xxxii of the

Introduction. Mr. Macartney only sent me impressions taken in sealing-

wax. From these plaster-casts were made by me for the photographic

plate. This accounts for the want of clearness in the figures.

No. 27. Elliptical signet-ring of brass. (Plate XIX, 3). Engrave-

ment too much worn to be clearly recognised ; apparently a woman

seated to left on a wicker-stool. From M. 2.

No. 28. Obverse and reverse of a square amulet, made of serpen¬

tine * ; thickness ftli of an inch, perforated with a string-hole. Obv.,

centaur to right, holding some object in each uplifted hand ; below a

four-rayed star, like X . Rev., lion to right, with open mouth threaten¬

ing a man in attitude of supplication before him ; above tail three dots.

From M. 3.

40 Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

No. 29. Round seal or button, of steatite *, with perforated pro¬

minence on back, engraved with rude radiate sun-face. From M. 3.

No. 30. Round intaglio of red cornelian, engraved with the head

of_a woman. From G. 10.

No. 31. Round intaglio of black glass f, engraved with the head

of a man (Parthian ?). From Gr. 4.

No. 32. Elliptical intaglio of red cornelian ; Grecian ; Pallas to

right, in long robe and helmet with spear and shield. From G. 4.

No. 33. Similar to No. 32. Apparently draped female figure to

left. Original gone astray.

No. 34. Square flat brass seal, with perforated peg, showing

kneeling figure of a man to left, with uplifted right arm, before some

small indistinct object (fire- worshiper ?). From M. 2, found at

Aq Sapil.

No. 35. Square brass flat seal, with broken, perforated peg at

back, engraved with two men, standing to front, but facing one

another, both in the same attitude, left arm uplifted, right arm

suspended From M. 2, found apparently in Kok Gumbaz.

No. 36. Elliptical intaglio of very thin brown glass, convex,

about one-half broken off and missing. Engravement, nude figure,

apparently a monkey, with feathered helmet, holding up a branch in

left hand. From G. 4.

No. 37. Round intaglio, of blue glass*; showing draped

(female ?) figure, seated cross-legged to front, with uplifted arms,

holding in left hand a wreath (or sun ?), in right hand a crescent.

Posture similar to that of king or deity on Indo-Scythian and Gupta

coins ; compare Ariana Antiqua , Plate, xviii, 4, 5 ; British Museum

Catalogue , PI. xxix, 4. From M. 3.

No. 38. Elliptical intaglio, of lapis lazuli*, showing head of a

bird to right, wearing helmet or cap. From G. ]0.

No. 39. Oblong intaglio of sardonix*, very deep red, with whitish

surface on the engraved side; showing nude dancing girl, wearing very

heavy girdle and anklets, standing to right, and presenting some

lengthy object in both forth-stretched hands, in posture similar to soldier

presenting arms. From G. 10.

No. 40. Round intaglio, of glass or obsidian,* showing a goat or

long-horned gazelle, walking to right. From G. 10.

No. 41. Square flat bronze seal, with perforated peg, showing a

hare sitting to right. From M. 2.

No. 42. Square flat brass seal, with perforated peg (broken),

showing a crude calf, walking to left. Similar to No. 43. From M. 2,

found at Aq Sapil.

i

1899.]

Section I. — Coins and Seals.

41

No. 43. Round brass seal, with perforated peg, showing a sheep

or goat walking to right. From M. 9.

No. 44. Square brass signet-ring (portion of ring missing),

engraved with large-horned sheep (ovis Poli?) walking to left. From

the Takla Makan desert ; consignment unknown.

No. 45. Round intaglio of blackish agate with whitish surface on

the engraved side (similar to No. 39), showing a deer running to right.

From G. 10.

No. 46. Round solid perforated seal-ring, of whitish agate, t

showing a crudely made hare (P), running to right. From G. 4.

No. 47. Round intaglio, of spinel *, showing a hare running

to right. From G. 10.

No. 48. Square flat brass seal, with perforated peg (broken),

showing head of a bull en face. From M. 2, found at Aq Sapil.

No. 49. Round intaglio, of sardonix *, very deep red, with whitish

surface on the engraved side (similar to No. 39), showing a cow stand¬

ing to right, with suckling calf. From G. 10.

No. 50. Round flat brass seal (much worn), with perforated

peg (broken) ; showing a sheep, walking to left, very crude. From

M. 6.

No. 51. Round intaglio of red corneliau ; showing a horse, bridled

and belted, jumping to left. From G. 4.

No. 52. Round flat brass seal, with perforated peg, showing an

elephant standing to left. From M. 2, found at Aq Sapil.

No. 53. Square flat brass seal, with perforated peg (broken),

showing an elephant, walking to left, harnessed with drapings and

carrying a Buddhist relic casket. From M. 2, found at Aq Sapil.

No. 54. Elliptical solid perforated seal-ring, of whitish brown-

veined agate f; showing Brahmani (humped) bull, standing to right,

within chaplet of astragals. From G. 4.

No. 55. Round flat brass seal ( much worn), with perforated peg

(broken), showing a lion, walking to left. From M. 2.

No. 56. Round intaglio, of red cornelian, showing a lion walking

to right. From G. 10.

No. 57. Elliptical intaglio, of lapis lazuli*, showing Pegasus

walking to right ; two unread letters (Pahlavi F ) above hind-quarters.

From M. 3.

No. 58. Round brass seal (broken), with perforated peg, showing

a goose walking to left. From M. 2.

No. 59. Round solid perforated seal-ring, of whitish agate, f

showing parrot walking to right. From G. 4.

No. 60. Square flat brass seal, with perforated peg (broken)

J. i. 10

42

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

showing a roaring lion, seated to right, similar to No. 61, but without

raised paw. From G. 7.

No. 61. Square flat brass seal, with perforated peg, showing a

lion or cat sitting to right, with open mouth and uplifted left paw. From

M. 2, found at Aq Sapll. Auother, but smaller specimen, from G. 7, has

the paw not uplifted.

No. 62. Round solid perforated seal-ring, made of whitish agate, f

showing a scorpion, walking to right, with raised sting. From G. 4.

No. 63. (PI. XIX, 7). Conical intaglio, of green glass (much

corroded) ; engravement indistinct, apparently a locust, flying to right.

From M. 4.

No. 64. Round flat brass seal (broken and much corroded), with

perforated peg, showing two birds (bulbul?) facing each other and

fighting. From M. 2, found at Aq Sapil.

No. 65. Round flat ornament of copper, with ring attached to rim,

for wearing. In the centre, remains of a nail, showing that originally

there was something attached to its back. From M. 9.

No. 66. (Plate XIX, 8). Brass nail with round ornamental head,

formed like a mushroom.

No. 67. Ruund flat brass seal, with perforated peg (broken),

showing radiate sun-face, or perhaps wheel.

No. 68. Triangular flat brass seal, with perforated peg, showing

a scorpion, walking to right. From M. 6.

No. 69. Elliptical intaglio of red cornelian, engraved with an

uncertain design (pair of cymbals ?) From G. 4.

No. 70. Square flat brass seal, with perforated peg (broken),

showing an ornamental design. From M. 6.

No. 71. Quarterfoil flat brass seal, with perforated peg, and

original piece of thread on which it was worn. Engraved with an

ornamental design. From G. 10. Another specimen was in M. 2.

See No. 79.

No. 72. Square brass signet-ring (larger portion of ring missing),

showing an ornamental design. From the Takla Makan ; consignment

unknown.

No. 73. Round intaglio of whitish glass, f engraved on one side

with the figure of two crossed swords or arrows, on the other, with some

kind of faintly incised writing. From G. 4.

No. 74. Obverse and reverse of a round amulet, of serpentine ;*

thickness, £th of an inch, perforated with a fine string-hole. From

the Takla Makan; consignment unknown.

No. 75. Elliptical solid perforated seal-ring, of Bowenite,t

engraved with fire-altar, as on Sassanide coins. From G. 4.

1899.]

Section 1. — Coins and Seals.

43

No. 76. Square flat seal of brass, with perforated peg, engraved

with an ornamental design. From M. 2, found at Aq Sapil.

No. 77. Square flat seal of copper, bearing four square ornamental

designs, two of them being different forms of the Svastika. From M. 6.

No. 78. Square flat copper seal, with perforated peg, engraved with

an ornamental design. From M. 6.

No. 79. Square flat seal of brass, with perforated peg, engraved

with an ornamental quaterfoil design. From M. 2, found at Aq Sapil.

Another specimen, of the same size, was in M. 9. See No. 71.

No. 80. Flat copper seal, consisting of a square surmounted with

a tridental crown, and furnished with a perforated peg. Bearing

ornamental designs, that on the square being the same as on No. 79.

From M. 6.

No. 81. (Plate XIX, 11). Elliptical intaglio, perhaps of felsite*;

Grecian ; showing a draped and helmeted figure, sitting on a stool (? ),

holding a bird (?) on his outstretched right hand. From G. 4.

There is one cameo in the collection. See Plate XIX, 9. From

G. 10. It shows the helmeted head of a young man, of Grecian

design.

Among the gems, shown on p. 779 of Dr. Sven Hediti’s Through

Asia , aud obtained by him in Kliotan, there are several which are strik¬

ingly like some in the British collection. Thus the second in his first

line of facsimiles resembles our No. 30. There are three other similar

ones on that page, but they differ in having a ribbon round the neck

of the figure shown on it. There is also there one gem strikingly like

our No. 29. The deer, hare, and large-horned sheep are also found

on several of them.

But what is more noteworthy is that exactly similar seals and

intaglios have been discovered in the ancient stupas of Afghanistan.

Samples of these are shown in Wilson’s Ariana Antigua , Plates i, iii

aud iv. Thus figs. 7-10 on PI. iv, show two square flat seals with

perforated peg, made of iron or brass. Plate i, fig. 8, PI. iii, fig. 7

and PI. iv, figs. 10, 11 are similar cornelian intaglios. The signet-rings,

shown on PI. i, 5, and PI. iii, 6, though similar in shape to our Nos. 27,

44, 72, differ in being more costly, being of gold with inlaid cornelians,

while those in our collection are made entirely of brass or copper.

I may also note the evidences afforded by these objects to the

prevalence of Grecian and Buddhist culture in ancient Kliotan.

Nos. 24, 26, 32, 33 on PI. Ill and No. 11 on PI. XIX are distinctly

Grecian ; so are the centaur on PI. Ill, No. 28, and the pegasus on

No. 57. Distinctly Indian are the Brahmani bull on No. 54, and the

elephant on Nos. 52, 53. Distinctly Buddhist are the Svastika on

44

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

No. 77, and the relic-carrying elephant on No. 53. Old Persian

(Zoroastrian) are the fire-altar on No. 75, and the fire-worshipper (F)

on No. 34.

With reference to No. 23, I maj^ explain that there is a certain sys¬

tem of divination, well-known all over Northern India. It is practised, I

believe, only by men of the extreme North-West, “ Kashmiris ” as they

are commonly called, who are, as a rule, Muhammadans. They use a

double set of four brass dice, strung on two short iron rods, round which

they freely revolve ; four dice on each rod. The eight dice are all made

exactly alike; being rectangular parallelopipeds (PI. XIX, 5), with

only four equal sides (not cubes), and marked, on the long oblong sides,

with the numbers 2, 3 and 4, denoted by dots, in such a manner that

2 stands on the side opposite to 4, and 3 opposite to 3, as shown in

the subjoined woodcut.

1

1 he short square sides, of course, which are perforated for the iron

rod, bear no numbers; nor is the number one used. The dice look

as if they were made of brass, but they are said to be of a special

alloy of seven metals, consisting of brass, pewter, iron, lead, silver,

gold, and copper. The operator throws the two strings of dice so

that they fall parallel to each other, and then counts the dots in

parallel lines; thus, lines a and b give 6 each, arranged as 2 + 24-1 + 1

and 2 + 1 + 1 + 2, or a combination of 12. Each of the two lines might

give any number from 4 to 8, and between them a great variety of

arrangements and combinations. From these variations the -diviner

makes his forecasts.441 The object, described under No. 23 looks very

much like one of such a set of dice, only that its faces are marked

differently from what is the custom at the present time.

44 A description of the alloy ns well as of the modus ojperandi in divining with

such dice will also be found in the “ Third Report of Operations in Search of

Sanskrit Manuscripts in the Bombay Circle, April 1884 to March 1886,” by

Piofessor Peterson, pp. 44-46, printed as an Extra-Number of the Journal of the

Bombay Branch of the Royal Asiatic Society. Both the description and the sketch

of the dice are not quite accurate.

No. 10.

1899.]

Section II. — Block Prints.

45

Section II.— BLOCK PRINTS.

General Remarks.

The following is a summary of the Block-prints in the collection :

Total ... 45 books.

With the exception of one, the block prints all alike resemble

_ _ European books in their style of binding1.1 A

Style of Binding. , / , . c , , , . ., ®

sheet of paper is folded in the middle to

form two leaves, with four pages. A number of such folded sheets or

“ forms ” are then fastened together, along the line of the fold, to make

up a book. For the purpose of fastening them, they are, as a rule, simply

laid one upon the other ; but there are three books, all belonging to the

Third Set, in which they are not laid one upon the other and outside

the other, but placed one within the other so that the entire book forms

but one folded bundle. Occasionally also, as in No. I of the First Set,

a double form is met with, made up of two folded sheets, placed one

within the other and thus consisting of four leaves or eight pages.

The fastening is done in three ways : either by thread (2), or by

twists of paper (12), or by pegs of copper (30).2 The last-mentioned

method is the commonest : the relative frequency is indicated by

1 See also my Note on some Block-prints in the Proceedings , Asiatic Society of

Bengal , for April 1898, p. 124.

8 I may add that Mr. C. Bendall informs me (in a letter dated the 1st October,

1897) that the British Museum possesses a book in which “ the peg is of wood, not

metal.’

46

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

the bracketed numbers. The pegs are applied like rivets, that is,

tbeir ends are bent over and beaten down fast ; but before doing so,

they are passed , through thin pieces of copper, as a protection to the

leaves. Sometimes (as in No. YII of the First Set, No. Ill of the Fifth

Set and No. V of the Seventh Set) these guards are round ornamental

pieces, resembling coins, as shown in fig. 3 of Plate IV ; or they are

elliptical (as in No. VIII of the First Set) or oblong (as in No. VI of

the Seventh Set) ; but many of them (as in Nos. I and VII of the First

Set, Nos. I and III of the Fourth Set, No. Ill of the Fifth Set, and

No. I of the Seventh Set) are evidently portions of a large ornamental

circular plaque which had been cut into pieces ; and in this case the

pieces are very large, as shown in Plate IV, fig. I. In two cases (in No.

I of the Second Set, and in the book of the Eighth Set) the guards

consist of two long slips of copper, extending the whole length of the

book. The twists of paper are applied, like pieces of string, to form

loops. The stitching with thread only occurs in two books, in No. II of

the Third Set and in the book of the Ninth Set. As a rule the fastening

is made in three separate places, by three nails, or three twists of paper,

or three loops of thread. Twice, however, in No. Ill of the Fourth Set

and in the book of the Ninth Set, four pegs and four threads respectively

are used; and once, in No. II (Pothi) of the First Set, only one peg. A

fastening in two places is more frequent ; four times (in No. VIII of the

First Set, Nos. V and VI of the Fifth Set, and No. IV of the Seventh

Set) only two pegs are used ; and twice (in No. VII of the First and

No. VIII of the Fifth Set) only two twists of paper. The three initial

and the three fiual pages are as a rule left blank ; and thus the first and

the last leaves, being blank, serve as covers to the book. In one case

(No. VIII of the Fifth Set) seven initial and seven final pages (i.e.,

three leaves on each side) are left blank. The probable object of this

arrangement is disclosed by No. II of the Second Set and No. I of the

Fifth Set, in which four and two leaves respectively have been pasted

together to form pasteboard covers at either end. The single exception

above referred to is a pothi (No. II of the First Set), that is, a book

arranged in the Indian fashion. In this case,

the whole collection of “forms” is placed

between two pieces of wood, and held in position by riveting it with one

copper peg, passed through the middle of one of the narrow sides, as

shown in Plate VIII. The arrangement is exactly the same as in the

case of a set of Indian copper-plates of a landgrant. The peg takes the

place of the seal-ring of the grant, or of the string of a manuscript.

The two ends of the peg are split in two, and the two splits are turned

over right and left, after having been passed through. the thin copper

The Pothi.

Section II. — Block Prints.

47

1899.]

Size and shape.

guard ; thus the whole pothl is kept firmly fixed. The two wooden

covers are thick rough pieces (8fx4fxl") of a very light kind of

wood, the outside surfaces of which are not planed.

In size and shape the block-prints vary greatly. Some are

narrow oblongs, measuring from 9^ to 14f by

4 to 4J inches ; but mostly they are broad

oblongs, the largest measuring 23§ x 13V', the smallest, 6| x 4j". Their

thickness, also, much varies, depending, of course, partly on the

number of forms contained in them, partly on the thickness of their

paper. The thickest is the pothi, its wooden covers alone measuring

together two inches. Farther details of measurements will be given

with the following description of the several xylographs. In a few cases

the corners are slightly rounded off : in one case this is done so much as

to render the shape of the book eliptical ; see fig. VIII in Woodcut

No. 12. The edges of the leaves are frayed, as if the sheets had been

cut with a blunted or notched instrument. Very exceptionally I have

found the edge of a pair of leaves uncut. In these cases, when fastening

the book, a folded sheet had been put in wrongly with the fold outside

instead of inside. On the other hand, in five cases (First Set, Nos. IV

and V, Fourth Set, Nos. Ill and VII, Seventh Set, No. V) I have found

all the folds cut through, so that practically the book consists of separate

leaves, instead of forms. This is also the case in No. VII of the Fifth

Set, where, however, the leaves appear to have become separated by

the wear and tear of the folds.

The xylographs are all printed on paper. The paper appears to in¬

clude, at least, three distinct classes. One class

is a soft paper, thin, and of even texture, much

like the white or whitish paper of the Weber and some of the Macartney

Manuscripts, published by me in the Journal of the Asiatic Society of

Bengal, Vols. LX1I and LXVI, which is believed to be made of the bark

of the laurel (Daphne). This class of paper is found only in the one book

which constitutes the Eighth Set. It has a deep yellow colour, which is

probably a tint artificially imparted. Another class of paper is only found

in the book and the roll, comprised in the Ninth Set. This is an exceed¬

ingly thin, almost transparent, tough paper, of even texture, with a light

yellowish tint, probably natural. In its present condition it has become

somewhat brittle, from age. Except in colour, it reminds one of what,

in the trade, is known as “ parchment overland paper ; ” in fact, at first

sight I thought it was very fine vellum, though on closer examination

and washing, it at once revealed itself to be paper. The most common is

a third class of paper, of a more or less uneven texture and thickness,

the prevailing colour of which is a more or less dirty yellowish -brown

Paper.

48

Dt'. Hoernle- — Antiquities from Central Asia. [Extra No. 1,

Four distinct varieties are observable. The first variety is a soft,

thickish paper, to the toucli resembling felt or cloth, of comparatively

even texture and rather brighter yellowish-brown colour. It is found

in nine books ; viz., Nos. I and II of the First Set, Nos. I, III, IV, V

and VI of the Second Set, and Nos. II and III of the Third Set. The

second and third varieties differ from the first variety only in being

progressively thinner, of more uneven texture, and of darker colour.

The third variety, indeed, is occasionally of an exceedingly flimsy make.

The second variety is found in nineteen books ; viz., Nos. II, III, IV and

V of the First Set, No. II of the Second Set, No. I of the Third Set,

Nos. I, II, III, IV, VI and VII of the Fourth Set, Nos. I, II and VIII

of the Fifth Set, and Nos. I, II, III and V of the Seventh Set. The

third variety is found in five books ; viz., Nos. VII and VIII of the First

Set, No. V of the Fourth Set, No. Ill of the Fifth Set, and No. VI of

the Sixth Set. All these three varieties are comparatively soft papers,

and in this respect differ from the fourth variety, which is a hard and

stiffish paper, of middling thickness, and of very uneven make. This

fourth variety much resembles, except in point of colour and age,

the kind of paper which is still made in Khotan at the present day.

It is found in nine books ; viz., No. VIII of the Fourth Set, Nos. IV, V,

VI, VII of the Fifth Set, and Nos. I, II, III and IV of the Seventh

Set. With the exception of one book, they all belong to M. 8 ; and

the single exception belongs to M. 9. The three other varieties do not

resemble the modern Khotanese paper, though it is probable that they

all are of Khotanese manufacture, being probably made of the same

material, and by the same or a similar system of preparation. I

am disposed to believe that the four varieties of this class of paper

represent four different periods and four successive degradations of

Khotanese paper manufacture. The texture of the modern Khotanese

paper is exceedingly coarse and uneven, its pulp having been prepared

very roughly and spread very unevenly. When fresh, the paper has a

creamy or greyish colour : the much darker colour of the corresponding

paper of the block-prints is the effect of age. Regarding its material I

have received two different statements. The Rev. Magnus Backlund,

Swedish Missionary in Kashghar, who has visited Khotan, informs me

in a letter, dated the 29th June, 1898, that “it is made of the bark of

the willow, softened in lye, and then taken up and beaten between flat

stones, which of course, cannot be made so well as to prevent small

pieces of bark remaining here and there.” According to Munshi Ahmad

Din, of the Kashghar Agency, in a note written for me on the 19th

December, 1898, “ the Khotan paper is a very coarse stuff, chiefly

composed of silk waste.” In the sequel these classes and varieties of

Section II. — Block Prints.

49

1899.]

Block-printing.

paper are referred to as I, II, Ilia, III&, IIIc, and Hid respectively.

The paper of the Pothi (No. II of the First Set) I am doubtful in

classifying, but it probably belongs to Class III b.

That these books were printed from blocks of type is apparent

from the fact that the text is repealed over

and over again, from page to page, the repe¬

titions being facsimiles, as shown by measurements made by me (see

below under the First Set). The type, cut on the block, was enclosed

in a square of raised straight lines, and occasionally these inclosing

lines are printed off along with the inclosed type ; but as a rule they

do not seem to have been inked, and only a few traces of them, here

and there, are seen (as, e.g., on Plates IX and X). The printing was

not always carefully done ; occasionally the blocks were inked too much,

and the impressions are smudgy : at other times they were inked

too little, and the impression is almost illegible. When the print is

repeated on the same page, the impressions, for the sake of economy,

were sometimes placed so close together as to cause the margins of

the prints to run into one another and obliterate the letters. From the

fact that sometimes one has to remove the rivets, in order to be able

to read the whole of the impression, it is evident, that, as a rule, the

sheets or pages were printed first before they were stitched or riveted into

books. In some books, especially of the Fifth, Sixth and Seventh Sets,

the paper appears to have been more or less strongly greased, before

printing, possibly with the object of sizing it ; but the process has

sometimes had the effect of rendering the impressions almost illegible.

A regular system appears to have been observed in printing the

xylographs. As already explained, the first

and the last leaves of a book were always left

blank, for the purpose of serving as a cover. For the same reason, the

exterior pages of the second and penultimate leaves were also left blank.

The printing almost invariably commences on the interior of the second

leaf ( i.e ., the 4th page of the whole book), and stops on the interior of

the penultimate leaf (i.e.y the ante-ante-penultimate page of the whole

book). There are a few exceptions, which will be noted in the detailed

description; see, e.g ., No. VI of the First Set. Thus supposing a

book had six leaves or twelve pages, the imprints would commence

on the fourth page and stop on the 9th page ; pages 1, 2, 3 and 10, 11,

12 being blank. With regard to the arrangement of the imprints on

the pages, the principle (to which there are only a very few exceptions)

was that they were placed alternately in an upright and reversed

position. Whence it follows that, in reading a book, one would at first,

read consecutively, throughout the book, all the upright impressions ;

J. i, 11

System of Printing.

50

Dr. Hoernlc — Antiquities from Central Asia. [Extra No. 1,

next, turning the book right round, one would commence at the back

of the book, and read consecutively, right through the book, all the

reversed impressions, which, however, would now of course, stand

upright towards the reader. The subjoined Woodcut No. 11 illustrates

this system, and the various modifications in which it is applied.3 The

dotted lines in the diagrams signify the lines of type, and the letters a

and b indicate the beginnings and endings respectively of the impressions

of the text. In No. I, there are two impressions in each column, stand¬

ing foot to foot ; those in the upper halves of the pages standing upright,

while those in the lower halves are reversed. In this book one would

read, first, consecutively all the impressions in the upper halves of

the pages, in regular order (4, 5, 6, 7, 8, 9, etc.), up to the end of tho

book. Next turning the book right round, in the direction of the arrows,

and thus bringing the reversed impressions into an upright position, one

+.£8. e/Mc.

would read consecutively all the impressions (of the formerly lower

half-pages) from the back to the beginning of the book, in regular order

(etc., 9, 8, 7, 6, 5, 4). The same system, in another, slightly modified form,

3 The Roman numbers in this and the following Woodcuts refer to tho books

of the First Set. The diagrams are drawn to the scale of 1 inch in the woodcut

to 12 inches of the original.

1899.]

Section II. — Block Prints.

51

is applied in Nos. II, III and IV. In these books the impressions stand

upright on pages 4, 6, 8, etc., but reversed on pages 5, 7, 9, etc. In

reading any of these three books, one would begin by reading consecu¬

tively all the even-numbered pages (4, 6, 8, etc.), throughout the book.

Next, one would turn the book right round, as indicated by the arrows,

and thus bring the imprints, which hitherto had stood reversed, into an

upright position; and now, commencing at what was the end of the

book, one would read consecutively all the odd-numbered pages (etc.,

9, 7, 5) up to the original beginning of the book. The same system,

again, in a third modified form, may best be seen in book No. II of

the Seventh Set. Here the impressions do not stand upright and

reversed on alternate pages, but in alternate columns, as shown in

the diagram on page 55. Thus they stand upright in columns 1

and 3, but reversed in columns 2 and 4. In reading one would

commence with all the left-hand columns consecutively throughout

the book; then one would proceed to turn the book right round,

and now read all the impressions (of the former right-hand columns)

from the back to the beginning of the book. In this way the reading

of the entire book would be completed. To this principle of arrange¬

ment there are only a few exceptions, in which all the impressions are

placed upright (or in the same direction) throughout the book, so that

the book can be read right through, from page to page, without turning

it right round. This is the case in books No. IV of the Second Set,

Nos. IV, V and VI of the Fourth Set, No. II of the Fifth Set, No. II

of the Sixth Set, and No. IV of the Seventh Set.

There is, however, a certain number of books, in which the orderly

arrangement of imprints, above explained, is

not observed. In some of these books, indeed,

no system of printing whatever can be discerned. The impressions

appear to be placed promiscuously ; the only apparent object being to

crowd as many impressions into a page as it may, by any device, be made

to hold. This may be seen from figs. V and VI of the subjoined woodcut

No. 12. Two conspicuous examples of this kind of book are No. VI

of the First Set, and No. Ill of the Sixth Set. With such an arrange¬

ment, obviously, no intelligent and orderly reading of the book is

possible. Beside these there are some other books in which the absence

of all orderly arrangement is not quite so conspicuous. In these the

imprints are orderly placed on each page taken by itself ; that is,

on some pages they are all placed upright, on others, all are placed

reversed ; but these differing pages do not follow one another in any

system. In any pair of pages one may meet with any of the four

possible arrangements : upright- upright, upright-reversed, reversed-

52 Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

reversed, and reversed-upright. All these four arrangements occur with

No. 12.

V

£ I'd ^ t-O

VI

almost equal frequency. A good example of this kind is hook No. V

of the Seventh Set. In such books, too, any orderly reading is out

of the question.

If it were certain what the object of these books was — whether they

were, or were not, intended for reading, — the

Object and Use of presence or absence of systematic arrangement

the Bk>ck^ print 0£ the imprints might afford a good test to

determine the genuineness, or otherwise, of

the books. So long as their purport remains undeciphered, their object

must be a matter of speculation ; but the fact that they contain nothing

but interminable repetitions of the same text seems clearly to indicate

that in these books we are dealing with set formulas — creeds, prayers,

or incantations, or whatever one may call them, — possibly or probably

Buddhistic, — the virtue of which was supposed to be in proportion to the

frequency of their repetitions. The mode of this repetition, however,

need not necessarily have been an intelligent one : it might have been

quite mechanical, like that of the prayer-wheel or the prayer-flag.

Turning the leaves of a book would serve the purpose of the devotee quite

as well as turning a wheel, or letting the flags be moved by the breeze.

If this were the object of the books, it is evident that the order or want

of order in the arrangement of the formulas would be altogether im¬

material, provided the page is well covered with them. In any case,

whether the leaves were intended to be read, or merely to be turned^

1899.]

Section II. — Block Prints.

53

it is plain that there was no need of numbering them, seeing that, the

contents being merely a repetition of a set formula, one might use the

leaves in any order. As a matter of fact, none of these books have

their |leaves or pages numbered. The want of pagination is to be

regretted, as the numbers might have served

Want of Pagination. aa a cllie

to distinguishing top and bottom of

the page, and thus of determining the beginning and end of the formula

imprinted on it. The large number of the block-prints and the multiplicity

of the scripts contained in them open up another problem as to their object.

It would seem that there existed somewhere in the Takla Makan a kind

of library, or store of books, the locality of which seems to have been

discovered by native treasure seekers, being perhaps an ancient monas¬

tery. Moreover the existence, among the block-prints, of collective

books (such as comprised in the Sixth Set), which contain impressions

of texts in several scripts, seems to show that in that place there must

have been a collection of all the different kinds of blocks ; and that the

place, in fact, was a sort of printing establishment, for the production

and distribution of books of (religious) formulas among communities

or in localities, using different scripts, and perhaps speaking different

languages or dialects.

Prima facie , there are not less than nine different scripts employed

in the block-prints. Accordingly I have dis-

"^^^ScriHts^nd^^ tliem into nine sets. It is not im-

Formulas. probable that hereafter it may be shown that

some of the scripts are allied to, if not

identical with, one another ; I mean in this way that one may be the

kaligraphic counterpart of a current script. This may be the case,

perhaps, with the two scripts shown on Plate V, for they agree in their

number of lines. I believe also to have noticed, here and there, the same

symbol, in slightly modified forms, in different scripts. In order to

arrive at any definite and satisfactory conclusion on this head, a more

detailed and minute examination is necessary, for which the time

allowed me at present does not suffice, but for which I hope to have

leisure after my retirement from India. With my present informa¬

tion, it appears to me likely that the scripts of the First and Second

Sets, those of the Third and Fifth Sets, and those of the Fourth and

Seventh Sets are pairs the members of which have some more intimate

connection with each other. Further, it seems to me possible that the

juxtaposition of several formulas in the collective books of the

Sixth Set and elsewhere may lead to the recognition of some kind of

identity obtaining among them with reference to their purport. As to

the language, or perhaps the number of languages, hid in these scripts

54

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

Orientation of the

Texts and Scripts.

and formulas, of course, it is impossible to venture to express any

opinion, before some advance has been made in their decipherment.

Some of the block-prints are furnished with guards which show

in their ornaments a curious resemblance to a

Age and Conservation certain coin of one of the Urtuqis of Maridin.

of the Block-prints. . D1 , ¥ir £ , n

they may be seen on Jrlate IV, ngs. 1-9 ;

the coin also on Plate I, fig. 20. The coin is ascribed to 1232 A. D.,

see the Section on Coins, p. 31. If the resemblance is not desceptive,

it will fix the upper date of the block-prints in question. They could

not be older than the middle of the 13th century A.D. There is reason

to believe, however, that some of the block-prints must be several cen¬

turies older. That there is nothing in the physical conditions of Eastern

Turkistan to render such a long period of conservation impossible, I

have already remarked in the Introduction, p. xxviii.

The question of what is top and bottom, right and left, of the text,

or of the formulas composing the text, is a

puzzling one. The determination of it would

help to determine the further question of how

the script of the texts is to be read, whether from the left to the right

in the European fashion, or from the right to the left as in Semitic

writing, or from top to bottom as in Chinese. I have not, as yet, come

across any absolutely decisive evidence. In some books regularly

recurring partial impressions of formulas are met with, — cases in which

only a portion (one-half or one-third) of the formula, divided either hori¬

zontally or vertically or both ways, is met with. Want of sufficient

space on the page is always seen to be the reason for such partial im¬

pressions. In such cases it may very plausibly be argued that, when

the printer had not sufficient space to print the w'hole formula, he

would preferably print the initial portion of it on the available space.

On this assumption we should have an indication of what is the beginning

or the end of a formula. Thus let a b c d in the marginal diagram

represent such a complete formula, in which the lines

of writing run parallel with a b (as e.g. on Plate XII).

If a b g e, that is, the formula horizontally divided, be

the only portion printed, this may indicate that it is

the initial portion of the formula. Similarly if a f Ji d,

or the formula vertically divided, were only printed,

this would show that portion to be the initial one. If

further, both portions a b g e and ajhd were found regularly printed

in certain delimited places, we should know for certain that the portion

A contains the beginning of the formula, and that, its reading must

commence in the corner a, and proceed from a to b. It would still

a

f

d

9

h

1899.]

Section II. — Block Prints.

£ Y

remain uncertain whether the script run from the left (a) to the right

(b) in European fashion, or from the top (a) to the bottom (6) in

Chinese fashion. In other words, a might be either the upper left-hand

or the upper right-hand corner of the formula. Similarly the beginning

of a formula might happen to be found to lie in the portion B, in which

case the script would run from the right to the left, in Semitic fashion.

The two alternative possibilities, here explained, are those actually

observed by me in the case of the formulas of the Fourth and Seventh

and the formula of the Fifth Set respectively. The former seem to

run from the left to the right, the latter from the right to the left.

The weak point in this argument is not so much the fact that occa¬

sionally the opposite portions (eg cd and fbch ) of the formula are found

printed ; for this might be due to a careless misprint ; and the detailed

description of the various sets will show that misprints are by no

means uncommon. A far more serious difficulty is the uncertainty as to

whether the books were intended for reading at all. If they were not

intended for reading, but for some kind of mechanical use, the circum¬

stance of what particular portion of the formula was printed in order

to represent the whole of it is obviously of no moment. But on the

other hand, the regularity in printing a certain definite portion points

to method and design, such as one would not expect in the case of

printing for mere mechanical use. In the latter case one would expect

the portions AB, AD, BC, CD to occur promiscuously. It seems, no doubt,

certain that the disorderly books, above mentioned, such as No. VI of the

First Set, cannot have been intended for intelligent reading, but, on the

other hand, it is by no means certain that some other books may not have

been prepared with that object. Book No. II of the Seventh Set is a case

in point. The marginal diagram shows the arrangement of its imprints.

. The formula is indicated by

abcdef; it is repeated 7 times on

the two pages 4 and 5 ; the fourth

impresion is divided between page

f 4 and page 5, about one-half of

the formula standing on either of

5 those two pages. The reader is

supposed to read down the left-

6 hand columns which stand upright.

Having read down column I on

7 page 4, to the middle of the 4th

impression, he continues with the

other half of that impression in

column II on page 5. He then

i

_4

2

56

Dr. Hoernle — Antiquities from Central Asia. [Extra No. Ir

turns the book, and similarly reads the remaining (formerly right-hand,

but now also left-hand) columns III and IV. Such an orderly arrange¬

ment can hardly be explained on any other supposition than that of

being made with a view to intelligent reading. Occasionally also

anomalies are met with, the only satisfactory explanation of which seems

to point to a similar conclusion. One such anomaly will be found

discussed in the detailed description of book No. VI of the Fourth

Set. It is on considerations such as these, that I have provisionally

determined, and shown in the illustrative Plates, the top and bottom

of the texts of most of the nine sets of block-prints, as well as the

direction of the script of some of them, such as those of the Fourth and

Seventh Sets. I do not claim for these determinations more than a

provisional character. Very possibly a more minute and thorough

examination of the block-prints, than the limited time at present at my

command admits, may hereafter lead to more definite results.

The case, above discussed, of book No. II of the Seventh Set, is

instructive on another point, namely, whether

Orientation of the these block-print books are to be read from the

Books. left the right, beginning with the first page,

or from the right to the left, beginning with the last page or at the back

of the book, to speak from the European point of view. From the

diagram it will be seen that the reader first reads the pages (i. e.,

columns I and II) from the left to the right (and so on, throughout the

book), and then, turning the book right round, from the right to the

left (t.e., columns III and IV, and so on, throughout the book). It

would, therefore, appear that there is really no right or left, beginning

or end of the book, in the sense of the modern European practice. This

conclusion seems to be confirmed by the books Nos. I and II of the Third

Set, of which, to judge by the arrangement of the text (see the detailed

description) No. I must be read from the left to the right, while No. II

must be read from the right to the left. See also book No. V of the

Second Set.

The question on which side of the page the process of printing com¬

menced, whether on the right or left side, the

Orientation of the Qr 0£ is fairly easy to determine

Printing. many cases. When there is a broad, blank

margin on one side of the page, while the print runs up to its very edge

on the other side ; or when a column of print begins with a complete

impression of the formula on one side, and ends with a part-impression

on the other side, it is fairly certain that the printer commenced his

work on the former of the two sides. Books Nos. VI and VII of the

Fourth Set afford a good illustration of this conclusion. The point is of

Section II. — Bloch Prints.

57

The Question of

Genuineness.

1899.]

no great importance in itself ; but it may in some cases prove auxiliary

in determining the orientation of a script or text.

Considering the abundance of the block-prints and the mystery of

their scripts, it is not surprising that the

suspicion of forgery should suggest itself. It

suggested itself to me at an early stage of my

acquaintance with the Khotanese books ; and I am informed that it has

also suggested itself to some of the British Museum authorities and

others. But it was not till the summer of 1898 that the suspicion took

a more definite shape in a letter, dated the 29th June, 1898, which I

received from Mr. Backlund, Swedish Missionary in Kashghar, in

response to a request by me for information on the subject ; for at

one time, in the course of my examination of the block-prints my suspicion

had been much strengthened by the observation of the extreme want of

order in certain books. This result was subsequently neutralised by

the observation of the striking consistency of order in other books. It

became clear that, as I have already shown, both phenomena are quite

compatible with a general genuineness of the block-print books ; and in

fact, all the evidence that gradually accumulated has tended to confirm

that conclusion. Mr. Backlund’s account is as follows :

“ It is my duty to own that till quite recently I have scarcely taken

any interest in that old Khotan literature. In April last [1898], however,

Islam Akhun brought to me three copies, which, according to what he

told me, had been found in the neighbourhood of Aq Safil, buried under

sand in a hollow tree, together with other books of the same kind. Some

days before, he had sold two or three copies to Mr. Macartney also,4 but

nevertheless he urged me not to say a word to Mr. Macartney about my

acquisition. Upon my having a look at the books to discover whether they

were print or hand- written, he felt somewhat uneasy and whispered, “ “it

is astonishing how attentively he is looking at the books.” ” I offered him

less than the half of what he asked, and he not only handed over to me the

books without haggling, but also gave me into the bargain some old coins

he had with him. When he had gone out, one of our servants, entering

my room said, “ “ Sahib, I want to tell you that these books are not

so old as they are pretended to be. As I know how they are prepared,

I wish to inform you of it. When I lived in Khotan, I wished very

much to enter into the business, but was always shut out and could

even get no information about the books. At last I consulted my mother

4 These are probably included among the eight books which were sent to me

with a letter dated the 13th April, 1898. They were acquired from Islam Akhun,

but were stated by him to have been found in Kiang Tuz on the road to Cherchen.

J. i. 12

58 Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

about it ; and she advised me to try and find it out of a boy with

whom I was on very intimate terms, and who was the son of the headman

of this business. So, one day I asked him, how they got these books,

and he plainly told me that his father had the blocks prepared by a

cotton-printer, ” ” etc. Now it is evident that the servant might have

said all this from jealousy only, but I now determined to examine the

books with more critical eyes than before. Then the following facts

became clear to me immediately :

(1) The rich supply of books, which may be purchased at any

price we are pleased to put on them, although every European traveller

who has been in Khotan has taken a great interest in them, not men¬

tioning the Russian Consul and Mr. Macartney who have bought what

they have come across.

(2) The apparent freshness of them, as for instance —

(a) the sharp corners of the copper plates and nails which are

covered only with a very thin layer of rust ;

(b) no rust from the plates sticking to the paper under the plates;

(c) the corners of the books quite square (not round, as they

usually are in old books), and the edges recently out

though in such a manner as to make them look old ;

(d) although without proper covers, the outside leaves as well

as the leaves in general were well preserved, but one here

and there destroyed betwixt two fresh ones;b

(e) no yellow spots or marks of handling by readers, as usually

occur in old books ;

(/) the paper, though very ill-treated (burnt and smoky), still

strong almost as if it were new ;

(g) the paper exactly of the same kind, as prepared in Khotan

in the present day.

Now if these books are forgeries, must not there have been some

genuine ones, after which these are made ? Certainly, I think so,

especially in order to account for the characters. But I do not think

they took the pains to copy any text — they may have or they may have

not — of the original, but very likely put the letters in a preposterous way

to make it look like writing. As for the hand- written ones, I have no

particular opinion, as I have had no opportunity to examine them. But

I do not think that it is at all impossible that they should be forgeries.

You see I purchased the volumes I have spoken of — three of the longest

that have been sold — for a total of Rs. 7, and certaiuly I payed too

much. If they can get twice as much, very likely they would not

hesitate to prepare actual manuscript.”

6 The italics are Mr. Backlund’s.

Section II. — Block Prints.

59

1899.]

In a letter written by Mr. Backlund to Mr. Macartney, on tlie 8th

April, 1898, tbe day after he had purchased the above-mentioned three

block-print books, I find the following additional information: — “ If has

been communicated to me by a person, who is well acquainted with these

things [apparently the servant above referred to], that these books are

not old, but are continually made now-a-days also ; and he pretends to

know the printer also. The books are said to be prepared like this :

after being printed, the sheets are hung up in the chimney in order to

make them look old. They are now burnt in parts and covered with

soot. When they have assumed as dark a colour as seems to be suitable,

the soot is wiped off and the papers are nailed together into a book and

taken out into the desert, ’where they are buried in the sand. Having

remained there for some time they are “ discovered ” and brought out

into the market in order to — make fools of the Europeans. Examine

the paper in the books and you will find it quite of the same kind, as is

produced in Khotan now-a-days ; and the white spots in it here and

there point it out not to be of an ancient date.”

With regard to the three books, purchased by Mr. Backlund as

related above, he informed me in a subsequent letter, dated the 10th

October, 1898, that “ as he considered them useless, he handed them over

to an English traveller, Mr. R. P. Cobbold ; ” and that “ soon after having

got rid of them, a man offered him some very fresh prints, which ho

refused to take.” The books thus obtained by Mr Cobbold afterwards

passed into the possession of the British Museum, and I shall have occa^

sion to refer to them again.

I quote these letters so fully, in order that the case of the forgery-

theory may be stated quite fairly. To Islam Akhuti’s behaviour and the

servant’s denunciation too much force should not be attached. They

are nothing more but what may be expected under the circumstances.

The points enumerated by Mr. Backlund are those deserving considera¬

tion. And here it should be noted, in the first place, that they only

refer to Khotanese block-prints, not to manuscripts, and secondly, that

they are based on a very limited number of specimens. Mr. Backlund

admits — what indeed is obvious — that forgery presupposes the existence

of a genuine original which was imitated. The suggestion is that a

distorted imitation of this original was made purposely, and that that fact

accounts for the mystery of the scripts. This does not seem a plausible

hypothesis. Ho intelligible original, such as the suggestion assumes to

have existed, has been produced ; if it existed, the finder, surety, would

have disposed of it first, and when his genuine stock was exhausted, he

might then have had recourse to forgery to replenish his stock in trade.

Something of this kind, indeed, as I imagine and as I shall presently

60

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

show, has probably actually happened ; but not in the way required by

the hypothesis referred to. Moreover, as my detailed description of the

block-prints shows the varieties of the (ex hypothesi) forged script are so

numerous and so intricate as to require the allowance of a much longer

time for their elaboration, than has actually passed since forgeries

can have commenced, at most about ten years ago. The trade in

forged prints could only have arisen with the advent of modern Euro¬

pean travellers. The earliest of these is General Prjevalski who

visited Khotan in 1885, and at that time these books were unknown

and unthought of. The first objection, mentioned by Mr. Backlund

refers to the cheap price of the books. This is a point which may be

argued either way, and is usually considered to speak rather in

favour of genuineness. Mr. Backlund obtained his three books for Rs. 7 ;

but for some block-prints in the British Collection a rather good price

has been paid. For the book G. 9 (Eighth Set) Rs. 40 were paid ; for

the book G. 8 (No. VI in the Seventh Set) Rs. 45 ; for the two books

in M. 5, Rs. 40 ; for the two books in M. 6, purchased from Badruddin,

Rs. 40; for the four books in M. 6, purchased from Mr. Hogberg

together with a lot of antiques, Rs. 200. On the other hand, for the

two sets of nine books in M. 7 and eight books in M. 8, only Rs. 40,

each set, were paid ; and for the two books in M. 4 ( plus sundry

antiques) and the two books in M. 9 (also plus sundry antiques) even

only Rs. 11-3 and Rs. 20 respectively. The fact is that latterly (early in

1898) when suspicion had once been aroused regarding the genuineness of

these books, which tended to interfere with their saleableness, the dealers

found it advisable to lower their prices. This is a question of demand

and supply, and has little direct bearing on that of genuineness. It is

quite possible that a large store of genuine books may have been dis¬

covered somewhere in the desert.

The second objection refers to the supposed freshness of the books.

I have examined 44 books and my observations do not altogether agree

with those of Mr. Backlund. There are distinct marks of old rust on

the guards and beneath them in the case of some books ; in others the

corners are by no means “ quite square,” but irregular aud even round ;

the leaves of some books (outside as well as inside) are in a very

damaged condition and rotten, and show the dirty signs of having been

handled ; some books are printed on a kind of paper which is quite

unknown in Khotan.

The probability seems to be that latterly when the store of genuine

old books gave out, an attempt was made to produce new ones by

imitating some of the old genuine ones. The commencement of this

attempt would seem to fall in 1897; and the books offered to Mr.

Section II. — Bloch Prints .

61

1899.]

Backlund may have been part of the result of this attempt. Mr. Bendall

wrote to me on the 15th July, 1898, “I have been comparing your

pamphlet about the xylographs from Central Asia6 with a block-print

recently acquired by the British Museum from Lieut. Cobhold. What

is curious is that it is a duplicate of the hook figured on your first

Plate, but does not contain the writing between the two columns of

print to which you call attention.” This observation of Mr. Bendall

very possibly gives the key to the situation. If there exist any

forgeries, they are, in all probability, duplicates of genuine books that

have been discovered. The preparation of a duplicate is probably well

within the capability of a modern Khotanese forger, but the hypothesis

that he is capable of inventing not only one hut several scripts, and

of intricate, but self-consistent systems of their arrangement in books,

and finally of binding them after a method, quite unknown in Khotan

at the present day, contains more elements of improbability than the

hypothesis of the genuinenesses of the hooks.

The manufacture of duplicate block-prints postulates the existence

of old blocks from which new ones may have been prepared, and from

which (or from their new facsimiles) the modern reprints (if there

are any) must have been made. I have shown in the description of

the First and Second Sets, how utterly improbable it is that the blocks

of type can have been invented by the forger. The overwhelming

probability is that sets of old blocks of type have been discovered in the

Takla Makan, and from these reprints may have been made. But

moreover, actual old books printed from those blocks and representing

each of the nine Sets must have been found. For the systems of print¬

ing and binding which are used in the books are unknown in Khotan

in the present day, and imitations could not have been made, unless

models had been found.7 Add to this not only that most of the books,

though printed (as I believe) on Khotanese paper, are printed on

varieties of it (viz., Ilia b c) which are not known in Khotan at the

present day ; but also that there are others (as those in the Eighth and

Ninth Sets) which are printed on paper of a kind which is not Khotanese

at all. That some of the block-print books are printed on paper of the

variety Hid is quite true ; but this fact, by itself, does not prove forgery ;

for it cannot be doubted (considering Oriental conservative habits)

6 Published in the Proceedings of the Asiatic Society of Bengal, for 1898.

7 An alternative hypothesis would be that no blocks have been found, but only

books ; and that from these books new blocks have been prepared, and then employed

to print new books. The prints, however, as shown by measurements, are so

accurate facsimiles, that considering the inveterate inaccuracy of Orientals this

hypothesis of the imitation of new blocks from old prints seems excluded.

62

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

that the particular process of paper manufacture which is still followed

in Khotan may have been in vogue there for centuries before. At all

events, it caunot well be supposed that those books, which are printed

on old paper of a kind never known, or no more known in Khotan, are

modern Khotanese forgeries. If they are modern forgeries, they must have

been forged somewhere else than Khotan ; and this complicates the theory

of forgery with additional improbabilities. Further, some of the books,

admittedly or probably written on paper of Khotanese manufacture,

exhibit peculiarities which it may safely be said, would not have occurred

to a forger to introduce. I refer, for example, to the sketches of heads,

which are found in books No. I of the Fourth Set and No. V of the

Seventh Set,8 to the occurrence of the recensions Iq and lh in book No.

VIII of the Fifth Set, and to the Pothi with its entirely Indian arrange¬

ment. Such books cannot well be forgeries.

Further, forgeries may be admitted to be quite possible in the case

of block-prints, in the reprinting of which from genuine old blocks

there is no serious difficulty. But it is different with manuscripts ; and

let it be noted, that there are not only fragments of manuscripts, but

whole books — some of them fairly large books — of manuscript. The

difficulties of forging these would be enormous. In this case there are

no duplicates. There are, indeed, a fair number of them in the collec¬

tion ; but they are all different from one another. It would mean that

they had all been forged, within a comparatively short time, from no

models whatsoever. Some are written on paper which is not Khotanese

at all ; others are on paper, similar to that of some of the block-prints,

but of a variety now obsolete {viz. Ilia). Some are bound in the

Indian fashion of a Pothi ; others in the Khotanese fashion with copper

pegs or twists of paper. These manuscripts cannot be forgeries ; and

pro tanto they make against the hypothesis of forgery in the case of the

block-prints.

The mystery of the scripts — so many, and so intricately arranged —

is, no doubt a difficulty. But to solve it by the hypothesis of forgery

is only to substitute one riddle, and a harder one, for another. How

can Islam Akhun and his comparatively illiterate confederates be credited

with the no mean ingenuity necessary for excogitating them ? More¬

over the riddle of one of the scripts, which occurs in two of the manu-

8 These sketches are not easily observable. The books were some months in

my hands, before I discovered them, and I did so only on carefully examining them

page by page. Their existence does not appear to have been known either to

Mr. Macartney or to Islam Akhun who sold them to him. It does not seem pro¬

bable that a forger would have omitted to draw the buyer’s attention to the exist¬

ence of such a valuable peculiarity in his own handiwork.

1899.]

Section II. — Block Prints.

G3

script books written on Khotanese paper (variety Ilia), has been

solved. In January 1898 I showed these books to Sir Charles J. Lyall,

and he agreed with me that the script seemed to resemble Pahlavi and to

be in verse. In December last, when I had an opportunity of showing

them to Dr. Aurel Stein, who has made Iranian scripts and languages

a special study, he at once recognized the Pahlavi script in verse. He

even read some portions of it, though, of course, as will be readily

understood by those who know the difficulties of reading unknown texts

in Pahlavi, it was not possible for him, at such short notice, to determine

what the purport of the text might be.

Finally to add a minor point, book No. VI of the Second Set, is a

mere fragment. One cannot easily conceive why a forger should sell a

portion of a forged book of a kind, of which he could with comparative

ease fabricate a large number of complete copies ; while it is perfectly

natural that he should dispose of a genuine old book, even if he had

found or secured only a portion of it.

To sum up, the conclusion to which, with the present information,

I have come, is that the scripts are genuine ; and that most, if not all,

of the block-prints in the Collection also are genuine antiquities ; and

that if any are forgeries, they can only be duplicates of others which are

genuine, and must be found among the books of M. 8 and M. 9 which

are written on the IHd variety of Khotanese paper. By duplicates, I

do not mean such in point of size or variety of paper, but with reference

to the arrangement of the impressions of the block, or blocks, on the

pages. The determination of whether or not there are any such

duplicates in the British Collection, must remain over for a future

opportunity of examination of the block- prints.

In addition to the block-printed text, two books (No. IV of the

Written Legends. Fi.rst Set and No' 111 of the Secoad Set) con‘

tain additional small legends, inserted in blank

intervals between the repetitions of the text. Their letters are larger

than, and their direction sometimes different from, that of the surround¬

ing text. They have every appearance of not being printed but written

by hand. Their shape is irregular, and their ink is darker and does not

look as if imprinted from type.

On the whole the block-print books are in a fairly good state of

preservation. Some of them are much torn

or otherwise damaged. Many of them bear

stains of oil or other fatty substance ; some

also seem to have been exposed to the action of fire or water, as their

leaves are found more or less strongly singed or rotten.

State of

Preservation.

64

Dr. Hoernle — Antiquities from Central Asia, [Extra Nd. 1,

First Set. (Plates V-YIII, XI and XIII.)

This set comprises eiglit copies, namely, one pothi and seven books.

Its text is also found on three pages of book No. II of the Second Set ;

and some portions of the text of it are also found on one page of book

No. VIII of the Fifth Set and in book No. Ill of the Sixth Set.

The text consists of five portions, which comprise 5, 2, 7, 2 and

4 lines respectively, and which I shall distinguish as the formulas A, B,

C, D and E. The total text, therefore, comprises 20 lines of type.

Each line appears to contain from 10 to 13 letters.

The entire text of 20 lines, with the five formulas, arranged in the

order above given, is found in four books (Nos. I, III, IV, V). One book

(No. VI) and the pothi (No. II) contain only the formulas A, C and E ;

with this difference, however, that in book No. VI the three formulas

are arranged connectedly in the same order as in Nos. I, III, IV, V, only

leaving blank intervals in the place of B and D ; while the pothi gives

the three formulas A, C and E independently of one another, in no con¬

nected series and on different pages. Of the remaining two books, one

(No. VII) gives only a portion of formula A, viz., lines 1, 2, 4 and 5,

omitting line 3 ; while the other (No. VIII) gives only a portion of

formula C, viz., lines 9, 11, 13 and 14, omitting lines 8, 10 and 12. In

book No. Ill of the Sixth Set the same three formulas are found, but

not in any connected series ; viz., (1) the formula A, mutilated as in

No. VII, (2) the formula C, as in No. VIII, (3) a portion of formula E,

viz., lines 17, 18, 19, omitting line 20. The two formulas B (lines 6

and 7) and D (lines 15 and 16) have never been found by me separately

from the entire text. These formulas, therefore, are found only in the

books Nos. I, III, IV, V.

Accordingly the text of the First Set exists in the following eight

forms or recensions, denoted by la, lb, Ic, Id, le, If, I g and Ih.

(1) The full text (la) of 20 lines, consisting of the formulas A, B,

C, D and E, in a connected series. This is found in books Nos. I, III,

IV, V ; also in book No. II of the Second Set.

(2) The shorter text of 16 lines (16), consisting of the formulas

A, C and E, with proportionate blank spaces for B and D ; the whole in

a connected series (as in la). This is found in book No. VI.

(3) The three formulas A (Ic), C (Id), and E (Ie), given separately,

and forming no connected whole. In pothi, No. II, also in book No. VI.

(4) The formula A (If), in a mutilated form, viz., lines 1, 2, 4, 5.

In book VII ; also in book No. Ill of the Sixth Set.

(5) The formula C (Ljf),in a mutilated form, viz., lines 9, 11, 13, 14.

In book No. VIII ; also in book No. VIII of the Fifth Set and in book

No. Ill of the Sixth Set.

Section II. — Block Prints .

65

1899.]

(6) The formula E (I/?-), in a mutilated form, viz., lines 17, 18, 19.

In book No. Ill of the Sixth Set.

I have carefully measured these various recensions of the text. The

measurements were made vertically, horizontally and diagonally across

the prints. I measured, in this manner, the whole text in all its forms,

as well as groups of lines, single lines, half-lines, groups of letters

and single letters. The result was invariably the same ; the corres¬

ponding measures in the several books exactly agree, in whatever

variation they may be taken. Thus measuring the recension I a, from

the top of the last letter of the first line to the bottom of the last letter

in the last line (both on the left-hand side of the column, see Plate

V), the distance is exactly 6 inches. Measuring similarly, the

distances in the formulas A, C, and E are If, 1t9q and ly1^ inches

respectively; and again measuring similarly, the width of each of

the intervals between A and C, and between C and E (within which

the two formulas B and D are placed) is f of an inch. Measuring the

recension 16, exactly the same result is obtained. This recension, as has

been already explained, consists of the formulae A, C and E, with

blank spaces for B and D. The distance from the top of A to the

bottom of E, including the blank spaces, is exactly 6 inches ; the width

of each blank space is § of an inch ; and the widths of A, C and E

are If, 1t9q and ly1-^ inches respectively. The widths of the latter three

formulas or the three recensions I c, Id, Ie (see Plates VI and VII),

when they occur separately in the Pothl (No. II), are precisely the

same ; and the same is the case with them in their mutilated forms

If, I g, lh (see Plates XI and XIII). Thus in the recension If, a blank

space is left, for the omitted line 3, between the lines 2 and 4 ; and

measuring from the top of line 1 to the bottom of line 5, the distance

is, as before, exactly If inches. It is evident, therefore, from these

comparative measurements that the xylographs of the first set were

all printed either from the very same block, or from a number of

blocks with facsimile type.

It can easily be shown that the latter alternative is the true one,

and that in all probability eight blocks have been employed in printing

the xylographs of the first set. A block was prepared in this wise.

From the smoothened surface of a rectangular (oblong) piece of

wood the type was cut out in relievo by counter-sinking the background.

Along the edge of the piece of wood a thin ridge was also left in relievo,

enclosing the type and the counter-sunk background on all four sides.

In order to take an impression the surface of the type was inked.

Of course, the ridge was also inked, but this appears to have been done

very imperfectly, for in no case did it give more than a very intermittent

J. I. 13

66

Dr. Hoernle — Antiquities from Central Asia , [Extra No. 1,

impression, and in most cases it gave no distinct impression at all.

Traces of the enclosing lines, made by the ridge, may be seen on Plates

V, VII, XI and XIII. It is obvious that these enclosing lines afford a

ready means for identifying a block. Sufficient of their traces remain

to render it nearly certain that there was a separate block for each

of the eight recensions of the text.

The blocks used for recensions la and 16 had the same dimensions.

They were long, narrow slabs of woods, measuring 6| by Iff inches.

There were no blocks of twice that width bearing two columns of

type. This is proved by the fact, that when two impressions are

seen side by side, their edges frequently touch or even overlap one

another, showing that the impressions were taken separately one after

the other, and not very carefully. Moreover occasionally when the im¬

pressions were taken wider apart, two parallel enclosing lines may be

seen between the inner margins of the two prints. On the other hand,

it is not probable that both recensions la and 16 can have been printed

off the same block. For the blank spaces (for lines 6, 7, 15, 16) in 16 are

quite clean : smudges would have been unavoidable from the old inked

surface, even if the omitted lines had afterwards been left uninked.

Moreover, though the surface of the blank spaces is clean, the enclosing

lines can occasionally be seen continuing on both sides, and thus

showing that the entire surface of the block had been inked. It follows

that for recension 16 a separate block must have been used, in which

the surfaces of the two blank spaces had been counter-sunk in order to

prevent their being inked. Further it is not probable that the recension

16 can have been printed by using in combination three smaller blocks

of type, containing the formulas A, C and E respectively. For (1) the

width of the blank interval is always exactly the same (f of an inch),

(2) the enclosing lines right and left run perfectly straight, (3) there is

never any trace of any top and bottom enclosing lines of the three blocks

between the lines of type. These three facts (especially in combination)

seem quite incompatible with the use of three blocks to print one text.

There were three blocks, one for each of the recensions Ic, Id

and Ie. They must have measured about If x Iff, 2J x Iff, and If x Iff

respectively, as may be calculated from the slight traces of the enclosing

lines discernible in a few places (see Plates VII, XI, XIII). The blocks

for Ic and le must have been of the same or very nearly the same size.

The recension Id, ( i.e ., formula C) is printed six times on a page of the

pothi (see Plate VI), being arranged in two columns of three impres¬

sions each. That the page was not printed off two blocks, each containing

a whole column of type, or off one block containing a double column of

type, is evident from the fact that the six impressions do not keep in

1899.]

Section II. — Block Prints.

67

straight line, but approach or overlap one another, both horizontally

and vertically. The same remarks apply to recensions I c (A) and

Ie (E), each of which is printed eight times on a page of the Pothi

(see Plates VI and VII).

The two recensions If and I g must have been printed from blocks of

the same size as those for the recensions Ic and Id. The two pairs of

recensions (1/ and Ic, I g and Id) cannot have been printed from identical

blocks ; for the same reasons which (as explained above) show that the

two texts la and I b cannot have been printed off an identical block.

The case is not quite so clear with regard to the recension I h (formula

E) ; for I have noticed occasionally smudges on the blank space cor¬

responding to the omitted line 20. They look like very indistinct traces

of the letters of that line, suggesting that its type existed on the block

but had not been inked. In the blocks for 1/ and I g, the type of the

omitted lines does not seem to have existed, the whole space correspond¬

ing to those lines being counter-sunk, excepting the ridge along the

edges, traces of which ridge are still occasionally discernible (see

Plate XI).

For the formulas B and D (that is, the lines 6 and 7, 15 and 16)

there do not seem to have existed any separate blocks. So far as the

evidence, at present available, goes, those two formulas were never

printed separately, but only existed on the block for recension la.

The PSthI is undoubtedly a genuine ancient relic. It possesses

every mark of antiquity in point of general appearance and condition.

It is unique in its form of an Indian 'potlii. Its paper, which is hard,

rigid, brittle and discolored, and its print which is faded, suggest

considerable antiquity. In point of material and texture its paper is

very similar to, if not identical with, the paper of the variety III5,

on which many of the books are written, but it differs distinctly

in colour, being more of a dirty greyish-brown, than of the

dirty yellowish-brown of the books. With reference to this Pothi

Sir A. 0. Talbot, in his demi-official letter, No. 5972, dated the

23rd October 1897, writes that “ it might be of interest to note that the

book enclosed between the rough wooden covers bears a strong re¬

semblance to the religious manuscripts still used in the Hemis and

other large monasteries of Ladakh ; and that among the metal objects

sent 9 is what seems to be an old iron arrow-head, very like those with

which the arrows in the treasure-rooms at Hemis are tipped. Possibly

the excavation was made from the site of some former Buddhist

monastery of which, according to Remusat, many must have existed in,

and around, the Takla Makan.” The evident antiquity of the Pothi is

9 This arrow-head as well as the Pothi were in the consignment M, 4,

G8 Dr. Hoernle — Antiquities from Central Asia. [Extra No. I,

a point of great importance ; for it is a guarantee of the genuineness of

the text. Whatever degree of suspicion may attach to some of the

books, they can only be forgeries in a modified sense. Their paper and

the actual print may be modern, but their impressions must have been

taken from ancient blocks. For, as I have shown, the blocks from

which the books and the pSthi are printed, show identical sizes and

facsimile types. It is almost demonstrable, therefore, that a set of

ancient blocks of type must have been found, from which the books, if

any are really modern fabrications, have been printed. The three

blocks (for recensions Ic, Id, Ie, or the formulas A, C, E) from which the

PSthi was printed, must certainly have been found. It may be

suggested that, with the help of these three blocks, the blocks for the

other recensions might have been fabricated. But this would not account

for the existence of the formulas B and D (lines 6, 7 and 15, 16) in

recension la. It is very improbable that a forger, though he might have

omitted portions of an existing text, would have gone beyond his pattern

and invented new lines of type. The probabilities, therefore, decidedly

are for the genuineness of the block of recension la. The preparation

of facsimile blocks, from existing patterns, is not at all beyond the

capabilities of a clever imitator ; and the genuineness of the blocks for

the recensions 16, If, I g and I h, which are only differentiated from those

for la, Ic, Id and Ie by the omission of certain lines of type, might,

therefore, be questioned ; but the occurrence of the recensions I g and I h

on one page of book No. VIII of the Fifth Set (see Plate XI) renders the

hypothesis extremely improbable. Such a solitary and casual insertion

of an alien text in a book entirely devoted to a different text would

hardly have occurred to a forger. Moreover the state of preservation

of that book seems to stamp it as genuinely antique. On this point,

however, further evidence is required. If once the writing is decip¬

hered, and its purport understood, that knowledge may very possibly

decide the question of genuineness. If it should be found that by the

omission of a portion of it, the text is rendered unintelligible, that

result might seem to prove that the blocks for the mutilated texts 16,

If, I g, I h are the work of an ignorant forger ; for, at the present day,

neither the writing nor the language of these block-prints is under¬

stood in Khotan. On the other hand, it must be remembered that

there is good reason to suppose that some of the books were not intend¬

ed for intelligent reading, but merely for mechanical use.

As regards the determination of the question of what is the

beginning and the end of the text, there is some indication given by

the arrangement of the text in Book No. I. This book shows two

columns on each page (see Woodcut No. 11), each column consisting

1899.]

Section II. — Block Prints.

69

of two impressions of the full text (or recension la). These impressions

are invariably10 placed so as to turn the same side towards the upper

and lower edges of the book. Ifc follows that that side of the impression

( as shown in Plate V ) must be its top or the beginning of the text ;

and that the feet, or ends, of the two impressions meet in the middle of

each column and of the page. For it is natural to assume, that

the reader was intended to commence reading at the top of the

page, and not in its middle. Of course, on the supposition of a forgery,

this conclusion would loose much of its force, as an ignorant forger

might by chance have misplaced the impressions ; but the peculiar

placement of the imprints is so regular as to render such an hypothesis

very improbable. Moreover there are other indications, such as the

texture of the paper (see the General Remarks on paper), which make

against Book No. I being a forgery.

No. I. Book. (Plate V.)

Belongs to M. 6. Acquired from the Rev. Mr. Hogberg. Size,

12x9". Number of forms, 18; but first form is incomplete, the first

blank leaf is missing : print accordingly commences on second page

(properly fourth page). Moreover second and penultimate forms are

double, consisting each of two folded sheets, one placed within the

other, and, therefore, having each four leaves or eight pages. Accord¬

ingly number of leaves, 39. Paper, variety Ilia ; fairly clean. Riveted

with three copper pegs.

Contains recension la, printed in two columns on each page ; each

column consisting of two impressions, placed foot to foot, the upper

one being complete (20 lines), the lower, more or less incomplete (a3 a

rule 15 or 16 lines) owing to want of space. (See Woodcut No. 11).

The foot-to-foot arrangement of the text is almost invariable. There

are only four exceptional pages, on which it stands head to foot. The

two varieties of arrangement may be represented thus : —

The second variety is obviously due to mere carelessness on the

part of the printer. The first variety, which occurs 62 times, is clearly

10 There are only a very few, apparently accidental, exceptions, which are noted

in the detailed description.

70

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

intentional. Its object evidently is to make it possible to read the

book in the way previously explained in the General Remarks. The

regularity of the arrangement seems to indicate that this book was

really intended to be properly read. If it had been merely intended for

mechanical use by turning the leaves, there would have been no necessity

for observing any such strict regularity. It may be further noted that

discounting the four exceptional and erroneous pages, one end of the

text (indicated by glii in the above diagrams) is always placed in

the middle of the page, while the other end is invariably found at the

top of the page, in whichever way the book is held. This circumstance

seems to prove clearly, which line of the text must be considered its

beginning.

The text is repeated four times on every page. There are only two

exceptions, viz., pages 48 and 63. On page 48 there are only two im¬

pressions, while page 63 has only one. The remainder of the space is

occupied with legends in an apparently different alphabet, but which

may also be only a written or “ current ” form of the printed one.

They are shown on Plate V, and are evidently not printed from a block,

but written by hand.

No. II. P5thi.

Belongs to M. 4. Brought from Khotan, together with No. VI

and other objects ; the whole purchased for Rs. 11-3-2. Size, 8Jx4J

inches.11 Number of forms 45. Leaves of a curious, bottle-shaped

form, see Plates VI-VIII; reminding one of the manuscript book

found under the skull (see Introduction, pp. xxi). Bound, in the

Indian fashion, between two rectangular (not bottle-shaped) blocks

of wood, measuring 8Jx4|xl inches, and rough and uneven on the

outer, but planed on the inner surfaces ; and exceedingly dry and light

of weight. Riveted like an Indian copper- plate grant, on the left-hand,

narrow side of the oblong, by means of one copper peg, which passes

through the “ neck ” of the bottle-shaped leaves. Paper, of a dirty

greyish-brown color, and hard, stiff, and brittle and in many leaves

badly fractured ; also with many fatty stains and occasional burns. The

whole appearance very suggestive of genuine antiquity.

Contains recensions Ic, Id and Ie, printed separately on different

pages, and arranged in two columns, so that there are six impressions of

recension Id (formula C), and eight impressions each of recensions Ic and

Ie (formulas Aand E) on a page. Recension Id (C) occtfrs most frequently ;

viz. , on forms 1, 2,6, 7, 9-12, 13 (pp. 2, 3, 4), 14-18, 20-24, 25 (pp. 1-3),

11 The middle of the pages of the pothi slightly projects beyond the edges of

the blocks.

Section II. — Bloch Prints.

71

1899.]

27, 28, 29 (pp. 2, 3), 30, 35, 36, 38-42, 44 and 45. Recension Ic (A)

conies next, on forms 3, 4, 8, 19, 25 (p. 4), 26, 29 (pp. 1, 4), 31-34 and 43.

Recension Ie (E) only occurs on the forms 5, 13 (p. 1), and 37.

As a rule, the lines of print run parallel with the narrow side, and

accordingly the columns run parallel with the longer side of the page

(see Plate YI). There is only one exception in which the lines of print

run parallel with the longer side, and accordingly the columns (which in

this case are four in number) with the narrow side of the page (see

Plate VII) ; this is the first page of the 33rd leaf. On two pages there

is an altogether exceptional arrangement, which may be shown thus :12- —

These three exceptional pages would seem to have been trial pages,

to find out the best way of disposing the impressions on the page.

As a rule the columns of the text are arranged so that they stand

turned towards the inside ( i.e ., the neck-like, peg-hole side of the

page), either head to head or foot to foot. There are altogether 86

pairs of printed pages, 43 pairs inside and 43 pairs outside the “ forms.”

Out of these 86 pairs, 78 have the columns of text thus arranged,

i.e., head to head or foot to foot. In the remaining eight pairs, the foot

of the column on one page adjoins the head of the column on the other

page ; and the arrangement on these exceptional pairs is clearly due

to carelessness on the part of the printer. Again among the 78 pairs

which have the normal arrangement, 37 have the heads of their

columns turned inside, i.e., towards the peg-hole side, while 41 have

the feet of their columns turned inside.13 The pairs are nearly equally

ccc denotes recension Ic, and eee denotes recension Ie.

IS In terras of “ forms ” the distribution stands thus. There are 45 forms. In

2 forms (first and last) only one page is printed. Of the remaining 43 forms, 14

have the head of the text turned inside ; 17 have it outside ; of 6 others, 2 have

the head inside on their interior side, but outside on their exterior side, and 4 have

the head outside on the interior, but inside on the exterior ; of the remaining 4

forms, 2 have the head inside on the interior and no order on the exterior, while 1 has

the head outside on the exterior, but no order on the interior side, and 1 is exactly

the reverse of the last-mentioned; (14 + 17 + 6 + 4 = 41). Obviously the last-men¬

tioned 10 forms (6 + 4) might be folded the other way, and then would show the

exact reverse of their present condition. The remaining two forms have no order

either on the interior or exterior side (41 + 2 = 43).

72

Dr. Hoerule — Antiquities from Central Asia. [Extra No. 1,

divided between the two systems. But what is puzzling is that the

pairs of these two systems, in their present succession, do not follow in

any intelligible order (e g., first the 37, and afterwards the 41, or vice

versa), but they succeed one another pell-mell. If either system had been

followed throughout, or if the whole of the pairs of one system had

followed the whole of the pairs of the other system, an intelligible order

of reading would have resulted ; one might have read first one set of

alternate pages, and then turning the book right round, one would have

been able to read the other set of alternate pages. As it is, the condition

of the book suggests that it was not really intended to be read intelli¬

gently, but to have its leaves turned mechanically. As the leaves,

however, are not numbered, and their serial succession is not fixed, it

is possible that the leaves do not now stand in their original order.

The latter might have been (say) first 41, then 37, and the break of

uniformity might have been due to a lapsus on the part of the printer.

No. III. Book.

Belongs to M. 6. Acquired from the Rev. Mr. Hogberg. Size,

8 x 5| inches. Number of forms 22. Paper, variety III6. One leaf

(23rd) damaged by two holes. Extensive fatty stains on many pages.

Riveted with three copper pegs.

Contains recension la, printed in two columns on each page and

standing upright and reversed on alternate pages ; thus upright on pages

4 and 6, but reversed on pages 5 and 7, and so forth.

No. IV. Book.

Belongs to M. 5. Size, llfxS". Number of forms, 30, all cut

into separate leaves. Paper, variety III b. Twenty leaves, more or

less damaged by holes, evidently due to singeing. Many pages soiled

with extensive fatty stains. Riveted with three copper pegs.

Contains recension la, printed in two columns on each page, with

a wide interval ; each column consisting of two impressions ; accord¬

ingly the full text, four times on each page ; but reversed on alter¬

nate pages ; see Woodcut No. 11. The arrangement of the columns is

similar to that in No. II, Pothi. As a rule, they stand either head

to head, or foot to foot, on any pair of juxtaposed pages. There

are altogether 57 such pairs. Taking their present order of suc¬

cession, on 19 of them the columns stand head to head, and on 29,

foot to foot ; while on the remaining 9 this arrangement is not

observed, the position being head to foot. Seeing, however, that the

forms are all cut into separate leaves, those leaves . on which the

exceptional nine pairs stand, may be easily arranged so as to preserve the

1899.]

Section II. — Block Prints.

73

regular arrangement. If this is done, we obtain 22 pairs head to head,

and 35 pairs foot to foot (total 57). Each of these two sets may easily

be read by observing the principle, previously explained, of reading by

alternate pages ; and the break of uniformity between the two sets may

be due to inadvertence on the part of the printer.

Within the intervals, in the centre of the page, additional small

legends are inserted, consisting of 4 to 7 letters, double the size of the

text, and running at right angles to the latter. They are not seen on

every page, but as a rule only on the two outside pages of each form ;

thus on pages 5 and 8, 9 and 12, and so forth. On eight forms (viz., 5,

8, 14, 16, 21, 24, 26) they occur on the two inside pages ; thus on pages

18 and 19, 30 and 31, and so forth. In one form, the 22nd, they are

omitted altogether. In the initial and final forms (on which three

pages are blank), of course, they occur only once ; viz., on pages

4 and 117.

No. Y. Book.

Belongs to M. 3. Purchased in July 1897 from Badruddin, an

Afghan merchant in Kliotan. Size, 11 J by 8 inches. Forms all cut

through at the back with a blunt instrument ; hence all leaves separate,

numbering 39 ; initial leaf missing. Paper, variety III b. Extensive fatty

stains on most pages ; many damaged by burns. Bound with

three twists of paper.

Contains recension la, printed on each page in one horizontal

column above and three vertical columns below, making four im¬

pressions on each page (see Woodcut No. 12). The three vertical

columns always stand in the same position towards one another, but

vary in their position towards the horizontal column, having sometimes

their head, sometimes their foot towards it. The position of the

horizontal column varies in having sometimes its head, sometimes its

foot turned to the outer margin of the book. The whole yields a great

variety of arrangements, in the choice of which no order is discernible.

It is impossible, therefore, to read the book in any intelligent order,

and it would seem to be intended for mere mechanical use.

No. YI. Book.

Belongs to M. 4. Brought from Khotan, together with No. II,

q. v. Size, Il§x44". Number of forms, 31. One blank leaf is torn

off, and missing ; hence only 63 leaves. Total of printed pages 126

instead of 124 ; because, contrary to the usual order of these books,

the print commences on the 3rd page instead of the 4th page ; simil¬

arly printing continues to the ante-penultimate page, instead of

stopping,, as usual, on the ante-ante-penultimate page. Paper, variety

J. i. 14

74 t)r. Hoernle — Antiquities from Central Asia. [Extra No. 1,

Ilia. Extensively soiled with fatty stains, and damaged by bums.

Bound with three twists of paper.

Contains recension 16 ; also recensions I r, I d, Ie, printed in two

columns on every page. The arrangement of the text of these recensions

on the pages is of the wildest kind ; there is no discernible order what¬

soever. The following diagrams of 19 different arrangements, which

I have noticed, will give a good idea of the disorder. Recension I c

(formula A) is indicated by aaa, I d (C) by ccc ; Ie (E) by eee ; and 16

by the corresponding (bracketed) combinations of these three notations.

The latter are represented in the same direction in which the recen¬

sions are printed in the book.

I. (pp. 1, 2.) 14 II. (pp. 3, 6.)

^ odd aaa

odd ccc

{^vvv eee

!aaa aaa

ccc ccc

IV. (pp. 5,8, 9, 15,

III. (p. 4.) 18,19,22, 24, 25,

27, 35, etc.)

VIII. (p. 13.)

IX. (p. 14.)

CD <D <D

<D <D CD

CD <D <D

CD <D CD

CD CD CD

CD

CD CD

%

CD CD

03 CD

X. (p. 28.)

co

CO

OS

co

CO

XI. (p. 29.)

{

ddd

000

VVV

aaa

ccc

8

e

a

8

XII. (p. 30.)

Q

a

£

&

a

Odd

ooo

vvv

odd

ooo

The numbers refer to printed pages. Thus printed page 1 = page 3 of the book.

1899.] Section II, — Block Prints,

XIII. (pp. 31, 34.) XIV. (pp, 32, 33.) XV. (p. 83.)

75

f aaa 999

I

CCC 000

I

(_ eee von

I

aaa aaa

ccc ccc

f 999 aaa

i

-{ 000 ccc

I

(jt>»u eee

aaa aaa

ccc ccc

{

a

<~>

a

<^

Ci

cs

Cl

eee J

XVII. (p. 114.) XVIII. (p. 115.) XIX. (p. 116.)

CD

Cl

o

Cl

CCC

eee

999 |

000 J-

VDV

Cl

ci

<*>

Cl

CCC

eee

999

000 ^

VDV

Si

»

S3

a

VDV

999

XVI. (p. 113.)

aaa

<T>

01

Cb

<*>

ct>

<n

Cfc

Ol

Ctl

Ct>

cn

0>

a»

Fourteen of these variations occur within the first 36 printed pages

alone. It is obvious that with such a planless arrangement of the texts

no proper reading of the book is possible. It can have been intended

only for mechanical use.

No. VII. Book.

Belongs to M. 7. Acquired by Mr. Macartney, together with eight

other books, from Islam Akhun for Rs. 40. Said to have been found

in a half-buried house in Aq Tala Tuz, see Introduction, p. xvii.

Size, 13X11J". Number of forms, 39. Paper, variety IIIc. Many

leaves bear fatty stains, and a few are damaged by burns. Riveted

with three copper pegs ; the guards consisting on one side of irregular

quadrangular pieces, about 2|" square, cut from an ornamented plaque,

similar to those shown in Plate IV, fig. 1, on the other side of circular

pieces, about 1^" diameter, resembling a coin (see Plate IV, figs. 2-9).

Printing very carelessly done. The block was not properly inked, the

impressions are very faint ; moreover they are not properly adjusted ; not

^infrequently one or two lines of two successive impressions are printed

over one another ; often the columns are not kept straight.

Contains recension I/, t.e., lines 1, 2, 4, 5 of A, printed in four

columns on each page, each column (except the second) consisting

of seven or eight impressions ; but in the second column on each page

76

Dr. Hoernlc — Antiquities from Central Asia. [Extra No. I,

(counting from the outer margin of the page) one impression, being

the fifth of the series of eight, is omitted, leaving a square blank.

To this rule there are only very few exceptions : on one page (form

23, page 1)1 have observed only three columns ; and on four pages

(f. 21, p. 4; f. 30, p. 4 ; f . 33, p. 3 ; f , 37, p. 4) there is no blank

space ; and on one page (f. 13, p. 4) the blank space takes the

place of the third impression of the series of eight. On each page

the four columns stand in the same position towards one another,

either all four upright, or all four reversed ; but there is no order

whatever with regard to different pages ; on one page, all four columns

may be upright, on the next page they may again be all upright or

they may be all reversed. No. VII in Woodcut No. 12 will give some

idea of the arrangement of the text on the pages of this book. In five

places an altogether different text is printed in the blank space, and

the print of this text, with one exception (on form 22, p. 1), always

runs at right angles with the proper text of the book. These five

places are : form 4, pp. 2, 3 ; f. 22, pp. 1,4; f. 23, p. 4 ; f. 27, pp. 2, 3 ;

f. 29, pp. 2, 3. The text, thus introduced, is formula A of the

Second Set.

No. VIII. Book.

Belongs to M. 7. Found and acquired in the same way as

No. VII, q.v. Size, 10J X 6£-", but elliptical in form, as shown in

Woodcut No. 12. Number of forms, 44. The usual three blank pages are

wanting at one end of the book ; on the other hand there is in the

middle of the book one form with three blank pages, from which it

would seem that in binding the book the final form has been mis¬

placed. Paper, variety IIIc. Numerous fatty stains, and a few burns.

Riveted with two copper pegs only, the guards being small elliptical

(about lxf") pieces, cut from a plaque, similar to the round pieces

shown in Plate IV, figs. 4-9.

Contains recension I#, i.e., lines 9, 11, 13, 14 of formula C,

printed in five columns on each page, each column consisting as a

rule of two impressions ; therefore ten impressions of the text on each

page. Exceptionally 2f impressions are found on a few pages. As

a rule the columns are turned head to head on two adjoining pages ;

out of a total of 84 pairs of such pages, that arrangement is found

in 58 pairs. In 6 other pairs the columns are turned foot to foot.

In the remaining 20 pairs, the foot of the columns on one page is

turned towards the head of the columns on the adjoining page.

1899.]

Section II. — Block Prints .

77

Second Set. (Plates IX, X, XIII and XIV.)

This set comprises six hooks. Some portions of the text are

also found in the three books of the Sixth Set, and in book VII

of the First Set. A detailed description of this second set, illus¬

trated by two Plates (Xos. I and II), was published by me in the

Proceedings of the Asiatic Society of Bengal, for April, 1898 (pp. 124-

131). I shall, therefore, here content myself with a briefer account,

but correcting some errors and adding such information as I have been

able to glean in the meantime.

The text, occurs in two different recensions, a shorter one of 12

lines and a longer of 13 lines, which I shall denote respectively by II a

and 116. The shorter recension Ila (see Plates IX, fig. 2 and X, fig. 1)

has its text arranged in two columns, with a wide interval, running

vertically, while another wide interval intersects the two columns

horizontally between the 7th and 9th lines. The two intervals thus

present the shape of a cross. The longer recension II b (Plate X, fig, 2)

fills up these cross-shaped intervals with additional texts. It consists,

therefore, of three columns, the additional column being placed in the

vertical interval, and of thirteen lines, the additional line occupying

the horizontal interval. It will be noticed that there are a few differ¬

ences in the type of the two recensions, e.g ., in the first letter of

lines 6 and 7 ; but this may be, merely due to imperfect inking.

Recension Ila is found in books I— III and recension 116 in books IV-VI.

The shorter text Ila which is common to both recensions, consists

of six portions or formulas, which I shall distinguish as A, B, C, D, E

and F. The formulas A, B, C comprise lines 1, 2, 3 — 5, 6, 7 — 9, 10, 11,

32, 13 of column I respectively, while formulas D, E, F are made up of

the corresponding lines of column II. The 4th line appears to be an

additional one, inserted into what must have been originally an interval,

similar to the additional line inserted, in the longer recension II&, into

the still existing horizontal interval. It would seem that a third recen¬

sion must have existed, which possessed two horizontal intervals, between

lines 3, 5 and 7, 9 respectively. I have not, however, met with this

third recension in any book. Of the six formulas, C ( i.e ., lines 9-13

of column I) is found in books Xos. land II of the Sixth Set (see

Plate XIV), while formulas A, B, D, E (each consisting of three lines)

are found in book No. Ill of the Sixth Set, and formula A is also found

in book VII of the First Set (see Plate XIII). As yet the formula F

has not been found by me separately in any book.

For printing these several texts seven different blocks must have been

in use : one for recension Ila, another for recension 116, and five more for

the five formulas A,B,C,D,E. This is clearly shown by the enclos-

78

Dr. Hoernle — Antiquities from Central Asia . [Extra No. 1,

ing lines which still exist. These lines, e.g., are seen running right round

the two columns of recensions Ila and life, see Plates IX and X. They

are also seen enclosing each of the five formulas A, B, C, D, E, see Plates

XIII and XIY. As shown by these lines, the dimensions of the blocks

must have been about If x If" for A, 2§ x If" for C, 2| x If" for D, and

If X If" for E. The additional portions, viz., the middle column III of

recension II fe, line 4 of recensions Ila and life, and line 8 of recension life,

do not appear to have been printed separately ; nor is there any evidence

to show that separate blocks existed for printing them.

It is probable that once there existed three blocks : (1) a block

holding a text of 11 lines, omitting line 4 of recension Ila, and therefore

showing twTo blank intervals and presenting the shape of a double cross ;

I may call this recension lie ; (2) a block holding a text of 12 lines,

with one blank interval, in the shape of a single cross, being recension

Ila ; (3) a block holding a text of 13 lines, with no blank interval, being

recension life. No book, exhibiting recension lie, has come to light. The

block for it, therefore, cannot have been found by the treasure-seekers. If

it had been found, it is morally certain (on the assumption of forgery) that

books would have been printed with it and brought into the market.

But, the block for recension lie not having been found, it is difficult to

understand, on the one hand, how the existence of recension Ila, should

have suggested to a forger to omit line 4 and manufacture blocks for

A, B, D, E ; or, on the other hand, how the separate existence of A, B,

C, D and E should have suggested to a forger to combine them into one

text Ila, and manufacture a block for it, containing the intermediate

line 4 and a blank interval between lines 7 and 9 ; or again, to combine

them into an alternative text life and manufacture for it another block

containing the two intermediate lines 4 and 8. One can imagine a

forger omitting extant lines, but not inventing new lines for which he

has no pattern. Add to this that the formula F has never been found

printed separately ; so that the forger would have had to invent, for the

recensions Ila and life, not only the intermediate lines, but also the whole

formula F. The improbabilities of such a theory are overwhelming.

It follows, therefore, in the alternative, that either the books are genuine-

or that at least the orginal blocks must have been found for the recen¬

sions Ila and life as well as for the formulas A, B, C, D and E. From

these original blocks, of course, books might have been printed; but

the forgery could have extended no further.

With regard to the question of the beginning and end of the text,

book No. Ill affords a similar test to that in book No. I of the First

Set. Jn that book one end of the text is always turned towards its

upper and lower edges, whence it may be concluded that that end

Section II. — Bloch Prints.

79

1899.]

holds the beginning or the top-line of the text. On the Plates the text

is represented in the position thus indicated.

One book, No. Ill, of this Set, as will be shown below, is provided

with additional small legends, similar to those in Book IV of the First Set.

No. I. Book.

Same as “ Block-print (3 ” in Proceedings. Belongs to M. 3. Size,

6f x4". Number of forms, 38. Riveted with three copper pegs, which

are held in position by two copper slips, running in front and at the

hack of the hook, along its longer side, and measuring 6 f x § inches.

Paper, variety Ilia. Fatty stains on many leaves ; no marks of burn¬

ing or singeing. •

Contains recension Ila, printed once on every page, and standing

upright and reversed on every second or third form ; thus upright on

forms 1, 2, 5, 6, 7, 9, 11, 14, 17, 39, 21, 23, 25, 27, 28, 31, 32, 35, 36, 38 ;

reversed on forms 3, 4, 8, 10, 12, 13, 15, 16, 18, 20, 22, 24, 26, 29, 30,

33, 34, 37.

No. II. Book. (Plate IX, fig. 2.)

Same as “ Block-print a” in Proceedings. Belongs to Gr. 7. Size,

11 J x 7f". Number of forms, 32. Peculiar in having covers of paste¬

board, made of four ordinary leaves pasted together. Paper, variety III6.

Most pages stained, and singed or burned. Bound with three copper nails.

Contains recension Ila, printed in three columns on each page ;

each column consisting of one impression; accordingly the text three

times on each page, but reversed on alternate pages, i.e., turned foot

to foot, similar to the arrangement of book No. IV of the First Set

(Woodcut No. 11). With the exception of one form, the columns of

the text are printed parallel to the narrower side of the book, so as to

turn their foot towards the inner margin of the book, and so close toge¬

ther that their edges touch, and sometimes overlap one another. The

exceptional form is the 22nd. It bears on page 1-3 also impressions of

recension la of the First Set ; and the impressions of the two different

texts, indioated in the subjoined diagrams by the letters dc ba and the

numerals 321 respectively, are arranged as follows ; the inner margin of

the page being indicated by parallel lines.

1st page. (Plate IX), 2nd page.

321

654

987

fO _

dc ba

hg fe

ml hi

(V

Ci

QO Or iO

^ K*

rO

Tt-tM

rO

.2 £

3rd page.

?s.

CM io GO

Co 'to Oi

rO

O _ '-a

pe 5* g

e *§ -p

o

4th page.

W \uidc ba

df bi[hg ef

vq opml Jci

pi puiq fill

&f bxpf 6x[

80

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

This would seem to have been the trial-form, employed to discover by

^<5* §■* §- which arrangement the largest number of impressions

could be crowded on to one page. The usual arrange-

5. » »

- <5 O

' — i-j O'*

IT. <v a

=e a.

ST

a a

a

e*.^» C*

S^h§-

a ^ a-

ment is shown in the margin.

No. III. Book. (Plate X, fig. 1.)

Same as “ Block-print y ” in Proceedings. Belongs to M. 3. Size,

ll|x4§ inches. Number of forms, 40. Bound with three twists of

paper. Paper, variety Ilia. A few fatty stains and burns ; on the

whole fairly clean. Printing ends on the penultimate page of the

last form, instead of the ante-ante-penultimate page as usual.

Contains recension Ila ; there being two impressions of it on each

page, placed foot to foot, the upper one being complete (12 lines), the

lower more or less incomplete (as a rule 10 lines), the arrangement

being similar to that in Book No. I of the First Set. There are only

four exceptional forms, on which the impressions are placed head to

foot; viz., forms 6, 12, 17 (1st, 2nd, 4th pp.), and 31 (2nd and 4th pp.).

Seeing that there are 40 forms (or 156 printed pages), these few excep¬

tions (13 pages) are evidently accidental misprints. As in the case of

book No. I of the First Set, this “ foot-to-foot ” arrangement of the- text,

seems to be a clear indication as to which is to be taken as its top-line.

Within the horizontal intervals, in the middle of the page, additional

small legends (similar to those of No. IV of the First Set), consisting

of 4 to 6 letters, are inserted, running in the same direction as the text,

but of somewhat larger size, and apparently written by hand. They

are all shown on Plate II of the Proceedings for 1898. As in the case

of Book No. IY of the First set, they do not occur on every page,

but only on the two outside pages of a form. Moreover, in nearly

one-half of the forms (viz,, 19 out of 40) they do not occur at all.

No. IY. Book.

Same as “ Block-print 8 ” in Proceedings . Belongs to M. 3. Size,

8x5| inches. Number of forms, 40. Riveted with three copper

pegs, the guards being small oblongs, 1 or 1 J x f ", Paper, variety

Ilia ; on the whole fairly clean, no burns. Printing rather indistinct,

owing to the inferior quality of the paper, on which the ink has a

tendency to run, so that the imprint occasionally shows on the other

side, in which case sometimes the reverse page is not printed at all.

1899.]

Section II. — Block Prints.

81

Contains recension 115, printed once on each page, but, excep¬

tionally, standing upright on every page, so that the book can be read

right through, from page to page, without turning it right round. To

this arrangement there are only a few exceptions; on 14 pages (out of a

total of 74 printed ones) the imprint is reversed, and these are clearly

accidental errors. There are 12 forms, which at first seem not to

agree with the arrangement, all the imprints on them being reversed.

But they only require to be folded the other way, and to be turned,

when they all come right. With respect to these forms, therefore, the

book has only been carelessly bound.

No. Y. Book. (Plate X, fig. 2.)

Same as “Block-print e” in Proceedings. Belongs to M. 3. Size,

9|x4J inches. Number of forms, 34. Riveted exactly like No. IY.

Paper, variety Ilia. Many stains, but no burns. Printing similar to

that in No. IY.

Contains recension II b, printed twice on each page, so that the two

impressions stand head to foot, the lower one being complete (13 lines),

the upper, more or less incomplete (as a rule 6 or 8 lines, i.e ., 11. 13-8,

or 11. 13-6) owing to want of space. There are only four exceptional

pages on which they stand foot to foot, viz. form 7, p. 4 ; f. 17, p. 3 ;

f. 23, p. 1, and f. 26, p. 1 ; and these, of course are careless misprints.

There are also two pages on which there is only one impression ; viz .

form 17, p. 4 and f. 26, p. 2.

In the ordinary, head-to-foot, arrangement, the pairs of impres¬

sions stand upright and reversed on alternate pages.16 The two

varieties of arrangement may be represented thus, the parallel lines

representing the inner edge of the pages or the fold of a form.

Ordinary. Exceptional.

The ordinary arrangement is very curious for two reasons : (1 )

because the page commences with the incomplete member of the pair

of texts, which must have been awkward in reading the book, if it was

meant for reading; (2) because it compels the reader to begin with

the left hand pages, that is, at the wrong end of the book, assuming

15 There are some three or four exceptional pages which do not keep the alter¬

nate order. These evidently are misprints.

J. i. 15

82

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

that the proper way of reading these books is from the right to the left

in the Semitic fashion. It is, however, by no means certain that the

direction of the scripts is from the right to the left. See also the

General Remarks on the Orientation of the Books.

No. VI. Book.

Not mentioned in the Proceedings. Belongs to M. 6. Acquired

from the Rev. Mr. Hogberg. Size, 9|x4 inches. Number of forms,

8 ; but the book is a mere fragment ; a large portion, including begin¬

ning and end and the rivets, is wanting. Paper, variety Ilia ; some

stains, but no burns. Printing similar to tli at in Nos. IV and V, but

even worse ; ink has run so badly through the paper that many pages

coaid not be printed at all, the print on one side showing through on

the other.

Contains recension 1I&, printed exactly as in No. V, two impressions

on each page, standing head to foot, the lower being complete, the

upper, incomplete.

Third Set.

This set comprises three books. The peculiarity of them is that

their forms are not placed one upon the other, as in the bound books of

all the other sets, but are inserted one within the other. Moreover the

printing does not commence upon the fourth, but on the second page,

and does not stop on the ante-ante-penultimate, but on the penultimate

page.

The text of this set consists of two short formulas A and B.

Formula A is made up of five short lines, of about 5 or 6 letters each. It

is found in all three books. Formula B consists of four longer lines of

about 7 or 8 letters each, and is only found in two books, Nos. II and III.

For printing these two formulas, two separate blocks must have

been used. This is evident from the lines, enclosing the formulas,

as well as from the different size of the two blocks. Measured between

those lines, the dimensions are l§xlf,/ and x 1TV' respectively.

There is nothing to indicate what is top and bottom of the formulas.

No. I. Book. (Plate XI.)

Belongs to M. 3. Size, 14§x4|". Number of forms, 16. It was

originally bound with three twists of paper. These having broken,

the book has been re-stitched with fresh thread, apparently by the

finder, or in Mr. Macartney’s office. It is, therefore, not certain whether

the number of the forms is complete. Paper, variety III b. Leaves much

burned and torn.

1899.]

Section II. — Bloch Prints.

83

Contains both formulas A and B, printed in two rows on each page,

but without any particular order. This is illustrated by the subjoined

diagrams, which also indicate the relative position of the two formulas

and occasional irregularities in printing them. The straight lines indi¬

cate the fold of the sheet. As a rule, the formulas occur 5-8 times in a

row, and occupy alternate rows. There are only three exceptions ; on the

15th sheet the foimula B occurs alone ; on the outside of the 13th sheet,

each formula occupies both rows of a page ; and on the outside of the

2nd, and the inside of the 9th sheet, there is only one row on one page,

and that row is filled with prints of formula B alone ; probably the A

row has been simply forgotten by the printer, as its proper space is left

blank.

Second Sheet.

Outside.

Oi A

A

WA

tdA

A

ha

A

“A

Inside.

Third Sheet.

Outside. Inside.

td y

V

td y

wy

“v

HT

Outside.

tdA

A

td

A

ttA

td A-

Fourth Sheet.

Inside,

td A

td y

y

td

y

td

*►

A

tdA

td

A

WA

tdA

y W

<1 cq

A

A «

> td

>

td

Thirteenth Sheet.

Outside.

yy

y y

yy

y v

yy

td td

to td

Ninth Sheet.

Inside.

a

ay

a v

a ^

a

y

a y

ay

No. II. Book.

Belongs to M. 6. Acquired from the Rev. Mr. Hogberg. Size

x6|". Number of forms 13. Paper, variety Ilia. A few fatty stains,

and marks of singeing. Stitched in three places with loops of loosely

twisted thread.

Contains both formulas A and B, printed in three and four columns,

each column containing as a rule five impressions. As may be seen

from the subjoined diagrams, there is no order whatever in the arrange-

84

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

ment of the impressions. Occasionally formula A occupies the whole

page, in which case there are four columns, with five (or six) impres¬

sions in each ; total 20 (or 24) impressions. At other times formula B

occupies the whole page, when there are only three columns, with five

impressions in each ; total 15 impressions. But in many cases both

formulas occur promiscuously on the page, in which case also there are

only three columns, with a total of 15 or 16 impressions. Clearly this

book cannot have been intended for intelligent reading.

First Sheet. Second Sheet.

Inside. Outside. Inside.

Third Sheet.

Outside.

A Ay A

y Ay A

yAy A

A y Ay

Ay A y

y A y A

y Ay A

y A y A

y Ay y

y A y A

y Ay y

Inside.

Ay y A

yAy A

y A y y

yAy A

y Ay y

y B B y

y BBB

y BB B

y BBB

No. III. Book.

Belongs to G. 7. Size, 9 x 7". Number of forms 18. Paper,

variety Ilia, A few burns and fatty stains. Riveted with three

copper pegs.

Contains only formula A, printed in three

columns on each page, the two inner columns consist¬

ing of seven, and the outer, of six impressions, which

latter stand at right-angles to the former ; otherwise

there is no sort of order in the relative position of

the imprints, as the marginal example may show.

1899.]

Section II. — Block Prints.

85

Fourth Set. (Plate XII).

This set comprises eight books. Two of the books contain also

other texts in addition to their own, viz., No. VII contains also the

text of the Fifth Set, and No. VIII, of the Seventh Set.

The text of the Fourth Set consists of a formula, containing five

lines of writing. There appear to have been two blocks, from which this

formula has been printed off ; for I have noticed a very slight differ¬

ence in tbe dimensions of the print : in Books I, III, V, the lines

differ by £ to yy an inch in length from those in Books II, IV. The

enclosing lines of the block of the longer print alone are preserved in

book No. VII, and this block measures, within tbe enclosing lines,

3§ x 2£ inches.

Two of the books seem to afford indications whereby to determine

what is top or bottom, right or left, and beginning or end of the text.

These are Nos. II and VI. No. II is printed with three columns on each

page, and a column consists, as a rule, of six impressions of the formula ;

but there are two pages on which the columns contain 6^ impressions, and

one page on which they contain only 5^ impressions ; as illustrated by

Dn the margin. On the other hand book No. YI is printed

with 2^ columns on each page ; that is, two columns contain,

each, two complete impressions of the formula, while the third

column contains only two half-impressions of it ; as illustrated

in diagram II. It would seem right to conclude from

diagram I, that ab c must be the top-line ; for if ghi

were taken to be the top-line, the printing would have com¬

menced with a mutilated formula. Similarly diagram II

would show that a is right and c left of the formula, and

that the reading of the latter must commence, not with c , but

with a ; or in other words that the formula must be read

from the left to the right. It would follow, therefore, that

if the formula is to be read in the European fashion, it commences in

the left-hand upper corner, as shown in the photographic facsimile on

Plate XII ; or, if it is to be read in the Chinese fashion, its com¬

mencement lies in the right-hand upper corner. Unfortunately, as the

case of the Fifth Set shows, the argument is not so conclusive as would

be desirable.

No. I. Book.

Belongs to M. 7. Said to have been found at Aq Tala Tuz.

Size, 23| X 13|.” Number of forms, 36. Riveted with three copper pegs,

S6 Dr. Hoernle — Antiquities f rom Central Asia. [Extra No. 1,

tlie guards being large pieces of a broken-up plaque (see Plate IV, fig. J).

In almost perfect preservation. Paper, varieties III b and IIIc mixed.

Text printed in the three columns on each page ; each column

consisting of nine impressions of the formula, which accordingly is

repeated 27 times on every page, or 3,726 times in the whole book

(i.e., 27x138 printed pages). The columns run parallel with the longer

side of the book, 'and stand regularly upright and reversed on alternate

pages; that is, upright on pages 4, 6, 8, etc., and reversed on pages

5, 7, 9, etc. The pages 4, 6, 8, etc., of course, are left-hand pages; and.

when the book is turned right round, in order to bring the reversed

formulas into the upright position to read them, the pages 5, 7, 9, etc.

now become left-hand pages. It follows, therefore, that on reading

this book, all the left-hand pages must be read first, and afterwards all

the right-hand pages, — wThich latter pages, of course, on turning the

book round to read them, also become left-hand pages. Accordingly

whichever way the book is placed for the purpose of reading, it is

always the left-hand pages that must be read.

This book is distinguished from all others, in having a finely executed

sketch of a head on the fourth page of the 15th form ; see Plate XVII.

The sketch is placed horizontally across the page, so that the bottom of

it is turned towards the inner margin of the page. From the fact that

the columns of the print curve round the sketch, it is clear that, it was

drawn on the page, before the latter was printed on. The head seems

to me to show Arian features, and is sketched with much artistic skill.

No. II. Book.

Belongs to M. 7. Found at Aq Tala Tuz. Size, 15§xllf".

Number of forms, 17. Bound with three twists of paper. Well pre¬

served ; a few slight burns and stains. Paper, variety III5. Printing

not quite distinct. It commences, as usual on the 4th page, but ends on

the penultimate page ( i.e ., on the 3rd page of the last form).

Text printed in three columns on each page ; each column consist¬

ing, as a rule, of six impressions of the formula, which accordingly is

repeated 18 times on every page. On two pages there are 6| impres¬

sions, and on one page only 5J. The columns stand regularly upright

and reversed on alternate pages ; that is, upright on the right-hand

pages 5, 7, 9, etc., and reversed on the left-hand pages 4, 6, 8, etc.

Accordingly this book must be read differently from book No. I ; that

is, in whichever way it is placed for reading, its right-hand pages must

be read. There are only five pages which do not conform to the

regular order ; these are p. 4 of form 3 ; pp. 1 and 4 of form 10, and

p. 2 of form 15 ; and these, clearly, are mere lapses of the printer.

1899.]

Section II. — Block Prints.

87

No. III. Book.

Belongs to M. 6. Purchased from Badruddin. Size, 22§ x 13".

Number of forms, 19 ; all (with the exception of forms 7 and 8) cut

through at the back, into separate leaves. Very well preserved ; only

very few and very slight burns and stains. Paper, variety III6. Riveted

with four copper pegs.

Text printed in three columns on each page ; each column consisting

of eight impressions of the formula, which, accordingly, is repeated

24 times on every page, or 1,680 times in the whole book ( i.e ., 24x70

printed pages). The columns run parallel with the longer side of the

book, and stand regularly upright and reversed on alternate pages ;

exactly as in No. II ; that is to say, whichever way the book is placed,

the right-hand pages must be read. There are only four exceptional

pages, due to blundering of the printer; viz., pp. 2 and 3 of form 6;

p. 4 of form 9, and p. 1 of form 19.

No. IV. Book.

Belongs to M. 7. Pound at Aq Tala Tuz. Size, 10§ x7f//. Number

of forms, 30. Bound with three twists of paper. Paper, variety IIIZ>.

Extensively marked with water and fat stains ; no burns. Print rather

difficult to read, owing to the stains and defective inking. Final blank

leaf lost.

Text printed, as a rule, in three columns on each page; each

column consisting of two impressions of the formula, which, accordingly

is printed six times on every page. The columns run parallel to the

narrower side of the book, but do not stand upright and reversed on

alternate pages, but keep the same direction on every page, that is,

on the left-hand pages the head of the columns is turned outside, and

on the right-hand pages, inside, as shown in the diagram

on the margin. Hence in reading the book, it does

not require turning round. There are a few exceptional

pages, on which the imprints take reversed positions,

but these are clearly blunders of the printer or binder.

There are, however, seven pages, on which the arrange¬

ment is altogether different ; viz., p. 4 of form 1 ; and pp. 1

and 4 of forms 6, 10, 38. On these pages the text is

printed in two columns, running parallel to the longer

side of the book, and each column contains four impressions of the for¬

mula, which, accordingly, is repeated eight times on each of these pages.

No. Y. Book.

Belongs to M. 3. Purchased from Badruddin. Size, 31x6J".

Number of forms, 45. Riveted with three copper pegs. Paper, variety

88

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

IIIc. Fairly clean ; a few fatty stains and small burns ; the two outside

leaves torn to shreds. Print almost illegible, owing to defective

printing.

Text printed exactly as in book No. IV, save that the exceptional

arrangement of the text in two columns never occurs.

No. VI. Book.

Belongs to M. 7. Found at Aq Tala Tuz. Size, 8 x5%". Number

of forms, 26 ; but the first form is in shreds, and the last form is

misplaced through a blunder of the binder, standing fifth in the present

series. Riveted with three copper pegs. Many large fatty stains and a

few burns. Paper, variety III6. Print almost illegible, owing to defec¬

tive inking.

Text printed in 2J columns on each page ; each column consisting

of two impressions of the formula, of which, accordingly, four complete

and two half -impressions stand on every page. The columns run

parallel to the narrower side of the book, and are arranged exactly

as in Book No. IV. Here, too, as in No. IV, there are a few exceptional

Form 11.

(margin).

2.

a a

Cr- O'1

o o

a a

C3* Cr-1

o o

a a

o-' cr1

o

rO

e

3.

a

<3-*

a

w a

lO CJJ

a »

o-* a*

pages on which the imprints take reversed positions. An

example, which shows also another misprinted anomaly,

is given on the margin from form 11.

It may be noted that there is a slight blank margin

at the top of the pages (as indicated in the diagram),

while at the bottom the print runs to the very edge.

This shows the modus operandi of the printer, who com¬

menced printing at the top of the page, though ( as shown

by the inner column on page 3 of the diagram) not

(no margin.) always with the initial edge of the block. But when he

came to printing the half-impression on that inner column, he turned

the block round, apparently with the object of getting the initial portion

of the formula on to the page. It seems that, as long as the whole of the

formula could be accommodated on the page, the printer did not care,

in what position he placed the impression of it ; but when he could only

accommodate a portion of it, he was careful in printing its initial

portion. This seems to be a reasonable explanation of the anomaly,

and it points to what apparently must be taken to be the beginning of

the formula.

No. VII. Book. (Plate XII.)

Belongs to M. 6. Purchased from Badruddin. Size, ll|x8J".

Number of forms, 12 ; all cut throtigh at the back into separate leaves.

Exceptional in being bound with only two twists of paper, and in

commencing its printing on the second, and stopping it on the ante-

1899.] Section II. — Block Prints. 89

penultimate page. Paper, apparently variety III&. Rather rotten ;

several leaves mutilated ; very much damaged by extensive fatty stains

and burns. Printing rather illegible on many pages ; two of them

(pp. 10 and 11) being altogether blank, owing to the print on the

reverses (pp. 9 and 12) showing through.

Contains its own proper formula, as well as that of the Fifth Set.

The latter, however, is only found on pages 2, 3, 5, 8, 9, 12, 14, 15, 21,

29, 32, 37, 40 and 45 ; and on these pages it is found in addition to the

proper formula of the Fourth Set. When so found, it always occupies

the outer side of the page, except on p. 14, where it stands on the

inner side; see, below, marginal diagram II. On page 13, it should

also have been found, but (apparently by an oversight) the space has

been left blank.

Text printed in three columns on each page ; each column consisting

of three impressions of the formula; that is, either

thrice the proper formula of the Fourth Set; or

twice that formula and once the formula of the

Fifth Set. As the latter formula is longer than

that of the Fourth Set, and hence occupies more

space, it is only found 2| times on a page ;

and on these pages, the lower portions of the

columns overlap one another. This will be under¬

stood from the diagram I shown in the margin, in

which a be , and 12 34 denote the formulas of the

Fourth and Fifth Sets respectively. Here, also, the

same modus operandi of the printer may be noticed,

as in Ho. YI. He commenced at the top of the page (as shown in the

marginal diagrams) where there is a broad blank

margin, while at the bottom the print runs right

up to the edge of the pages with the half-impression

of the formula of the Fifth Set. This circum¬

stance would seem to serve as an indication of

the beginning of the latter formula ; though this

conclusion is not quite borne out by diagram II,

which shows the exceptional page 14, above referred

to. This diagram would rather seem to show that

the beginning of the formula of the Fifth Set lies

at its other extremity ; but then both pages, espe¬

cially page 15, show an anomaly in the relative

position of the two formulas.

It will further be seen from the above two diagrams that the columns

run parallel to the narrower side of the book, and are arranged in a

J. i, 16

II.

p. 14. p. 15.

(broad margin).

(no margin).

I.

p. 2. p. 3.

(broad margin).

(no margin).

90

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

different position from that in books Nos. IV to VI. There the heads

of the columns are turned in the same direction on consecutive pages

(see diagram in No. IV), while here the heads are always turned in

opposite directions, against one another, that is, they are always turned

reversely on alternate pages.

No. VIII. Book.

Belongs to M. 8. Found at Kiang Tuz. Size, 14 y 8f. Number of

forms, 20. Paper, variety II TcZ. Surface greased, and a few burns.

Printing rather indistinct. Riveted with two copper pegs.

Contains besides its own proper text that of the Seventh Set. The

latter is found only on one form, the 3rd, where it is printed, on all its

four pages, exactly in the same way as the proper formula of the Sixth

1st form, p. 14. Set. This formula is printed in two columns on each

page, running, as a rule, parallel to the longer side

of the book and consisting each of five or six impres¬

sions. The columns stand in no particular order,

sometimes upright and reversed on the same page,

sometimes so on alternate pages, sometimes in the

same position on consecutive pages. On four pages

(viz., on the first and last printed pages, as well as on the 13th and

28th) the columns stand at right-angles to one another, the outer

column consisting of only four impressions, as shown in the marginal

diagram.

Fifth Set. (Plates XI and XII.)

This Set comprises eight books. Its text is found in two different

recensions, a shorter one of three, and a longer of four lines. These I

shall denote by Va and Yb respectively. Besides its own proper text,

there is also found on one page of book No. VIII the text of the First

Set.

The shorter recension Va is found in all the books of this set ;

also in No. VII of the Fourth Set. The longer recension Yb is only

found in No. VIII of this set, where it occurs together with recension

Va. The two recensions only differ from each other by Yb adding a

fourth line to the three lines of Ya.

Two distinct blocks have been used for printing the two recensions.

This is shown by their enclosing lines which may be seen on the

facsimiles. The block of Ya measured 4^x1^ inches; that of Yb

measured 4|- x inches.

There is nothing to show which is top and bottom, or right and

left, or beginning and end of the two formulas. From the way in which

a

o*

CS

a

O4

o

a

O'*

o

a

O4

Ci

abc

1899.]

Section II. — Bloch Prints.

91

the formula is printed in No. VII of the Fourth Set (PI. XII), it would

seem probable that the top of it is the line which adjoins the formula of

the Fourth Set ; for on this supposition it would lie on the page in the

same direction as that formula. It would also seem probable that it

commences on the right-hand side, and must therefore be read from the

right to the left ; for that half of the formula, which appears on the

page of No. VII of the Fourth Set, is, on the above supposition, its

right-hand half, and it seems reasonable to assume that, when only

one-half could be printed, it was the initial half that was printed. More¬

over in certain pages of book No. IV (see below the detailed descrip¬

tion) it is always what on the above theory is the left-hand side of

the formula which appears in the middle of the page ; and this circum¬

stance points to the same conclusion, as it seems also reasonable to

assume that, as usual, the reader was intended to commence reading

from the margin of the page. But unfortunately this argument is

weakened by the fact that in books Nos. I and III, where occasionally

only a portion of the formula is printed, it is indifferently either one

or the other of its two outer lines that is omitted. The fact is that the

force of all such arguments depends on the assumption that these

block-print books were intended for reading. If they were not meant

for reading, but intended for the mechanical use of merely turning the

pages, it was obviously quite immaterial which portion of the formula

was printed whenever the space did not suffice to print the whole.

No. I. Book.

Belongs to M. 3. Purchased from Badruddln. Size, 8J x 5|".

Number of forms, 37 ; the first and last forms, however, consist of

two sheets each, placed within one another, and pasted together 16 to

form thick covers, as in No. II of the Second Set. Paper, variety III6.

Fairly clean. Printing, indifferent. Riveted with three copper pegs.

Contains recension Va, printed six times on each page, in a

column which runs parallel to the longer side of the book, and stands

regularly upright and reversed on alternate pages. The latter rule is

not observed on 10 pages out of a total of 142 pages, and these pages,

therefore, represent clearly mere accidental lapses of the printer. On

a few pages, the formula is only found five-times repeated ; but on many

pages the sixth repetition, standing too close to the top or bottom of the

page, is incomplete, one of the three lines of the formula being omitted.

This omitted line is sometimes one, sometimes the other of the two

outer lines, so that no conclusion can be drawn as to what is the initial

and what the final line of the formula.

16 In the initial cover the paste has given way, and the leaves are now separate.

92 Dr. Hoernle — Antiquities from Central Asia. [Extra No. I,,

No. II. Book.

Belongs to M. 9. Size 7fx5f". Number of forms, 22|, the

seventh form consisting of only one leaf. Paper, variety III6. A few

leaves damaged by burns and fatty stains. Riveted with three copper

pegs, the guards on one side consisting of large fragments of the

plaque, shown on Plate IV, fig. 1.

Contains recension Va, printed, as a rule, in two columns on each

page ; the columns consisting, as a rule, each of four repetitions of the

formula, and running parallel to the narrower side of the book.

On a few pages there are only three repetitions in the column ; the

formula accordingly occurs, as a rule, eight times, and exceptionally

six times, on every page. The columns are placed in the same direction

on every page, that is to say, line 1 of the formula always stands

alternately near the outer and inner edges of the page, so that in reading

the book, it does not require to be turned round. There is only one

exceptional page (evidently a misprint) in which the columns are

reversed. Thus

Regular Pages. Exceptional Pages.

4th 5th Regular Reversed.

As the space is barely sufficient to accomodate the whole breadth

of the columns, it happens that occasionally the terminal letters on

one side (a, a, a, a) of the formula, at other times those on the other

side (d, d, d, d) are omitted. It is thus impossible to use these omis¬

sions as a test for determining the initial and final sides of the formula.

The eleventh form is a total exception : on it the text is printed in one

column, which consists of six imprints of the formula, and runs

parallel to the longer side of the book.

No. III. Book.

Belongs to M. 7. Found at Aq Tala Tuz. Size, ll^Xfif".

Number of forms, 52. Paper, variety IIIc. Initial and final leaves

damaged ; otherwise well preserved. Riveted with three nails ; guards

round pieces, as shown on Plate IV, fig. 3.

Contains recension Va, printed in one column on each page, run¬

ning parallel to the longer side of the book, and consisting of (as a

rule) seven, or (sometimes) eight impressions of the formula. When

1899.]

Section II. — Block Prints .

93

there are eight repetitions on the page, the space barely suffices for

them ; and accordingly sometimes one, sometimes the other of the

outer lines of the formula is omitted, the same as in No. I. The

columns stand upright and reversed on alternate pages ; though

there are many misprinted pages, about 12 per cent, of the total.

No. 1Y. Book.

Belongs to M. 8. Found at Kiang Tuz Size, 12§ x 8". Number

of forms, 20. Paper, variety Illd. One outer form damaged by

burns ; entire surface of all leaves greased with fat. Riveted with threo

copper pegs. Printing commences, contrary to the usual rule, on the

second page of the first form, but stops, as usual, on the ante-ante¬

penultimate page. Print throughout almost illegible, owing apparently

to the greasy surface of the paper.

Contains recension Va, printed in two columns on each page,

running parallel to the longer side of the book, and containing each

nine impressions of the formula ; there being accordingly 18 impressions

on every page. On some pages the columns are printed both upright ;

on others one column stands upright, the other reversed. Thus

Page 16. Page 17.

These two arrangements seem to have been observed quite promis¬

cuously ; thus : on 1st form upright and reversed throughout,

upright throughout,

upright on 3rd page only,

upright on outer pages only,

upright on two first pages only,

upright on outer pages only,

upright on inner pages only,

do. do. do.,

do. throughout, etc., etc.

On those pages where the impressions are placed in opposite direc¬

tions, it is always the left-hand sides of the formula (as seen on Plate

on 2nd

on 3rd

on 4th

on 5 th

on 6th

on 7 th

on 8th

on 9 th

33

5)

JJ

}»

35

3)

33

JJ

94

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

XII) which stand in the middle of the page, adjoining each other. The

right-hand sides are near the outer and inner edges of the page ; and

as the space is rather limited, occasionally the terminal letters on those

sides are omitted.

No. V. Book.

Belongs to M. 8. Found at Kiang Tuz. Size, 13x8". Number of

forms, 20. Paper, variety II Id. Entire surface of all leaves greasy ; no

burns ; the two outside leaves slightly damaged. Print throughout

rather difficult to read. Riveted with two copper pegs only. The first

(or covering) form is wrongly folded.

Contains recension Va, printed in two columns on each page, which

run in right angles to each other, the inner parallel to the longer, the

outer parallel to the shorter side of the book. The former consists

of ten (exceptionally nine) impressions of the formula, the latter of

six, arranged in two lines ; the total number of repetitions on every

page being 16 (exceptionally 15). The inner columns stand regularly

upright and reversed on alternate pages. The outer columns, as a

rule, have the third line of the formula (as shown in Plate XII) turned

towards the inner column ; but on 21 pages (out of a total of 76) the

formula is turned the other way, being probably misprints. The two

positions are shown in the subjoined diagrams.

Regular.

ns ns abed

8 e abed

abed

ns ns

u abed

10 abed

e « , ,

abed

ns ns abed

° ® abed

•O lO

e e abed

poqv

po qv

poqv

e e

cy cy

o o

a*

O S3

cy cy

Ci o

S3* S3*

a a

cy cy

o cs

fU S3*

Exceptional.

a a abed

^ ^ abed

o o

S3* S3* o 6 cd

abed

a a

cy cy

° ° a b c d

^ ^ 7 ,

abed

o anted

^ 'Tabcd

O o

a* S3* a b c d

poqvrx * ns .

poqv 8 8

poqv

ns ns

poqv 0 o

poqv

8 8

poqv

poqv n> ns

poqv ^ °

L * rO .O

poqv <3 $

No. VI. Book.

Belongs to M. 8. Found at Kiang Tuz. Size, 13 x 8J". Number of

forms, 20. Paper, variety Hid. All surfaces more or less greasy ; a few

burns ; large piece torn out of the fourth leaf. Print indifferent, and

rather difficult to read. Riveted with two copper pegs only.

Contains recension Va, the arrangement of which is exactly the

same as in No. Y, except that the inner column consists, as a rule, of

nine, and only exceptionally of ten impressions of the formula.

1899.]

95

Section II. — Bloch Prints.

No. VII. Book.

Belongs to M. 8. Found at Kiang Tuz. Size, 16|*7". Number of

forms, 20 ; some of them worn through at the back into separate leaves.

Paper, variety III d. Fairly clean and well preserved. Riveted with

three copper pegs.

Text printed in one column on each page, running parallel to the

longer side of the book, and consisting of twelve impressions of the

formula. As a rule the columns stand upright and reversed on alter¬

nate pages, there being only 13 exceptions (in a total of 74 printed

pages), probably misprints.

No. VIII. Book. (Plate XI.)

Belongs to Gr. 7. Size, 8|x5|". Number of forms, 32. Has

exceptionally seven blank pages both at the beginning and end of the

book ; that is, printing commences on the 8tli page from the beginning,

and stops on the 8th page from the end. Paper, variety III5. Book in

rather bad preservation, and print so faint as to be illegible in many

pages. Bound with two twists of paper.

Contains both recensions, Ya and Yb, but the latter only on the

outside pages of forms 5, 6, 7, 15, 16, 19, 21, 22, 30, and on the inside

pages of forms 14, 23, 29, 17 also on the first (outside) page of form

24; altogether on 25 only out of a total of 114 pages. Curiously

enough on the fourth (outside) page of the 24th form there are printed

two formulas of the First Set.

Both recensions are printed in two columns on each page, running

parallel to the narrower side of the book, and consisting each of three

(very exceptionally four) impressions in the case of recension Va, and

of two (exceptionally three) impressions in the case of recension Yb. The

a a £ space is not sufficient for the exceptional number

^ J4 ^ of impressions ; accordingly in those exceptional

^ 55,4 ^ cases one or two lines, from either extremity, are

p 0 0 omitted. As a rule the columns stand turned in

SJ* ^ opposite directions on alternate pages, as shown in

^ ^ ^ the margin.

Page 4 of form 24 is a curiosity. It is shown on Plate XI. The leaf

is badly damaged by burns. On one side (page 3) it shows the formula

of the Fifth Set in the usual arrangement of two columns, with two

impressions in each, running parallel to the narrower side of the book.

On the reverse side (page 4) formulas C and E, in the recensions I g

H These too become outside pages, if the fold of the forms is turned the other

way.

^

o o o

rO i-O rO

<s c e

^

Sj o o

rO rC)

m

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

and I h respectively, of the text of the First Set are imprinted,

arranged in two columns, at right angles to one another, one run¬

ning parallel to the longer, the other with the shorter side of

the book. The former stands near the outer, the latter near the inner

edge of the page. The outer column consists of two impressions each of

the recensions I g (11. 9, 11, 13, 14) and I h (11. 17, 18, 19) ; while the

inner column consists of four impressions of the recension I g. The two

pages are shown in the marginal diagram, the formula of the Fifth Set

being indicated by abed , and those of the

First Set, by ccc ( = C=I g) and eee

(=E=I/i) respectively. The circum¬

stance of the two mutilated formulas of

the First Set appearing here in this un¬

expected and purposeless way seems to

render the hypothesis of a forgery al¬

most impossible, both with regard to

the whole book No. VIII, and to the

blocks for the recensions I g and Ih.

Sixth Set. (Plate XIII and PI. XIV, fig. 1.)

This Set comprises three books. Its peculiarity is that it is not

appropriated to one particular text only, but presents a collection of

several texts. What was found occasionally as a rare exception in the

other sets — the introduction of a few impressions of an alien text into the

midst of its own proper one — forms in the Sixth Set its main feature.

Of the three books comprised in it, No. I gives the texts of the

Second and Fifth Sets. No. II gives the texts of the Second, Fourth and

Fifth Sets, and No. Ill those of the First, Second, Fourth and Fifth

Sets.

No. I. Book. (Plate XIV, fig. 1.)

Belongs to M. 7. Found at Aq Tala Tuz. Size, 11 x8J". Number

of forms 30. Paper, variety III5. Greatly damaged by exposure to wet ;

also some large fatty stains and burns. Print almost illegible. Bound

with three nails ; guards being large rhomboid pieces of flat, thin, orna¬

mental copper, as shown in fig. 1 of Plate IV.

As a rule, the pages of this book present the texts of the Second

and Fifth Sets. There are only 6 exceptional pages (out of a total of

114 printed ones), viz., 6, 43, 94, 95, 116, 117, which give the text of

the Fifth Set alone.

The text of the two sets are printed in three' columns on each

page, running parallel to one another and to the longer side of the

1899.]

Section II. — Block Prints.

97

book, and standing upright and reversed on alternate pages. It may be

also noted that what has been indicated in the facsimiles (Plates X and

XII) as the first line of the formulas of the two texts, stand, as a rule

(though not always), alongside of one another, pointing to the conclusion

that these two lines occupy the same position in the respective formulas,

that is, that both are the head-lines (as assumed in the facsimiles) or

both the foot-lines. Of the three columns one gives the text of the

Fifth Set, and two give that of the Second Set.

The column containing the text of the Fifth Set always occupies the

inner side of a page, and consists of seven, or more usually eight, impres¬

sions of the formula of that set. But as the space is barely sufficient to

accommodate all eight, one line of the 8th repetition is often omitted ;

this seems to be invariably line 3, as indicated in the facsimile

(PI. XII) ; which fact also points to line 1 being really the head-line.

The two columns containing the text of the Second Set, always

occupy the outer side of a page, and consist each of four impressions

of the formula, so that there are altogether eight impressions of it on

each page. The formula here printed is only a portion of the text of

the Second Set, viz., formula C, or lines 9-L3 of column I of that

text (PI. X).

There is one exception to the arrangement above explained. On

p. 7 there are only two columns, standing at right angles to each other,

the outer one of which consists of four impressions of the text of the

Second Set. The two arrangements are shown in the subjoined

diagram ; a b c d ef denoting the text of the Fifth Set, and 1 2 3 4 5 6

that of the Second Set.

Regular. Exceptional.

The correspondence in the arrangement of the texts of the two

sets suggests that they may also correspond in their scripts and their

meaning. We should have here a bi-script, and perhaps a bi-lingual,

book.

J. i. 17

98

Dr. Hoernle — Antiquities from Central Asia. [Extra No.

No. II. Book.

Belongs to M. 7. Found at Aq Tala Tuz. Size, 11 x8''. Number

of existing forms 28 ; but originally it must have been 30, as the two

outside forms, forming the covers, are missing. Paper, variety IIIZ>.

Corners rounded off, somewhat similar to those of book No. VIII of the

First Set. Condition similar to that of No I. Bound with three twists

of paper.

Contains the text of the Second, and Fifth, as well as that of the

Fourth Sets ; on alternate forms ; commencing with the latter. That

is, the text of the Fourth Set is found on forms 1, 3, 5, etc. ; that

of the Second and Fifth Sets on forms 2, 4, 6, etc.

The latter texts, on their own forms, are arranged exactly as in

book No. 1 ; only as the space is insufficient, a portion of those impres¬

sions of the formula (C) of the Second Set, which are nearest the outer

edge of the page, is omitted ; on two pages (93 and 96) they are omitted

altogether.

The text of the Fourth Set is printed in three columns on each page,

running parallel to the narrower side of the book, and' consisting each

of twro impressions of the formula which accordingly is repeated six

times on every page. The columns keep throughout the same direction.

In fact the arrangement of the text is exactly the same as in No. IV of

the Fourth Set ; see the diagram on p. 87. There are, however, four

exceptional pages. On page 107 there are two columns, each consisting

of four impressions of the formula (altogether eight repetitions), and

running parallel to the longer side of the book.

On pages 81 and 84, the arrangement is the same,

except that there is only one column in the middle

of the page. On page 106, as shown in the marginal

diagram, the two arrangements are combined, there

being two columns standing at right angles to each

other, the outer consisting of three, the inner of four

impressions of the formula, and running parallel to the shorter and

longer sides of the book respectively.

No. III. Book. (Plate XIII.)

Belongs to M. 8. Found at Kiang Tuz. Size, 13X11". Number

of forms uncertain, as the outside leaves are missing, the book, in its

present condition, beginning and ending with a printed page ; but

the existing number of leaves is 81. Paper, variety IIIc. Many of the

leaves torn, and otherwise damaged by fatty stains and burns. Print

often barely legible. Bouud with three twists of paper.

Page 106.

oq v

o L

rO

e o qv

ro oqv

e

1899.]

Section II. — Bloch Prints.

99

Contains tlie texts of the First, Second, Fourth, and Fifth Sets.

Those of tlie First and Second Sets never occur complete. Portions of

them only are found, and always standing separately. Of the First

Set the formulas I /*, I g, and I h (i.e , A, C, E mutilated) only are found ;

of the Second Set only the formulas A, B, D and E. In the distribu¬

tion of these texts, no particular order is observable. Sometimes the

text of a set occupies a whole page by itself ; at other times, two or

three (never all four) texts are combined on one page. This will bo

best seen from the Table of analysis of the pages, given below.

If a single text occupies a page by itself, it is always printed in 3

to 5 columns, generally running parallel to the longer side of the book,

and consisting of 4 to 6 impressions of the formula, according to the

size of the latter. See below the diagrams of pages 8, 9, 11, 12, 59, 60.

If a single text occupies several consecutive (from 4 to 12) pages, the

columns always stand upright and reversed on alternate pages.

If several texts together occupy the same page, they are arranged

in the most varying order. No principle that may have guided the

printer is observable. The subjoined diagrams illustrate this ; the

arrangements shown are only a few out of a much larger number.

The formulas of the Fourth and Fifth Sets are indicated by a b c and

12 3 respectively ; those of the First and Second Sets by I/, I y, I h ,

and A, B, D, E respectively.

Page 8.

/i/i/i /i /i

/ 1 /i/i /i/i

/ 1 /i /i /i /i

/i/i /i/i / 1

/i/i /i /i /i

/i /i/i /i / 1

Page 9.

1/ If If If If

If If If If If

HH t— I PH I— I (— t

^ ^

M I— 1 rH HH t-H

Page 11.

ab c ab c ab c

abc ab c ab c

abc abc abc

Page 12.

oqv oqv oqv

o qv oqv oqv

oqv oqv oqv

100

Dr. Hoernle — Antiquities from Central Asia. [Extra No. !r

Page 48.

abc £ oqv

abc h_i o q v

abc o qv

t— i

abc *** o q v

abc £ v

Page 49.

ab c Ig ab c

abc hh abc

abc abc

hH

ab c abc

ab c £ abc

Page 59.

^ '-s ^ '+-■> *+•*>

h- r t—i h-r h-r h-r

N-*, ^

hi t—r h-r i — r hi

m-.

Kh'h'h'h'

^ ^ ^ ^ ^

*+~» ^ ^ ^ ^

hi h-r h-r i— r t—r

^ *+-> *+-> *+-*

h-r h-r h-r h-r i— r

*+-^ s—. w

h-r h-r h-r h-r i — r

hV h-r h-r h-r h-r

Page 60.

< <! C ^

j— 1 r— - r— i r-i , h-t

'-H '-+» ^ ^

J— I J-H .hH >-l >H

*-+> ^~h. ^

. I— • H ,1—1 >H J— <

'"S '—H t~“S '"“S *~'S

fhH I ,hH ,1-H J— I ,h-t

r—i r— ■* .hh >—< .hi

^*S ,''H *'-S ^

r— I r-H .hH .hH >— I

'-s *-+» '-s

Page 47. Page 91.

Page 139.

a be J**

abc

123 05

abc 7

abc to

abc 123

t— *

abc abc

Co

Page 140.

Page 163.

Page 164.

123 123 ^

123 123 2$

123 123

123 123 co

123 123 3

123 123

123 123 e*

123 123 £.

fij Ife

fiI I h

*11

6t H*

1 I h

fir Ih

I h

fij I h

-4 If

< 1/

<1

V

1/

V

<1 1/

-4 ¥

-4 1/

1 9 ^ BI i;l

I 9 <j ^1 bl

I g < Bi Bl

Ig tel 6i fix

Iff *4 fij fij

Iff <i fix fix

!? -4 Bl B1

I have taken the trouble to analyse the whole of the pages with

reference to the frequency of occurrence of the several formulas and

their combinations. The result is shown in the subjoined Table.

The formulas of the Second Set ( A, B, 0, D), except where stated

otherwise, occur only once on the page.

Formulas.

Sets IY.

Set Y

Sets 1Y and Y

Set IY and Set I g

Set IY and Set Ih

• t •

1 1 •

• • •

# t •

Pages of Occurrence. Frequency.

11-18, 23-27, 30, 39-42, 51-58,

67-70, 75-82, 87-90, 92, 94,

99-102, 107-118, 131-134,

140-150, 155-162 Total

28, 29, 140, 141

139, 142

48, 49 ... • ...

80

4

2

1

1899.]

Section II. — Block Prints.

101

Formulas. Pages of Occurrence. Frequency.

The combination of formulas in different scripts on the same

page seems to suggest some sort of correspondence or identity. It is

difficult, however, to fit the varieties of the combinations into a con¬

sistent theory. Possibly this may be due to misprints. Provisionally

I would suggest the following theory. Books Nos. I and II seem to

show that the formulas of Set IIC, Set IY, and Set Y correspond to

one another. Now pages 47-50 show formulas IY and I g in combina¬

tion ; and it may be also noted that these two particular formulas

occur most frequently by themselves in book No. Ill ; viz., formula

IV, 80 times, and formula I g, 38 times. Assuming that formula I g

is equal to formula IY, the combination on page 2 {viz., II E and Ifg )

would tend to identify formula 1/ with formula HE. Consequently

the combination of formulas II A and II E with formulas 1/ and Ih

on page 4, would identify formula Ih with formula II A; and in

corroboration of this equivalence it may be noted that on pages 4 and

163, there are whole columns of II A and Ih corresponding with each

other. At the same time this theory does not seem quite consistent

with the combination of the formulas If, II A, and IY, on page 93. As

formula IY is assumed to be equal to formula I g, it seems to follow

that I/=IIA ; and with this conclusion agree the two facts that on

page 4 the formulas 1/ and Ih are combined with the formulas II A and

II E, and that on pages 4 and 163 columns of II A correspond to columns

102

Dr. Iloernle — Antiquities from Central Asia. [Extra No. 1,

of If. In fact, these two facts fit in equally well with both theories.

The result accordingly would be

formulas I V = V = II C = Ig ;

I/=IIE or = II A ;

lh=ll A or ^IE-

Seventh Set.

This Set comprises six books. The text contained in it consists of

seven lines. The dimensions of its block cannot be given, as no traces

of any enclosing lines are seen in any of the books. The dimensions of

the text itself are about 3£ x 3f

With regard to the question what is top and bottom, right and left,

beginning and end of the formula, the following circumstances may

be noted. In book No. 1, about one-half of the formula, divided

horizontally, is occasionally found ; similarly in book No. IV, also about

one-half of it, but divided vertically, is occasionally met with. The two

halves that are thus found are invariably the same. On the assumption

that, if only a portion of the formula could be accommodated, the

printer would naturally choose to print its initial portion, it follows that

the two halves, between them, accurately define the corner which

contains the beginning of the formula. This is the upper left corner or

the upper right corner, according as one has to read the lines of the

formula in the European or the Chinese fashion. In the facsimile on

Plate XIY, the formula is represented in the position conforming with

the view here explained. The argument, however, is by no means, con¬

clusive, as it pre-supposes that the books were intended for intelligent

reading, not merely for the mechanical turning of pages.

Another point that may be worth noticing is that some of the letters

of the formula of this Set and of that of the Fourth Set show consider¬

able similarity.

No. I. Book. (Plate XIV, fig. 2.)

Belongs to M. 9. Size, 17Jx6|". Number of forms, 20. Paper,

variety Hid. Well preserved ; no burns ; but paper greasy, and print

rather indistinct. Bound with three twists of paper.

Text printed in one column on each page, running parallel to

the longer side of the book, and containing, as a rule, four impressions

of the formula. This arrangement allows a wide margin at the top and

bottom of each page ; accordingly on a few exceptional pages (about

half a dozen) a half-impression is added to fill up the blank space. As

this is always the same half-impression (viz., lines 1-3 or 1-4), it may

possibly indicate the beginning of the formula; and I have so used it

1899.]

Section II. — Bloch Prints .

103

for the arrangement of the formula on Plate XIY ; but the argument

only holds good, if the book was intended for reading, which is doubtful.

As a curiosity I may note the arrangement on page 1 of form 8 ; I have

not noticed it elsewhere. Between the 3rd and 4th impression of the

formula, there is inserted an impression of the first line by itself. This

is probably a mere misprint ; though it is not quite easy to understand

how it happened. For as the uniformity of the intervals of the lines

shows, these lines were not printed each by itself ; on the other hand,

if the block was twice applied to the paper, in different places, the

paper should show smudges of ink, which it does not do.

The columns stand regularly, without any exception, upright and

reversed on alternate pages.

No. II. Book.

Belongs to M. 8. Found at Kiang Tuz. Size, 14x9". Number

of forms, 12. Paper, variety Hid. Surfaces greased, and print rather

indistinct; otherwise well preserved. Biveted with three nails.

Text, printed in two columns on each page, running parallel

to the longer side of the book, and standing alternately up¬

right and reversed on the same page, but keeping the same position on

alternate pages, as shown in the diagram on the margin. Each column

consists of four impressions

(1)

(2)

(3)

(I)

S

Page 4.

ab c o qv

def fop

fab c oq

l def fop}

v

ab c oqvj

lef fgp

abc o qv

i _ _ _ >

-V

(3)

(I)

(a b c d e /) of the formula,

but as the space is barely

sufficient to accommodate

them, not unfrequently one

or two lines of the formula

are omitted at the top or the

bottom of the page ; and,

whenever this occurs, the

mutilated portions are arran¬

ged so as to make up, be¬

tween them, a complete

formula. The intention, evidently, seems to be that the left hand

columns and the right hand columns respectively should be read

consecutively throughout the book, as indicated by the arrows in the

marginal diagram. The four impressions are placed so as to keep the

corresponding lines as much as possible on a level with one another.

No. III. Book.

Belongs to M. 8. Found at Kiang Tuz. Size, 12f x 8J". Number

of forms, 20. Paper, variety Illd. Surfaces greased, and print rather

104

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

indistinct ; otherwise well preserved. Bound with three twists of

paper. Printing stops on the penultimate, but begins, as usual, on

the fourth page.

The arrangement of the text is exactly the same as in No. II,

except that the columns only consist of three impressions of the formula,

whence a wide margin is left at the top and bottom of every page.

There are, however, four exceptional pages, probably misprints, on

which the columns run in the same direction.

No. IV. Book.

Belongs to M. 8. Found at Kiang Tuz. Size, 13 x 8f ". Number

of forms, 20. Paper, variety IIIcZ. Surfaces greasy, and print rather

indistinct ; otherwise well preserved. Bound with two nails.

Text printed in three columns on each page, running parallel to

the narrower side of the book, and consisting each of two impressions

(a b c d e f) of the formula. The columns, as a rule, run in the same

direction (e.g., pp. 4, 5); but there are many exceptions (<?.y., pp. S, 9),

as shown in the subjoined diagram.

Ordinary. Exceptional.

The three columns do not fill up the whole available space ; there is

a wide margin on both sides. Accordingly on 14 pages (out of a total

of 74 printed ones) a half-column is added, consisting of two half¬

impressions of the formula divided vertically. As a

rule, this half-impression stands on the margin of the

same side, but on page 16, shown on the margin, it

stands once on the left, and once on the right margin.

As it is invariably the same half of the formula, this

circumstance might be considered to indicate the

side with which the formula commences ; and 1

have used it for that purpose in arranging the

facsimile on Plate XIV. But, as already observed,

the argument is not altogether conclusive.

1899.]

Section II. — Block Prints.

105

No. V. Book.

Belongs to G. 10. Size, 17|xl4". Number of forms, uncertain,

as they are all cat through, along the folds, into separate leaves ; the

latter number 41 ; but possibly one leaf is torn off. One of the existing

outside leaves is torn in sbreads. Paper, variety III6. Surfaces greasy,

and print indistinct; otherwise fairly well preserved ; no burns. Print¬

ing commences on the second, and ends on the penultimate page of the

now existing leaves. Riveted with three nails, the guards being round

pieces of copper resembling coins, like fig. 4-9 on Plate IV.

Text printed in three columns on each page, running parallel to

the longer side of the book, and consisting each of four impressions of

the formula, which, accordingly, is repeated 12 times on every page.

The columns stand upright or reversed on different pages ; but there

is no perceptible order in this respect.

Page 44.

On one of the pages of the 22nd

leaf, there is seen the sketch of a

man’s bust, twice repeated side by side,

one somewhat smaller than the other.

See Plate XVIII. They were evi¬

dently sketched on the page, before the

formula was printed around them, as

the arrangement of its impressions is

adjusted to the sketches, running

regularly round them. The sketches

are placed horizontally across the

page, as indicated by the two figures in

the marginal diagram.

No. VI. Book.

Belongs to G. 8. Purchased for Rs. 45. Size, 22 x 13'\ Number of

forms, 43. Paper, variety Illc. Rather clean, but many leaves torn, on

account of large size and flimsiness of paper. Print not very distinct.

Riveted with three nails, the guards being regular oblong pieces (If X

II") with rounded corners, showing embossed head and symbols, as seen

in fig. I of Plate IV.

Text printed in three columns on each page, running parallel to

the longer side of the book and consisting each of six impressions of the

formula. They stand, as a rule, upright and reversed on alternate

pages.

J. i. 18

106

Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

Eighth Set. (Plates XIV, fig. 3 and XV, fig. 1.)

This set comprises only one book. It belongs to G. 9. It was

purchased by Sayyid Gul Muhammad, a well-known Kashghar mer¬

chant, for forty rupees and was sent, as a present, to Captain Godfrey.

The book, of course, could not be accepted as a present, but it was

purchased on behalf of the British Government. It measures 11 x6J".

The exact number of its forms is unknown, for the beginning and end

are missing, and a large number of leaves exist only in fragments.

The number of still complete forms is 29 ; most of these even are more

or less damaged along the edges. The book is bound with three copper

nails, and the guards are formed of two thin copper slips, measuring

8 x |", and covered with ornaments like those on figs. 4-9, Plate IV.

Irrespective of its script, this book strikingly differs in several

points from the books comprised in all the previously described seven

sets. In the first place, it is clean ; there is no trace of any burn or

fatty stain. In the second place, the paper, to all appearance, is of an

entirely different quality. It is thin and soft and more nearly resembles

the paper of the Weber and Macartney Manuscripts procured from

Kuchar. It differs, however, from their paper in colour ; for while their

paper is white or whitish, the paper of this book is of a bright yellowish-

brown. It looks as if it were artificially tinted ; but the colouring,

if any, is fast, for it is tolerant of washing. It is a pity that its find-

place is not known ; but that it comes from some spot in the Takla

Makan is shown by the fact of all its leaves being, like those of- all

the other blockprints, very thickly covered with the fine yellow sand

of the desert. Another curiosity is that a small special formula, which

occasionally occurs in it, is printed with an apparently faded, red-

coloured fluid, which almost resembles blood. Its ordinary formula is,

as usual, printed with black ink. Minor peculiarities are the following:

(1) most of the existing leaves show a clean cut on one of the narrow

sides, (2) two of the pages have the text printed diagonally across them,

and (3) a few leaves are only printed on one side. The last mentioned

peculiarity is due to the extreme thinness of the paper, owing to which

the print on one side shows through on the other. The leaves have, as

in the case of all other block-printed books, frayed edges, but in the

present case one of the narrower sides of most leaves has been clipped

with a sharp knife or scissors, for it shows a clean cut, which occasion¬

ally passes right through a line of print, showing that the clipping

was not done with little care.

The text of the book consists of two formulas which I shall call

VIII a and VIII b. The formula VIII a consists of three long lines,

containing apparently about 16 letters each. It is the proper formula

1899.]

Section II. — Block Prints.

107

of the book, as it covers every printed page but one. Formula VIII b

is evidently a special one ; it is very small, consisting of four lines, of

2, 3, 3 and 6 letters ; and it is only found on a very few pages. On one

page it is found twice, printed in the middle and at the top of it, the

rest of the page being filled with the ordinary long formula VIII a .

On two other pages it is found similarly at their top ; and lastly there

is one leaf, on which it occupies the entire surface of both pages.

Curiously enough this is an isolated leaf, which is stuck in between the

two leaves of a folded form. But from the page which exhibits the

double imprint of formula VIII b (see Plate XV, fig. 1) it is evident,

that both formulas were printed at the same time ; for the needful

space (though only just barely sufficient) is purposely left for formula

VIII b between the impressions of formula VIII a.

The latter formula is printed ten times in a column on each page ;

the column running parallel to the longer side of the book. Within

the column the impressions of the formula stand, as a rule, upright

and reversed alternately ; though occasionally two upright or two

reversed impressions follow consecutively, as may be seen on the

facsimile page in Plate XV, fig. 1.

Formula VIII b is also printed in a column consisting of ten lines

of impressions ; but each line itself is made up of four impressions,

standing alternately upright and reversed ; so that the formula is

repeated 40 times on each of the two pages of the leaf the surface of

which it entirely occupies. On all other occasions (as on the facsimile

page ) where formula VIII b occurs, it only occupies one line consisting

of four impressions.

Among the fragments, found by me with the book there are two,

which have a peculiar interest in bearing, in addition to the ordinary

formula VIII a, a second small text, which I shall call formula VIII c.

One of the fragments consists of a very narrow oblong sheet, folded in

the middle into two leaves. Each of these leaves (see fig. 3 on Plate

XIV) measures 6|x3" ; and shows a clean cut along either of its long

.sides. As these sides measure exactly the same as the breadth (or

narrow side) of the book; it seems probable that the whole oblong

sheet is simply a slip cut out of one of the forms of the book. And

seeing that the slip is nearly blank on one side, it is further probable

that the form, from which it was cut, was one of the outside, or

covering, forms of the book which are now missing. The other piece

is of a very irregular rhomboid shape, being apparently a piece torn

off one of the leaves of one the outside forms of the book; for it

shows on one side three full and one fragmentary imprints of the

ordinary formula Villa in the usual column arrangement, while the

108 Dr. Hoernle — Antiquities from Central Asia. [Extra No. 1,

other side must have originally been blank, but is now covered with

imprints of formulas VIII b and VIII c in a promiscuous and dis¬

orderly way.

Ninth Set. (Plate XV, figs. 2 and 3, and PI. XVI.)

This set comprises two items, a roll and a book. The latter, when

received, was enclosed in a carved wooden box ; and the former probably

was also originally within it. The whole belongs to G. 10. It was

received by Captain Godfrey from Leh, and is said to have been dug

out in the Takla Makan, which, seeing that it is more or less thickly

encrusted with the fine yellow sand of the desert, is probably correct.

But it is a pity that the exact find-spot is not known.

No. I. The Box.

The box (PI. XV, fig. 2) has a height of 4§"; its diameter externally

is 4J", and internally, 3|" ; inclusive of the projecting carved figures, its

breadth is 4f/'. It is drilled out of one piece of wood, and is ornamented

with six carved projections, which run, like pillars, round it parallel with

the length of its wall, and at equal distances (about 1'') from one another,

and consist alternately of standing human figures and inscribed boards.

Close to one of the figures, there is a crack right through the wall of

box, gaping asunder about J of an inch. Above the head of the next

figure to the right, there is a large semi-circular notch cut into the rim

(shown on Pl. XV), and there is also a smaller triangular one over the

inscribed board which stands between those two figures. These notches

seem to have been made intentionally. There are also two small,

irregular holes in the wall (one shown on PI. XV), nearly opposite to

each other, but these appear to be due to injury. There is no lid to close

the box ; nor do appearances point to its ever having had any. The

projections go down to the bottom of the box, but do not reach quite

to the edge of the rim, being short of it by -f of an inch.

Of the three human figures, one is represented with his arms

a-kimbo, his hands resting on his abdomen (shown on PI. XV), while the

other two figures have their arms hanging down straight by their sides.

There are some similar crude figures of copper in the collection, which

will be described in the section on Miscellaneous Objects. All three

figures on the box appear to be represented nude. Two of them

(including the one with the arms a-kimbo) bear curious lines marked

regularly across both sides of the chest and upper arms. They might

be intended to denote a short jacket ; but similar lines are used to mark

the hair on the heads of all three figures. This hair is marked very

regularly, long hair with a parting in the middle. One of the figures —

1899.]

Section 11. — Block Prints.

109

he with the arras a*kiuibo — has also a beard, marked by similar lines

all round the lower part of the head. The other two are represented

beard-less. The heads are made disproportionately large ; and

altogether the figures are very crude.

The three bands (PL XY, fig. 3) of writing are oblongs, measuring

about 3f X 1 ". One of them is divided, by indented lines, into three nearly

equal compartments. Their top and bottom seem to be clearly indicated

by their correspondence to the heads and feet of the figures. They are

shown on the Plate in the position thus indicated. Accordingly the

legend of No. II which consists of two lines containing each six

symbols must be read either from top to bottom, or from right to left.

The legend of No. I appears to consist of a narrow column of nine short

lines, each containing three or four symbols. The three compartments

of No. Ill seem to contain 3, 2, 3 short lines respectively. The prob¬

ability seems to be that all the legends run from the right to the left.

On Plate XY, fig. 4 I show an inscription which exhibits a

curious primd facie resemblance to the writing on the bands. This

inscription stands on a hone of slate, measuring 5 X 1". It was found

at Mazyhund, close to Tiran, at the foot of Mahaban in the Swat

country, and brought to Major Deane, who very kindly gave it to me

to be added to the British Collection of Central Asian Antiquities.

No. II. This Boll. (Plate XYI).

The roll measures 16f by 4f inches. The paper is very different

in texture from that of the block-prints books comprised in Sets I to

YIII. It is exceedingly thin, tough and hard ; it is also oiled or

greased, apparently as a kind of sizing, to tolerate being printed on.

When washed, it shows a very light yellowish or creamy tint. In

general appearance it resembles thin parchment. It is only printed on

one side, the paper being so thin that the print of one side shows

through on the reverse, wherever there is an excess of ink, as in lines

8-11, 30-33, 38, 39. For that reason, clearly, the ink was, as a rule,

put on very sparingly, so that in many lines the print is so fine as to be

almost illegible.

The roll is covered with 45 lines of print, which run parallel to

its narrow sides, and which contain each from 13 to 15 symbols. A

closer inspection reveals the fact that this text of 45 lines consists of

five formulas, which are repeated at irregular intervals, and each of

which comprises two lines of the text. I shall distinguish these five

formulas as IX1, IX2, IX3, etc. The two lines of the several formulas

are made up of a number of symbols varying from seven to fourteen.

Sometimes, as in lines 8 and 9, which comprise formula IX4, the lines

110

Antiquities from Central Asia.

[Extra No. 1, 1899.

of the formula practically coincide in length with the lines of the text.

In other cases, as in lines 20 and 21, comprising formula IX2, the lines

of the formula are much shorter than those of the text. In these cases

the latter lines are filled up with repetitions of the formula, in a more

or less complete state.

Formula IX1 occurs five times, in lines 2 and 3, 16 and 17, 24 and

25, 34 and 35, 44 and 45. Accordingly, considering that there are two

columus, the formula is repeated fen times. In lines 2 and 3, it stands

reversed ; in all the other lines, it stands upright.

Formula IX2 occurs six times, in lines 4 and 5, 10 and 11, 20 and

21, 22 and 23, 30 and 31, 38 and 39. Altogether it is repeated twelve

times on the roll. This formula and the fourth are the onty ones in

which the symbols stand sufficiently apart to permit of being discrmi-

nated and counted. Its first line consists of eight, and its second line,

of seven symbols. The second (or sixth, according as the series is read

from the left or right) symbol of the latter line has a striking resem¬

blance to the Sanskrit (Brahmi) letter for a, as written in Nortli-India

about 800 A.D. ; but this must be a mere accidental coincidence, as no

resemblance can be seen in any of the other symbols.

Formula IX3 occurs three times, in lines 6 and 7, 12 and 13, 14 and

15. The number of symbols comprised in its two lines is uncertain ;

probably 11 and 13 respectively.

Formula IX4 occurs twice, in lines 8 and 9, 32 and 33. Both lines

appear to consist of eleven symbols.

Formula IX5 occurs six times, in lines 18 and 19, 26 and 27, 28 and

29, 36 and 37, 40 and 41, 42 and 43. In the two pairs of lines 18, 19

and 26, 27 it stands reversed ; in the other four pairs it stands upright.

The two lines appear to comprise 12 and 14 symbols respectively.

No. III. The Book.

Belongs to Gf. 10. Size 6x44,h Number of forms 34| ; many cut

into separate leaves ; number of leaves 69. No blank covers. Many

leaves torn. Paper rather brittle, and of the same kind as that of the

Boll No. II. Stitched with four loops of thread.

Contains the identical text of the roll ; as a rule, arranged in

columns, running parallel to the narrower side of the book ; but on a

few exceptional pages, they run parallel with the longer side. As in

the roll, the formulas are repeated at irregular intervals, two or three

formulas being found repeated on each page.

•Jr%pik

M'Mh/M?

turkistan

THE Repoet on

N antiquities

' A\'/,

* Hr V

,\{/4

Find-places

-S'ea/e 7 /„c;t

40 iftfo,.

MAX''

*9° Mites

> > 1 1 1 r,''

it/,

«#w

IldLol

1 1

ii, o ’ * 1 1 1 ■

Wiv""1*"'

NewIopNoj;

S""

^wmW*

Mcu'a.lb a s hi.

. ^Kcutaht

SvenHedin’s

Old/ Guty

Yarkand

'YrsVuXj

Q'oVi

(Old/ City)

%f PiT^

^ Sven H edxn's

' Old. COy

(him i

Jnbu (fhsrr

% TaLa/Tux

} 'Dundarv Ux lxq-

OutaiJTiliif

''l\ i

xum

Qash/ *jti(fSapih

**W Qwcai

Kianp Tux ^

Niya/ fX

MjiyLqh.il

KHOT;

/^Ww.'V ■

'^:o

^‘■llUViUUVU.

SWUllK

Annu'/yM"',

tyuir&ifiw

SWWV.feajm

Jfuy

MEMOIR

ON MAPS ILLUSTRATING

THE ANCIENT GEOGRAPHY

OF

KASMIR.

BY

M. A. STEIN, Ph.D.

PRINCIPAL, MADRASAH COLLEGE, CALCUTTA.

JOURNAL OF THE ASIATIC SOCIETY OF BENGAL, YOL. LXYIII,

PART I., EXTRA-NUMBER 2. — 1899.

CALCUTTA :

PRINTED AT THE BAPTIST MISSION PRESS

AND PUBLISHED BY THE

ASIATIC SOCIETY, 57, PARK STREET.

1899.

“ ... quae loca fabulosus

Lambit Hydaspes

Q. Hor. Flacci, Od. T, xxii.

N.B.— For Table of Contents and List of Abbreviations

see pp. 223-231.

JOURNAL

OF THE

ASIATIC SOCIETY OF BENGAL.

Part I. -HISTORY, LITERATURE, &e.

Extra-Number 2. — 1899.

Memoir on Maps illustrating the Ancient Geography of Kasmir. —

By M. A. Stein, Ph.D.

CHAPTER I.

INTRODUCTORY.

1. The maps accompanying this memoir are primarily intended

to show the results which a detailed study of

Ancient topography Kalhana’s RajataranginI has furnished re¬

in Kasmir Chronicle. , , • , „ _ , .

garding the early topography of Kasmir and

the adjacent territories. From the first when engaged in preparing a

critical edition of that text,1 2 * * the earliest and most important of the

Sanskrit Chronicles of Kasmir, I had realized that an exact identifica¬

tion of the very numerous old localities mentioned in it was indis¬

pensable for a correct understanding of the narrative. This conviction

forced itself even more strongly upon me in the course of the labours I

devoted to the preparation of the commentated translation of the work

which is now passing through the press.8

1 Kalhana’s RajataranginI or Chronicle of the Kings of KaSmir , Bombay Educa¬

tion Society’s Press, 1892, pp. xx and 296, 4to.

2 To be published, with a Historical Introduction, by Messrs, A, Constable and

Co,, London, in two volumes, 4to.

J. I. 1

2

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Many of the questions thus raised were so detailed and intricate

that it would have been manifestly impossible to attempt their solution

without carefully studying on the spot those topographical facts which

alone could elucidate them. It was, therefore, fortunate for my re¬

searches in this direction that I was able during successive years to

make a series of antiquarian tours in Kasmir. These acquainted me

not only with the extant ancient remains of the Valley, but also with

its actual topography and that of the neighbouring mountain regions.1

I cannot feel too grateful for the advantage I thus enjoyed. It has

allowed me in more than one case to fix with certainty the position of

important ancient sites, which no amount of philological acriby would

have sufficed to locate correctly.

2. In order to place before the student of the Kasmir Chronicle

the results of these researches, as well as the

Prepaiation of evidence on which they were based, the pre-

^■a^S* paration of maps appeared necessary that

would show the modern topography of the country in full detail together

with the ancient sites and local names identified. While considering

the means for the execution of such maps I received in the autumn

of 1896 the generous offer of the Asiatic Society of Bengal to bear

the cost connected with their preparation, on the understanding that

the maps would be published also in the Society’s Journal with a

separate explanatory memoir. I accepted this offer all the more

readily as it gave me the desired opportunity of treating the subject

of the early geography of Kasmir in a connected form and before ' a

larger public. For the liberal assistance thus rendered to me, I wish

to record here my sincere thanks. I owe special obligations to Dr.

G. A. Grierson, c.s., c.i.e., and Dr. A. F. R. Hoernle, c.i.e., whose

kind offices were mainly instrumental in securing the above arrange¬

ment.

The successful execution of the maps as now published was

rendered possible by the ready co-operation of Colonel J. Waterhouse,

i.S.C., late Assistant Surveyor General, in charge of the Lithographic

and Photographic Office of the Survey of India Department. The

preparation of a new ground map to show on a sufficiently large scale

the details of the modern topography of Kasmir would have cost much

trouble and entailed very heavy, almost prohibitory, expense. At the

same time it had to be considered that there were no other materials

available for such a map but those supplied by the Trigonometrical

1 The tours referred to occupied the greatest part of my summer vacations

in 1888, 1889, 1891, 1892, 1894 and were supplemented by shorter visits to parti¬

cular sites during the summers of 1895-96.

1899. J

INTRODUCTORY.

3

Survey operations in Kasmir, 1856-60, which had been embodied on the

scale of 4 miles to 1 inch in the corresponding sheets of the ‘Atlas of

India.’

It hence appeared to me the most convenient plan to use as a

ground-map a mechanical reproduction of that portion of the ‘ Atlas

of India’ which contains Kasmir and the adjacent territories. Over

this ground-map the entries relating to the ancient topography of the

country could be printed in a distinguishing colour. This plan having

received Colonel Waterhouse’s approval, the required portions of the

engraved plates containing Sheets 27 and 28 of the ‘ Atlas,’ were trans¬

ferred to the stone and the copies of the larger map reproduced from

the latter by lithography. A similar process was used for the smaller

map showing the capital, S'rinagar, and its environs on the enlarged

scale of 1 mile to 1 inch. But in this case the original map which was

not engraved but only zincographed, had to be retraced on the stone.1

In the case of either map the entries marking ancient sites and names

were printed in red over the ground-map from a separate stone. In

order to distinguish at a glance the old local names iu the Raja-

tarangini from those known only to the later Chronicles and other sources,

the former were shown in Grotesque type and the latter in Italic.

By following the method here briefly explained it was possible to

provide maps which exhibit in all needful detail the latest and most

authentic survey of Kasmir and at the same time show clearly all im¬

portant features of the old topography. The success of the technical

execution is due mainly to the great care and attention bestowed on it

by Colonel Waterhouse and his staff. For this as well as much valu¬

able advice accorded to me in connection with the work I may be

allowed to offer here my grateful acknowledgments.

3. It has already been stated that the maps here presented are in

the first place intended to illustrate those data

Arrangement of 0f ppg anc{ent geography of Kasmir which are

memoir. contained in Kalhana’s Chronicle. But in

addition to the old local names and sites taken from this our most

important source of information those mentioned in the later Sanskrit

Chronicles and other Kasmirian texts have also been inserted as far as

they can claim antiquity and interest. These maps may, therefore,

equally well serve to illustrate a comprehensive account of the historical

topography of Kasmir, such as I shall attempt to give here, up to the

close of the Hindu epoch.

1 The original of this smaller ground-map had appeared as an inset in the

“ Map of Jummoo, Kashmir and Adjacent Territories ” 4 miles to 1 inch, published

by tire Survey of India, 1861.

4

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

In treating tliis subject it appears to me most convenient to exa¬

mine first the sources of information from which our knowledge

regarding the ancient topography of Kasmir is drawn. I shall next

proceed to notice what we can learn from these sources as to the general

physical features of Kasmir geography and their bearing on the histo¬

rical and economical conditions of the country during the Hindu period.

In the last chapter I intend to discuss the political divisions of the

territory and to indicate briefly the information available to us regard¬

ing the particular places of historical or religious interest.

Most of the data upon which this account of the old topography of

Kasmir is based, are contained in Kalhana’s Chronicle, and have there¬

fore already been explained by me in the notes which accompany my

translation of that work. In order to avoid unnecessary repetition I shall

refer to these notes for all such detailed evidence as could not con¬

veniently be set forth within the limits of the present paper.

4. Before, however, closing these preliminary remarks it is neces-

_ sary to refer briefly to those few publications

Earlier publications. • • i £ , i • n ,, .

m which tacts bearing on the old topography

of Kasmir have received an earlier treatment.

In view of what has been said above we naturally turn first to the

works which have dealt directly with the interpretation of the Raja-

taranginl. Dr. W ilson who was the first European to study the Chronicle

in the original, published an abstract of the contents of the first six

Books as early as 1825. 1 He seems to have fully realized the importance

of an accurate and sober examination of the geographical questions

connected with the narrative. The textual materials at his disposal

were, however, extremely defective, and European knowledge of Kasmir

restricted at the time solely to the accounts of Bernier and Forster.

He could hence scarcely do more than indicate the more or less corrupt

modern equivalents by which the Persian Chronicles render some of the

Kasmir local names taken from Kalhana’s account. The judgment and

accuracy with which Dr. Wilson discussed the Chronicle’s notices of coun¬

tries and places situated outside Kasmir and better known at that time,

shows sufficiently that only the defective character of the available

materials prevented that distinguished Sanskrit scholar from doing

justice to the task.

The elaborate commentary with which Mr. Troyer accompanied his

French translation of the Rajatarangini,* does not represent any mate¬

rial advance beyond the contents of Wilson’s Essay. Yet Mr. Troyer

1 An Essay on the Hindu History of Cashmir, in Asiatic Researches , vol. xv., 1825,

pp. 1 sqq.

8 Radjatarangim on Histoire des rois du Kachmtr, Paris, 1840-52.

1899.]

INTRODUCTORY.

5

liacl already the whole text of the Chronicle to refer to, and in the

meantime a considerable amount of information about Kasmir had

become available through tbe works of travellers like Moorcroft,

Jacquemont, Vigne, Von Hiigel, and others. The serious shortcomings

which characterize Mr. Troyer’s labors notwithstanding his patient devo¬

tion to the task, have already been fully indicated by Prof. Biihler.1

Detailed reference to the defects of the topographical notes is hence

unnecessary.

The English translation of the Chronicle published in the years

1879-87 by Babu Jogesh Chunder Dutt makes no attempt whatever to

elucidate the many points of topographical interest. Though the trans¬

lation itself is decidedly better than that of Mr. Troyer, yet it necessarily

shares the defects arising from the use of the same corrupt text. Both

versions strikingly demonstrate the importance of topographical re¬

searches by the frequent instances in which the translators have mis¬

taken local names for words of ordinary meaning or vice versa .2

The advantages offered for enquiries of this kind by a personal

acquaintance with the country were fully illustrated by the valuable

contributions which General (then Captain) Cunningham was able to

make to our knowledge of ancient Kasmir in connection with his visit

to the Valley in November 1847. Though his stay was short and pri¬

marily devoted to a survey of the more conspicuous of the temple-ruins

still extant, he succeeded in identifying correctly a number of important

ancient sites such as Puranddliisthana 1 the old capital,’ Jyesthesvara,

Mdrtdnda, Padmapura , Pattana , Khonamvsaf

General Cunningham subsequently had occasion to discuss compre¬

hensively these localities in his Ancient Geography of India , a work

which, notwithstanding its deficiencies in detail, amply testifies to the

great antiquarian experience and natural acumen of its author. The

chapter on the “Kingdom of Kashmir ”4 utilizes the evidence afforded

by the Chinese sources and Alberuni, and indicates correctly the old

names of the petty hill states to the south and south-east of Kasmir

( Rdjapuri , Valldpura , Gampd , Kdsthavata ). It further adds to the

identifications already mentioned equally important notes on Pravara-

pura , the present STinagar, Vijayesvara , HusTcapura , JnsJcapura, Jaya-

pura. If General Cunningham was less successful in his attempts at

1 See Report on a tour in search of Sanskrit Manuscripts made in Kashmir

J. Bo. B. R. A. S. 1877, pp. 55 sqq.

2 For some of the imaginary territories and places which figure in these trans¬

lations, see Vienna Oriental Journal , 1898, pp. 67 sqq,

8 See his Essay on the Arian Order of Architecture as exhibited in the temples of

Kashmir, J. A. S. B., 1848, pp. 242-327.

* See Ancient Geography of India, 1871, pp. 89-103, 128-141.

6

ANCIENT GEOGRAPHY OF K ASM III. [Extra No. 2,

locating Parihasapura and some other ancient sites, this may fairly be

attributed to his inability to consult the Sanskrit sources in the ori¬

ginal.1

Professor Lassen’s “Indische Altertliumskunde ” gives an exten¬

sive analysis of the historical contents of Kalhana’s work.2 But his

explanations as to the ancient localities mentioned are generally only

there well-founded where they are based on General Cunningham’s

researches. Ancient territories and places are often connected with

modern localities merely on the ground of a faint resemblance of the

names and without sufficient internal evidence. This tendency has

frequently led that distinguished scholar to ignore the narrow territorial

limits within which most of the local and ethnic names occurring in the

later portion of Kalhana’s narrative have to be looked for. It is only

natural that identifications of real (or imaginary) localities which trans¬

ferred the scene of contemporary events described by Kalhana to

territories so distant as Lahore, Eastern Afghanistan or Ajmir,3 have

helped to produce a very ill-focussed picture of the political power and

extent of the Kasmir kingdom in those later times.

The merit of having definitely shown the right methods and means

for re-constructing the ancient geography of Kasmir belongs to Professor

Buhler. This great scholar by whose lamented death so many branches

of Indian research have suffered irreparable loss, had in the masterly

1 If particular proof were wanted to show that a through acquaintance

with the modern topography of a country is in itself not sufficient to lead to

useful results in regard to its historical geography, Mr. Vigne’s work, Travels

in Kashmir, Ladak, Iskardo, (London, 1842, two Yols.) would supply it. This

estimable artist and traveller evidently took a great deal of interest in the

antiquities of the country which he traversed in many directions. His book,

however, as far as the old geography of Kasmir is concerned, furnishes scarcely

anything more than a series of amusingly naive etymologies of local names. Tlius

Hiir<}pdr (S'urapura) is “ the Diamond City,” Pdndrethan (Puranadhisthana) the

place of the ‘ Pandus and Duryndun ’ ( i . e., Duryodhana), Sopur (Suyyapura) ‘the

Golden City,’ etc. ; see i. p. 267, ii. pp. 37, 157.

Mr. Vigne is responsible for the strange derivation of the name of the Kasmir

capital, Srinagar (S'rinagara), or as he spells it, ‘ Siri-Nagur,’ from “ Surya Nagur ,

the city of the sun” (p. ii. 137). Judging from the persistence with which the

error has been copied by a succession of modern writers on Kasmir, this etymology

bids fair to establish itself as a piece of orthodox creed with European visitors to

the Valley.

2 See Indische Alterthumskunde (second ed.), ii. pp. 885-915 ; iii. pp. 984-1128.

3 I refer to locations like those of Lohara (Loh^rin) at Lahore ; of the

[imaginary] province Kampand in eastern Afghanistan ; of the Lavanya tribe near

the Sambhar salt lake ; of the feudal chief Kosthesvara at Kotgarli on the Satlej,

etc. ; comp. Ind. Alterth. iii. pp. 1057, 1041, 1069, 1105, and for the supposed terri¬

torial extent of the Kasmir state, iii. p. 1119.

1899.]

INTRODUCTORY.

7

report on liis Kasmir tour lucidly set forth the work that remained to be

done in connection with the Rajataranginl.1 He had there shown that

for a full comprehension of its contents a minute study of the ancient

geography of Kasmir was indispensable. He was the first to call atten¬

tion to the ample materials which are offered for such a study by the

later Sanskrit Chronicles, the Nilamatapurana and other Kasmirian

texts. But he also realized that “ some of the geographical questions

would probably require a final re-examination in Kasmir.”

Other labors prevented my lamented master from undertaking this

task himself. But the most graphic and accurate notices which his

Report gives of those sites in the Valley he had himself been able to

visit,2 prove convincingly — if any proof were needed — that no impor¬

tant point connected with the old topography of the country could easily

have escaped his attention. The particular identifications first made

by Prof. Biihler will be duly mentioned in their proper places. It was

a source of true satisfaction to me that I was able during my last year’s

visit to Europe to present personally the departed with the first clean

copies of the maps now published. That the results recorded in them

were such as obtained his approval, will always appear to me the

highest reward for the labour their preparation and the preceding re¬

searches had cost me.

1 See Report on a tour in search of Sanskrit manuscripts made in Kashmir ,

Bombay, 1877, p. 58.

2 See loc. cit.y pp. 4-18.

s

ANCIENT GEOGRAPHY OF KAS'AllR, [Extra No. 2,

CHAPTER II.

ACCOUNTS OF OLD KAS'MlR.

Section I. — Classical Notices.

5. Our sources for the early geography of Kasmir may be con¬

veniently divided into foreign notices and indigenous records. As the

information supplied by the former is on the whole earlier in date

though by no means more precise or important, we shall commence our

review with them. Having learned what little the outer wrorld knew

or recorded of the secluded alpine land, we shall appreciate all the more

the imposing array of Kasmirian authorities which offer themselves as

our guides iu and about the Valley. With the foreign accounts but

in a kind of intermediate position we may class those Indian texts the

authors of which may have possessed some more detailed information of

Kasmir, but have not thought it necessary to vouchsafe it to us.

It is significant for the isolated position which its mountain barriers

assured to Kasmir, that we do not find any

mention of the country in those accounts to

which we are accustomed to look for the first

truly historical notices of the North-West of India. I mean the relations

of Alexander’s invasion. The march from Taxila to the Hydaspes

(Jehlam) took the Macedonian forces along a line of route which lay

comparatively near to the confines of Kasmir. Yet there is no notice in

the accounts of Alexander’s expedition which can be shown to imply

even a hearsay knowledge of the Kasmir Valley. On the other hand

the names of the neighbouring territories on the West and South have

long ngo been recognized in the names of their rulers Arsakcs and

Alexander’s

invasion.

1890.]

CLASSICAL NOTICES.

9

Abisare.s. These names clearly represent ethnic appellations derived

from Uras'd (Ptolemy’s Ovapcra ) and Abhisdrci.1

The only certain reference to Kasmir which classical literature has

, , T_ . . preserved for us, is found in Ptolemy’s Geo-

Ptolemy’s Kaspema. There can be nQ donbt that D.Anvi]le

was right in recognizing its name in that of the region of Kacr7rapia

situated ‘below the sources of the Bidaspes (Vitasta) and of the

Sandabal (Candrabhaga) and of the Adris (Iravati)’.2 Ptolemy men¬

tions this territory correctly enough between that of the Daradrai or

Dards on the Indus and Kylindrine or the land of the Kulindas on the

Hyphasis (Bias) and eastwards. In his subsequent detailed description

of Indian territories, however, he makes the region ‘held by the

Kaspeiraaans ’ extend eastwards from the land of the Pandoouoi on the

Bidaspes as far as Mount Ooindion or the Vindhya.3

It is clear that the limits here indicated which would embrace

a great portion of the present Panjab with parts of the North-West

Provinces and Central India, can have nothing to do with Kasmir.

It has been suggested that Ptolemy’s statement refers to a period when

the power of the dynasty ruling over Kasmir actually extended over

the wide territories above indicated.4 The assumption, put into a form

more in keeping with historical probability, would be that Kasmir was

then subject to a great foreign dominion the rulers of which, for one

reason or the other, were in Ptolemy’s source designated from this part

of their realm.

However this may be, it is curious to note that we meet with the

name Kdo-7retpa also in the long list of cities located within the region

belonging to the Kaspeirseans. The geographical position assigned to

it by Ptolemy’s table (or map) would bring Kaspeira close to the junc¬

tion of the Hydaspes and Zaradros (Satlej), i.e., the neighbourhood of

Multan.5 Yet it seems difficult to believe that the information origin¬

ally underlying this entry referred to any other locality but Kasmir.6

1 See Lassen, Ind. Alt., ii. p. 174 ; Wilson, Essay, p. 116 ; also my notes on

Rajatar. i. 180 ; v. 217.

2 See Ptolemy vn. i. 42 and pp. 21, 40 sq. in Antiqnite Geographique de Vlnde ,

par 41. D' Anvil le, Premier Geographe du Roi, etc , Paris, 1775. — The accuracy and

sound judgment displayed in this work fully justify the great fame it has enjoyed.

3 Ptolemy, vn. i. 47.

4 Compare, e.g., Lassen, Ind. Alt. ii. p. 898 ; Y. de St. Martin, Memoires de

V Academie des Inscriptions, Sav. etrang., Ire Serie, t. v., p. 880.

6 See the old map reproduced in Dr. McCrindle’s Ancient India as described

by Ptolemy, Bombay, 18S5.

6 This had been rightly seen already by D’Anville. He points out, p. 40,

that the error in latitude implied by Ptolemy’s position of Kaspeira (if STmagar

J. i. 2

10

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

It would be useless to attempt to seek now for an explanation of tlie

erroneous location. The researches of the most competent scholars

liave amply proved how little reliance can be placed on the apparent

exactness of Ptolemy’s latitudes and longitudes in the Asiatic portion

of his work.1 None of the other city names in the same list can be con¬

nected with Kasmir. Nor is the identification of any one of them certain,

expect that of MoSovpa rj w ©€a>v, the sacred Mathura. This alone

suffices to show how far away from Kasmir we are liable to be taken.

The value of Ptolemy’s notice of Kaspeiria lies mainly in the fact

that it presents us with an accurate enough transcription of that form

of the country’s name which on independent phonetic evidence we must

assume as an intermediate stage between the Sanskrit Kasmlra and the

modern Kasmiri form Kasir. The explanations given below (§ 36) will

show that a well-established phonetic law presupposes a form *Kasvira for

the earlier Prakrit stage of Kasmiri. Of this form we have in Kaspeira

(pronounced Kaspira) as close a rendering as Greek waiting permitted.2

The Sanskrit form of the name, Kasmlra , has, as far as we can go

back, been always the one in official use. By it the country lias been,

and is still to this day, generally known abroad (Hindi Kasmir, Persian

Kashmir.) The preservation of the popular Prakrit *Kasvira by

Ptolemy deserves hence attention with regard to the original source

from which this particular item of information was obtained.

6. It is very probable that we have also to connect with Kasmir

a curious notice which Stephen of Byzance has

Kaspeiroi of Diony preserved from the Bassarika. a lost poem

sros and Nonnos. r . r

ot Dionysios ot Samos. The passage, first

apparently noticed by D’Anville, mentions the Kaspeiroi as a tribe

famous among all Indians for their fast feet.3 We do not know the

is really meant) is not greater than that which can plainly be proved in the case

of his entry for Barbarei , the port at the mouth of the Indus.

1 I cannot refrain from quoting here in full the very just remarks of

Sir Henry Yule, Cathay and the Way Thither , p. cli, which ought ever to be

remembered by those who have to deal with Ptolemy on Indian soil. “ We see

here how Ptolemy’s Asiatic Geography was compiled. It is evident that he first

drew his maps embodying all information that he had procured, however vague

and rough it might be. From these maps he then educed his tables of latitudes

and longitudes and his systematic topography. The result is that everything

assumes an appearance of exact definition ; and indications on the map which

meant no more than (somewhere hereabouts is said to be such a country), became

translated into a precision fit for an Act of Parliament.”

2 Tims the tribal name Aspasioi of Arrian (iv. 23) reproduces the Sanskrit

AsvaTca ; comp. McCrindle, Invasion of India, p. 333.

3 The text of the passage is reproduced by Trover, ii. p. 307. Another short

quotation from the same text mentions the Arienoi along with the Kaaneipoi.

1899.]

CLASSICAL NOTICES.

11

time of this Dionysios. Nor is there any indication as to the source

from which he may have taken the reference. That the Kasmiris had

abroad the reputation of being good pedestrians may be concluded from

a remark of Albeiuni.1 It is clear that the natural conditions of an

alpine valley enclosed by difficult mountain ranges are likely to develop

the marching powers of its inhabitants. The Rajatarahgiin gives us in

fact several instances of very respectable marching performances. It

shows at the same time the scant use made of riding animals in the

mountains.2 There is thus more than the mere name to justify us in

referring the notice of Dionysios of Samos to Kasmlr.

We meet with the name of the Kaspeiroi also in the Dionysiaka of

Nonnos. There they are mentioned among the Indian tribes rising in

arms against Bacchos.3 As Nonnos’ list names in the same passage

also the Arienoi whose name we see coupled with that of Kaspeiroi

in the fragment of the Bassarika, it is probable that Nonnos has

taken his referencej either from the latter work or from some common

source.

Kaspatyros of

Herodotos.

7. We should, indeed, have a far earlier reference to Kasmlr in

classical literature, and one by no less an

authority than the ‘ Father of history,’ if the

opinion of those scholars could be accepted who

h ive thought to recognize the name of the Valley in the Kaspatyros of

Herodotos. The facts are briefly the following. Herodotos mentions

the city of Kaspatyros as the place at which the expedition under

Scylax of Koryauda, sent by Darius to explore the course of the Indus,

embarked.4’ He distinctly places this city in the Paktyan land

(H aKTvtKT] yrj). This is described as being to the north of the other

Indians and apparently bordering on the Baktrian territory. The place

meant by Herodotos is evidently the same that Hekataios knew before

him by the name of Kaspapyros and as a city of the Gandarians.6

The notice of Hekataios (circ. 549-486 b.c.) makes it clear that

Kaspatyros or Kaspapyros, whichever form may be more accurate,

must have been situated in that territory where the Indus first becomes

navigable, i.e., in the ancient Gandhara, the present Peshawar District.

That the designation P aid y ike used by Herodotus refers to the same

1 India, trail si. Sachau, i. p. 206.

2 Compare Rdjat. vii. 140, 1301 ; viii. 192, 379, 1588, 1796, 1887, 2673 sq.

5 See Dionysiaka, xxvi. 165 sqq. I take this reference from Troyer, ii. p. 308.

4 See iv. 44, also iii. 102. The points bearing on the interpretation of the

passage have been fully discussed by Sir E. H. Bunbury, History of Ancient Geogra¬

phy, i, pp. 228, 256.

6 See Stephanos Byzant., s.v. TANAAPIKH ; also Muller’s Fragmenta historic,

graec., i. p. 12.

12

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

territory and represents the earliest mention of the ethnic name

Pakhtun or the modern Indian Pathdn, seems also probable,1 The

exact site of Kaspatyros lias not been identified. Considering the great

changes which local nomenclature in Gandhara has undergone, it

perhaps never will be.2

Wilson was the first who distinctly attempted to connect the name

of Kaspatyros with Kasmir. s But the idea seems to have occurred

earlier. For D’Auville thought it necessary to refer to it and to refute it.

Wilson saw clearly enough that the city of Scylax must have been

situated close to the Indus and hence far away from Kasmir. If not¬

withstanding this important fact he yet proposed to identify its name

with that of Kasmir, on the assumption that the borders of the latter

kingdom extended as far as the Indus, the mistake must be traced to a

fanciful etymology of the latter name.

Wilson assumed that the name Kasmir was derived from * Kasyapa-

pura , a name which he supposed to have been given to the country

owing to its colonization by the Rsi Ka^yapa. He supported this

strange derivation by a reference to the uniform assertion of ‘ Oriental

writers.’4 But it is difficult to believe that he could have meant any

1 This identification seems to have been first made simultaneously by Dorn

and Lassen ; compare V. de St. Martin, E'tude sur la gdographie grecque de VInde,

Mem. de l’Acad. des Inscriptions, Sav. E'trang., He Seine, V., p. 17 sqq. His

note on Kaspatyros, ib. pp. 81—36, contains a judicious review of the whole question

from the geographical point of view and a detailed account of earlier opinions. For

a more recent resume compare Darmesteteii, Chants Populaires des Afghans, pp.

clxxx sqq.

2 Proper navigation begins now at Jahanglra, a place situated on the left bank

of the Kabul ltiver, some six miles above the confluence of the latter with the

Indus at Attock. The lower Dart of the Kabul River’s course lies in a well-defined

i

single bed which, in view of the natural configuration of the banks, cannot have

changed materially in historical times. Above Jahanglra the current becomes too

strong for safe navigation.

I doubt very much whether the Indus immediately above Attock cau ever have

beeu suitable for proper navigation. The river is cut up there into many, often very

shallow, channels and obstructed by continually shifting sandbanks. On the eastern

bank spreads the low plain of Chuch, which must have always left a wide scope to

the vagaries of the great river. Taking into account these circumstances I should

not be surprised if Scylax’s expedition had chosen some place near Jahanglra for

the start on their voyage. There are many ruined sites near the latter place, and

near Alladher closeby on the Indus.

3 See Essay, p. 117 ; for a reproduction of the argument, also, Ariana Antiqua,

p. 136 sq.

4 “ This (the name of Cashmir) was derived, it is uniformly asserted by the

Oriental writers, from the colonization of the country by Casyapa, the first settle¬

ment or city being named after him Casyapapur, converted in ordinary pronuncia-

1899.]

CHINESE RECORDS.

13

better authorities than the Persian Tayikhs of Kasmir, of the 17th and

18th century, which he had occasion to consult in connection with his

above-quoted Essay. They, indeed, indulge in whimsical etymologies

like Kashmir, i.e., Kashap (Kasyapa) 4- mar (matha), etc. But nei¬

ther these etymologies nor the name * Kasyapapura are in any way

known to our genuine sources.

Wilson would scarcely have chosen to put forth such a deriva¬

tion, had the whole of the Chronicle or the other Kasmirian texts been

at the time accessible to him. Extensive as this literature is, it does

not furnish any evidence whatever for * Kasyapapura or a similar name

having ever been used as a designation of the country. This fact is all

the more significant as allusions to the legendary origin of the country

are otherwise so frequent. The philological impossibility of deriving

Kasmira from * Kasyapapura need scarcely be specially indicated at

the present day.1 A reference to the theory was, however, here neces¬

sary, as it has found its way into works of authorities like Ritter,

Lassen and Humboldt, and has hence been reproduced even by recent

writers.2

Section II. — Chinese Records.

8 If classical literature has thus nothing to tell us of Kasmir but

the bare name, it is very different with the

Earliest^ Chinese « Chinese records. Buddhist pilgrims from

China on their way to the sacred sites of the

Indian plains visited Kasmir and chose it as a resting place. Their

itineraries as well as the records of the political relations established

with Kasmir during a period of Chinese extension to the west, furnish

us with a series of interesting data, for the old geography of Kasmir.

It seems difficult to ascertain from the materials at present accessi¬

ble in translations or notices of European scholars, which is to be con¬

sidered the earliest Chinese reference to Kasmir. The difficulty is

connected with the use of the geographical term Ki-pin. This name

tion into Casliappur or Caspapur, the latter of which forms is the proper reading of

the Greek text ; ” Essay , p. 117.

1 It is curious to note that Kasyapapura was, according to an Indian authority

quoted by Alberuni, India, transl. Sachau, i. p. 298, one of the old names of Multan.

S See Ritter, Erdkunde, ii. p. 1087 ; Lassen, Ind. Alt., ii. p. 635 (where for

# Kasyapapura > Kasmira an equally unfounded derivation from #Kasyapamira

is substituted); Humboldt, Asie Centrale , i. p. 102; for modern works, e.g.,

McCrindle, Ancient India, p. 108; Beal, Si-yu-ki , i. p. 148.

14

ANCIENT GEOGRAPHY OF KASAI IR. [Extra No. 2,

originally and properly designated t lie Upper Kabul Valley.1 It ap¬

pears, however, at a period when Chinese knowledge of India was

less developed, to have been used in a vague and general fashion for a

variety of territories on the northern confines of India, among them also

Kasmir. However this may be, our loss seems scarcely to be great, as

these notices of the Chinese Annalists regarding Ki-pin do not seem to

give characteristic local details.2

The first clear reference to Kasmir which I can trace at present,

is contained in a record dating from 541 a.d. It is taken from the

account of an Indian envoy who reached China during the early part of

the reign of the T'ang dynasty.3 The name of Kasmir is not mentioned.

Yet it is evident that M. Pauthier who published the extract, was right

in referring to Kasmir the description given of the northern portion

of India as a country “ situated at the foot of the snowy mountains and

enveloped by them on all sides like a precious jewel. In the south

there is a valley which leads up to it and serves as the gate of the

kingdom.” The points noticed here are exactly those with which we

meet in all Chinese accounts of Kasmir.

9. Ninety years after the date of this notice Kasmir was visited by

Hhjen Tsiang. He reached the Valley from

Urasa in the west and resided in it as an

honoured guest for fully two years. The

records of the great Chinese pilgrim contain by far the fullest and most

accurate description of Kasmir that has come down to us from a foreign

pen during the period with which we are here concerned.4

Hiuen Tsiang must have entered Kasmir by the valley of the

Vitasta as he describes his route as leading to the south-east of

Urasa, the present Hazara District. After ‘ crossing over mountains

and treading along precipices’ he arrived at the c stone gate which is

the western entrance of the kingdom.’ We shall see below that this

gate known also to Ou-k'ong and Alberuni, was the frontier watch-

station or JDcara in the gorge of Baramula ( Varahamula). He passed

the first night on Kasmir soil at R usTcayura, the modern Uskiir, opposite

Baramula. Thence he proceeded to the capital which he describes

1 Compare the explanations of Messrs. Levi and Chavannes in their paper

‘ L’ltineraire d’ Ou-k'ong, Journal asiatique, 1895, vi. pp. 371 sqq., together with the

supplementary and modifying statements, ib., 1896, vii. pp. 161 sq.

2 These notices are enumerated by Messrs. Levi and Chavannes, Journal asiat.,

1895, vi. pp. 378 sq.

8 G. Pauthier, Examen metliodique des faits qui concernent le Thian-Tchou

on VInde, Paris, 1839, p. 40.

4 See Si-yu-ki , transl. Beal, i. pp. 148 sqq. ; Life of Hiuen Tsiang, transl. Beal,

pp. 68 sqq.

J899.]

CHINESE RECORDS.

15

exactly in the position of the present Srinagar. There he was lodged in

the convent known as the Jayendravihara which is named also in the

Rajataraiigini. 1 A two years’ stay, though chiefly passed in the study of

‘ the Sutras and S'astras’, must have enabled Hiuen Tsiang to acquaint

himself thoroughly with the Yalley.

His description of the kingdom Kia-shi-mi-lo shows clearly that

the geographical application of the term Kasmlr must have been then,

exactly as now, restricted to the great basin of the Yitasta and the

side valleys drained by its tributaries above the Baramula defile. He

notices that the country is enclosed on all sides by mountains which

are very high. “ Although the mountains have passes through them,

these are narrow and contracted.” These natural bulwarks protected

the country from neighbouring states ‘ which had never succeeded in

subduing it.’ Though the climate is cold and the snow plentiful, the

soil is fertile and abounds with fruits and flowers. The inhabitants

seem to have changed as little as the soil since Hiuen Tsiang’s days. It

is still easy to recognize in them the people whom he describes as “ light

and frivolous, and of a weak, pusillanimous disposition. The people

are handsome in appearance, but they are given to cunning. They love

learning and are well-instructed.”

“ Since centuries learning had been held in great respect in this

kingdom.” Hiuen Tsiang dwells with evident pleasure on the re¬

collection of the learned conferences he had with the Kasmlr doctors

of the sacred law.2 Kasmlr had in earlier times played a great part in

the traditions of the Buddhist church. Hiuen Tsiang relates at length

the legends how the Arhat Madhyantika had first spread the law of

Buddha in the land ; how in the time of Asoka the five hundred Arhats

had taken up their abode there ; and how finally under the great

Kaniska, king of Gandhara, Kasmlr had been the scene of the universal

Council which fixed and expounded the Sacred Canon. Yet he observes

that in his own time the kingdom ns a whole was “ not much given

to the faith, and that the temples of the heretics were their sole

thought.”3

It is probably owing to this not very flourishing condition of con¬

temporary Buddhism that Hiuen Tsiang mentions only a comparatively

small number of Viharas and Stupas in the Yalley. Among the Stupas

there were four ascribed to Asoka. Beneath another Kaniska was

believed to have deposited the canonical texts as fixed by his Council,

engraved on sheets of copper. None of these structures have yet been

1 Compare note iii. 355.

* See Life, p. 71 sq.

8 See Si-yio-ki , i. p. 158.

16

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

identified with any certainty. But in their description the pilgrim

furnishes us incidentally with a valuable topographical indication.

Speaking of the convent which prided itself on the possession of a

miraculous tooth of Buddha, he indicates its site as being about 10 li

(circ. 2 miles) to the south-east of the new city and to the north of the

old city.1 This proves that the capital of Hiuen Tsiang’s time which

corresponds to the present STinagar, was then a comparatively new

foundation, exactly as the Chronicle’s account has it. At the same time

the reference to the ‘old city’ enables us to fix with absolute certainty

the earlier capital of S'rinagari at the present Pandrethan , the Purdna-

dhisthana of Kalhana.2

The two full years which Hiuen Tsiang, according to his own

statement spent in Kasmir,3 represent a longer halt than any which the

pious traveller allowed himself during his sixteen year’s wanderings

through the whole of India and Central Asia.4 With all due respect

for the spiritual fervour of the pilgrim and the excellence of his

Kasmman preceptors, it is difficult to suppress the surmise that the

material attractions of the Valley had something to do with his long

stay. The cool air of Kasmir, the northern aspect of its scenery and

products, have at all times exercised their powerful charm over those

visitors who themselves born in colder climes have come to the Valley

from the heat and dust of the Indian plains. Just as these advantages

attract in yearly increasing numbers European visitors from India

Proper, so the modern Turk! pilgrims from Kashgar, Yarkand and

other parts of Central Asia, whether on the way to Mecca or on their

return, never fail to make a long stay in Kasmir.

We should undoubtedly find the example of the modern Hajis

followed also by Buddhist pilgrims if there were still any from those

northern regions to take their way through Kasmir to the holy places

of India. It would be an interesting task to examine to what extent the

fame of Kasmir as the ‘ paradis terrestre des Indes,’ is the creation of

the Valley’s northern visitors, both European and Asiatic. Here it may

suffice to add that Hiuen Tsiang before he reached Kasmir, must have

had already his experience of the torrid heat and the other amenities of

a Pan jab summer.5 & We shall also see that the example of the other

Chinese pilgrim whom we are able to follow on his visit to Kasmir,

points exactly to the same conclusion.

1 Si-yu-ki, i. p. 158.

2 See below §§ 88, 89.

See Life, p. 72.

4 Compare the table of dates for Hiuen Tsiang’s itinerary, Cunningham, Ancient

Geography, pp. 563 sqq.

& See Cunningham, Ancient Geography, p. 563 sq.

1899.]

CHINESE RECORDS.

17

Hiuen Tsiang’s narrative tells us that he left the Valley going in a

south-westerly direction. He reached Pun-nu-tso , the Parnotsa of the

Chronicle and the modern Prunts, after crossing mountains and passing

precipices.1 As the Tos^maidan route is the direct and most frequented

route to that territory, it is very probable that Hiuen Tsiang also follow¬

ed it. Parnotsa as well as Rajapuri (Ho-lo-she-pu-lo) to which the

pilgrim subsequently proceeded, had at the time of his visit no inde¬

pendent ruler, but were subject to Kasmlr.

10. The next Chinese notice of Kasmlr, and one which is of con¬

siderable historical interest, is contained in the

Annals of the T'ang dynasty.2 They inform us

that the first embassy from Kasmlr arrived at the imperial court in or

shortly after a.d. 713. In the year 720 Tchen-t' o-lo-pi-li} ruler of Kasmlr,

the Candraplda of the Chronicle, was accorded by imperial decree the

title of king.

T'ang Annals.

His brother and successor Mou-to-pi in whom Kalhnna’s Muktaplda

or Lalitaditya has long ago been recognised, sent after the first Chinese

expedition against Po-liu or Baltistan (between 736 and 747) an envoy

called Ou-li-to to the Chinese court. He was to report the alleged

victories of his master over the Tibetans but at the same time also to

solicit the establishment of a camp of Chinese troops by the banks of

the lake Mo-ho-to-mo-loung (the Mahapadma Naga or Volur lake).

The Kasmlr king offered to provide all necessary supplies for an auxi¬

liary force of 200,000 men. But the ‘ Divine Khan ’ found it more

convenient to content himself with issuing decrees for the sumptuous

entertainment of the ambassador aud for the registration of Muktaplda

with the title of king. Since that time the relations of Kasmlr with

the celestial empire and the receipts of tribute from the former are said

to have continued without interruption.

The description of Kasmlr which is coupled with this record of the

T'ang Annals, appears to be in the main copied from Hiuen Tsiang’s

Si-yu-ki. But in addition it furnishes us with an exact statement as to

the Kasmlr capital at that time. In my Notes on Ou-h' ony s Account of

1 Si-yu-ki, i. p. 162 Life p. 72.

2 The notice was first made known by A. R^musat’s translation of the

corresponding extract in Matuanlin’s encyclopsedia ; see Nouveaux Melanges

asiatiques, Paris, 1829, i. pp. 196 sqq. An abstract of the same notice, bnt from

the original text of the Annals, where the names are more correctly rendered, will

be found in Messrs. L£vi and Chavannes’ V ItimZraire d’Ou-k'ong, Journal asiat.,

1895, vi. pp. 354 sqq.

From Reinaud, Mdmoire sur VInde, pp. 189 sq. it would appear that the names

of Kasmlr kings in this Chinese record and that of the Mahapadma lake were

first correctly identified by Klaproth, Mdmoires relatifs d VAsie, ii. pp. 275 sq.

This work is at present not accessible to me.

J . i. 3.

18

ANCIENT GEOGRAPHY OF KAS'MlR.

[Extra No. 2,

KaSmlr 1 I have shown that the Po-lo-ou-lo-po-lo of the Annals is a

correct reproduction of Pravarapura, the old and official name of S'ri-

nagara. In the same way the name Mi-na-si-to given to the great river

which flows to the west of the capital, represents a correct enough

transcription of Vitasta. Both the names are recorded in the form

which they bore in the official Sanskit, and are, therefore, evidently

taken from the information given by the Kasmir envoys.

11. Not many years after Muktaplda’s embassy Kasmir was visited

by another Chinese pilgrim, Ou-k'ong. Though

Ou-k'ong. greatly inferior to Hiuen Tsiang in learning

and power of observation, he has yet left us

information regarding the country which is of interest and value. The

itinerary of Ou-k'ong the discovery and recent publication of which we

owe to Messrs. L6vi and Chavannes,1 2 * contains the reminiscences of forty

years’ wanderings, taken down after the pilgrim’s return to China and

in a form regrettably brief. But whether it be due to Ou-k'ong’s long

stay in Kasmir or to other causes, his account is fortunately far more

detailed in the case of Kasmir than in that of any other territory visited

by him. His description of the Valley and the several sites mentioned

by him have been fully discussed by me in the separate paper already

quoted.8 I need hence indicate here only the main results of this analysis.

Ou-k'ong reached Kasmir in the year 759 from Gandliara, presum¬

ably by the same ronte as Hiuen Tsiang had followed. He took

there the final vows of a Buddhist monk and spent there fully four

years engaged, as his itinerary tells us, in pilgrimages to holy sites and

in the study of Sanskrit.4 5 Though he is said to have studied from day¬

break till night-fall, his diligence does not seem to have brought him

much literary culture. This is curiously shown by the popular Apa-

bhramsa forms iu which our pilgrim records the names of the monas¬

teries he specially singles out for notice. Four of these I have been

able to identify with Viharas mentioned in the Chronicle,6 * and two of

them have left their names to villages which survive to the present day.

1 See pp. 26 sqq. in the above-quoted paper, published in the “ Proceedings ”

of the Imperial Academy, Vienna (Philos. -histor. Class), 1896, vol. cxxxv.

2 See L ’ Itindraire d’Ou-k'ovg, Journal asiat., 1895, vi. pp. 341 sqq.

& See Notes on Ou-k'ong’s account of Kasmir, loc. cit.

4 See L’ Itineraire d’Ou-Tc'ong , p. 356.

5 Thus the monastery of Ngo-mi-t'o-p'o-ivan (* Amitabhavana) corresponds to the

Amrtabhavana Vihara of Rajat. iii. 9, which has given its name to the present

Ant^bavan near S'rinagar. The ‘ monastere du mont Ki-tchS, (# Kicd < Skr. krtyd)

is no other than the Krtyasrama Vihara, at the modern village Kitsqhom, the legend

of which is related at lergbh by Kalhana, i. 131 sqq. The Vihara of the great

king Moung-ti (* Mutti) was one of Muktaplda’s foundations, probably the #Mukta-

1899.]

CHINESE RECORDS.

19

While Hiuen Tsiang mentions only about one hundred convents in

the country, Ou-k'ong found more than three hundred and speaks in

addition of the number of Stupas and sacred images as considerable.

We may conclude from this that there had been a rise in the popularity

of Buddhism in the century intervening between the visits of the two

pilgrims.

Ou-k'ong describes the kingdom of Kasmir correctly enough as

enclosed on all sides by mountains which form its natural ramparts.

Only three roads have been opened through them, and these again are

secured by gates. In the east a road leads to T'ou-fan or Tibet ; in the

north there is a road which reaches into Poliu or Baltistan ; the road

which starts from ‘the western gate’ goes to K'ien-t'o-lo or Gandhara. 1

We have here a clear enough description of the great routes

through the mountains which since ancient times have formed the main,

lines of communication between the Valley and the outer world. The

road to T'ou-fan corresponds undoubtedly to the present route over the

Zoji-La to Ladakh and hence to Tibet. The road to Po-liu is represented

by the present “ Gilgit Road,” leading into the Upper Kisangariga Valley

and thence to Skardo or Astor on the Indus. The third road can be no

other than the route which leaves the Valley by the gorge of Baramula

and follows the Vitasta in its course to the west. We have seen already

that Hiuen Tsiang followed it when he entered Kasmir by ‘the stone

gate, the western entrance of the kingdom.’ There can be doubt that

in the gates ( fermetures ) closing these roads we have a reference to the

ancient frontier watch-stations of which we find so frequent mention in

our Kasmirian records.

Besides these three roads Ou-k'ong knew yet a fourth. “ This, how¬

ever, is always closed and opens only when an imperial army honours it

with a visit.” It is probable that this curious notice must be referred

to one of the roads leading over the Pir Pantsal range to the south.

Owing possibly to political causes these routes may have been closed to

ordinary traffic at the time of Ou-k'ong’s visit.2

The political relations between China and the northern kingdoms

of India seem to have ceased soon after the time of Ou-k'ong. This was

probably due to the Chinese power under the later T'ang gradually

losing ground in Central Asia before the Uigurs and the Tibetans. The

vihtira at Huskapura : Uskiir, iv. 188. In the c monastere du general {tsiang -khm) ’

it is easy to recognize the Vihara of the Turk (Tuhkhara) Cankuna who was one

of Muktapida’s ministers. He is reported to have founded two monasteries called

after his own name (iv. 211, 215).

1 See Ii’ Itineruire d’ Ou-k'ong, p. 356.

2 See Notes on Ou-k'ong, p. 24 sq.

20

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

pilgrimages, however, of Chinese Buddhists to India continued during

the next two centuries, and of one at least of these pilgrim parties it is

recorded that it took the route throughKasmir.1 But no detailed account

bearing on Kasmir has yet come to light of these later pilgrimages.

Section III. — Muhammadan notices.

Kasmir closed to

Arab geographers.

12. After the Greeks and the Chinese the early Muhammadan

writers are our next foreign informants regard¬

ing the historical geography of India. If with

one very remarkable exception they have

nothing to tell us of Kasmir topography, the explanation is not far to

seek. The first rush of Arab invasion in the Indus Valley duriug the

eighth century had carried the Muhammadan arms at times close enough

to the confines of Kasmir.2 No permanent conquest, however, had been

effected even in the plains of the Northern Panjab. Protected in the

West by the unbroken resistance of the S'ahis of Kabul and in the South

by a belt of war-like Hindu hill-states, Kasmir had never been seriously

threatened. Even when Islam at last after a long struggle victoriously

over-spread the whole of Northern India, Kasmir behind its mountain

ramparts remained safe for centuries longer.

Conquest and trade were the factors which brought so large a part

of the ancient world within the ken of the early Muhammadan travel¬

lers and geographers. Both failed them equally in the case of Kasmir.

For a classical witness shows us that a system of seclusion, — ever easy

to maintain in a country so well guarded by nature as Kasmir, — hermeti¬

cally sealed at that time the Valley to all foreigners without exception.

Even the well-informed Al-Mas‘udi who had personally visited the

Indus Valley, is unable to tell us more about Kasmir than that it is a

kingdom with many towns and villages enclosed by very high and

inaccessible mountains, through which leads a single passage closed by a

gate.3 The notices we find in the works of Al-Qazwini and Al-Idrisi

are practically restricted to the same brief statement. The references

in other geographical works are even more succinct and vague.4

1 Compare Yule, Cathay , p. lxxi., and Julien, Journal astat., 1847, p. 43.

2 See Reinaud, Mdmoire sur VInde, pp. 195 sq. ; Alberuni, India, i. p. 21.

* See Al-MasludV s “ Meadows of Gold,” transl. Sprenger, I. p. 382.

4 The silence of the early Muhammadan geographers as regards Kasmir was

duly noticed by Ritter, Asia, ii. p. 1115. — For Al-Qazwir.T, see Gildemeister,

De rebus Indicis, p 210 ; for Al-Tdrisi, Elliot, History of India, i. pp. 90. sq.

For the notices of other Arab geographers, see Bibliotheca geographorum

J 809.]

MUHAMMADAN NOTICES.

21

AlTberunl’s interest

in Kasmlr.

13. Notwithstanding the circumstances above indicated, Arabic

literature furnishes us with a very accurate

and valuable account of old Kasmlr. We owe

it to the research and critical penetration of

AlberunI of whom indeed it might be said as of an early British ex¬

plorer of Afghanistan,1 that he could look through the mountains.

The great Muhammadan scholar had evidently utilized every opportunity

during his long stay at Grhazna and in the Panjab, (a.d. 1017-30) for

collecting information on Kasmlr.

His interest in the distant alpine valley is easily understood. He,

himself, tells us in the first chapter of his great work on India, how

Hindu sciences when the victories of Mahmud had made the Hindus

‘ like atoms of dust scattered in all directions,’ had retired far away

from the conquered parts of the country. They “ fled to places which

our hand cannot yet reach, to Kasmlr, Benares and other places.”2

In another passage he speaks again of Benares and Kasmlr as the high

schools of Hindu sciences.3 He repeatedly refers to Kasmirian authors,

and from the notices shown below it is evident that among his infor

mants, if not among his actual teachers, there were Kasmirian scholars.4

The curious fact that AlberunI himself composed some Sanskrit

treatises for circulation among ‘ the people of Kasmlr,’ 5 proves beyond all

arabicorum, ed. De Goeje, i. p. 4 ; ii. pp. 9, 445 ; v. p. 364; vi. pp. 5, 18, 68; vii. pp.

89, 687 ; also Abii-l-Fidd , ed. Reinaud, pp. 361, 506.

1 Mountstuart Elphinstone.

2 Alberuni’ s India , transl. Saohau, i. p. 22.

8 India, i. p. 173.

4 AlberunI, ii. 181, refers particularly to Kasmirian informants with whom he

conversed regarding the miracle of the ‘ Kudaishahr ,’ i.e., the Kapatefaara Tirtha

(see below § 112). The way in which the pilgrimage to this spot was described to

Alberuni, makes it quite certain that his informants were personally familiar with

the Tirtha. The same must be said of his note on the pilgrimage to the temple

of S'arada (i. 117 ; see below § 127). The details regarding a local Kasmlr festival

(ii. p. 178), the anecdote about the propagation of the S'isyahitavrtti in Kasmir

(i. 135), are such as could not well have reached Alberuni otherwise but by verbal

communication.

Writing himself in A.D. 1030 he refers to a statement contained in the almanac

for the S'aka year 951 (A.D. 1029—30) ‘which had came from Kashmir’ (i. p. 391).

He could scarcely have secured such an almanac except through Kasmirian Pandits

who even at the present day, wherever they may be, make it a point to provide

themselves from home with their local naksatrapattrikd.

For references to Kasmirian authors or texts specially connected with Kasmlr,

see i. pp. 126, 157, 298, 334, i. p. 54 (Visnndharma), etc. Compare also the very

detailed account of the calendar reckoning cux-rent in Kasmlr and the conterminous

territories, ii. p. 8.

8 See India, Prof. Sachan’s preface, p. xxiv., and the introduction to his edition

of the text, p. xx.

22

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

doubt the existence of special relations between the great Mleccha scho¬

lar and that jealously guarded country. These relations seem strange

considering what Alberuni himself tells us so graphically about the

rigid isolation of Kasmir. We can scarcely explain them otherwise

than by personal intercourse with Kasmirian Pandits.

In view of these indications we can hardly go wrong in attributing

a great portion of Alberuni’s detailed knowledge of Kasmir topography

to these learned informants. But we also know that the chances of war

had given him an opportunity of supplementing this knowledge in part

by personal observation. Alberuni refers in two places to his personal

acquaintance with the fortress Lauhur (or Lahur) on the confines of

Kasmir. In an extract from my commentary on the Rajatarangini

already published, 1 I have proved that Alberuni’s Lauhur is identical

with the castle of Lohara , so frequently mentioned in the Chronicle.

Its position is marked by the present Loh^rin on the southern slope of

the Pir Pantsal range.

‘ Loharakotta ’ is undoubtedly the same as the Fort of Loh-kot

which according to the uniform report of the Muhammadan historians

brought Mahmud’s attempt at an invasion of Kasmir to a standstill. It is

hence certain that Alberuni had accompanied this unsuccessful expedition.

It probably took place in a.d. 1021. Though it failed to reach Kasmir, it

must have given Alberuni ample opportunity to collect local informa¬

tion and to acquaint himself with the topography of those mountain

regions which formed Kasmir’s strongest bulwark to the south. The

result is yet clearly traceable in the accuracy with which he describes

the relative position of the most prominent points of this territory.

Is it too much to suppose that Alberuni had at one time or the other Kasmirian

Pandits in his employ ? We know that in preparing the vast materials digested in

his book he worked largely with the help of indigenous scholars. Judging from

his own description of the state of Hindu sciences iu the conquered territories and

the bitter enmity prevailing there against the dominant Mlecchas, it is doubtful

whether he could have secured there such assistance as he required.

Alberuni himself, when describing the difficulties in the way of his Indian

studies, tells us (i. p. 24) : “ I do not spare either trouble or money in collecting

Sanskrit books from places where I supposed they were likely to be found, and in

procuring for myself, even from very remote places, Hindu scholars who under¬

stand them and are able to teach me.”

Kasmir has always been distinguished by an over-production of learning. Its

Pandits have been as ready in old days as at present to leave their homes for distant

places wherever their learning secured for them a livelihood (compare Buhler,

Introd. to the Vikramdnkadevacarita, p. xvii ; also Indische Palxographie, p. 56).

1 See my note oh the ‘ Castle of Lohara ,’ Indian Antiquary , 1897, pp. 225 sqq.,

or Note E, on Rajat. iv. 177, §§ 12, 13.

MUHAMMADAN NOTICES.

23

1899.]

Alberunl’s account

of Kasmlr.

14. Alberuni’s main account of Kasmlr is contained in Chapter

xviii. which gives 1 various notes on the coun¬

tries of the Hindus, their rivers and their

ocean.’1 Compared with the description of the

rest of India, it is disproportionately detailed. Alberuni first sketches in

broad but correct outlines the political division of the mountain region

which lies between the great Central Asian watershed and the Panjab

plain. He then refers to the pedestrian habits of the Kasmirians and

notes the use by the nobles of palankins carried on the shoulders of men,

a custom fully illustrated by the Chronicle and accounted for by the

nature of the communications in the mountains.2

What follows deserves full quotation. “ They are particularly

anxious about the natural strength of their country, and therefore take

always much care to keep a strong hold upon the entrances and roads

leading into it. In consequence it is very difficult to have any commerce

with them. In former times they used to allow one or two foreigners

to enter their country, particularly Jews, but at present they do not

allow any Hindu whom they do not know personally to enter, much

less other people.”

We have here a full and clear statement of that system of guard¬

ing all frontier-passes which we have found alluded to already in the

Chinese records. It explains the great part which is played in the

Kasmlr Chronicles by the frontier watch-stations, the Dvaras and

Drangas. It is of all the more interest as the last traces of the system,

in the form of rahddrl , have disappeared in Kasmlr only within quite

recent memory.3

Alberuni then proceeds to describe the ‘ best known entrance to

Kashmir.’ Though the starting point of his itinerary cannot be identi¬

fied with absolute certainty, it is clear that he means the route which

ascends the Jehlam Valley. From “the town Babrahdn , half way

between the rivers Sindh (Indus) and Jailam, 8 farsakli are counted

to the bridge over the river where the water of the Kusnarl is joined by

that of the Mahwi , both of which come from the mountains of Shamilan

and fall into the Jailam.” Though there seems to be here some slight

confusion, I have little doubt that the point meant by ‘ the bridge over

the river ’ corresponds to the present Muzaffarabad, at the confluence

etc.

1 See India, i. pp. 206 sqq.

8 Compare e.g. Bdjat. iv. 407 ; v. 33, 219; vii. 478; viii. 2298, 2636, 2674, 3165,

The word katt which Alberuni gives as the indigenous term of the palankin is

perhaps a corrupted Apabhramsa form of Tearniratha, often named in the Rajat.

3 Compare my Notes on the Ancient Topography of the Pir Pant^al Route

J. A. S. B., 1895, pp. 382 sqq. ; also below § 40.

24

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

of the Jehlam ami Kisangahga. The easiest route to Kasmir from

the west leads through the open central portion of Hazara (Urasa) to

Mansahra ; hence across the Kunhar and Kisangahga rivers to Muzaf-

farabad, and then up by the right side of the Jehlam Valley to

Baramula.1 * * In Kusnarl it is easy to recognize with Prof. Sachau the

present Kunhar River which falls info the Jehlam a few miles below its

great bend at Mnzaffarabad.a The Mahwi is evidently meant to designate

the Kisangahga.8 If thus interpreted the oidy error in Alb§runi’s de¬

scription is that it makes the Kunhar join the Kisanganga whereas

in reality it falls into the Jehlam after the latter’s junction with the

K isanganga.

I have shown in my note on Rajat. v. 215 that the route here indi¬

cated, which was a favorite one until the modern “Jehlam Valley

Tonga Road ” was constructed, is distinctly referred to already in

Kalhana’s account of S'amkaravarman’s marcli to and from Urasa. The

distance of 8 farsakli corresponds according to Alberuni’s reckoning to

about 39 English miles.4 * Referring to the map and the modern route

measurements 6 * this distance carries us to a point between Mansahra

and the next stage Abbottabad, i.e., exactly into the neighbourhood

where according to the evidence given in the above-quoted note the old

capital of Urasa must be located. ‘ Babrahan ’ which cannot be identi¬

fied at present, is perhaps intended to represent the name of this old

town which could fairly be described as situated midway between the

Indus and Jehlam.

From Muzaffarabad onwards, — where there is still a bridge over the

Kisangahga just as at the time (1783) when Forster crossed here on his

way from Kasmir to Attock,6 and as, if our explanation is right, in the

time of Alberuni, — we can follow the route quite plainly. Alberiini

counts five days of march “ to the beginning of the ravine whence the

l This route is described, e.g., by Drew, Jummoo, p. 528, ‘ as the easiest route

from the Panjab to Kasmir.’

8 Kunhar represents the regular phonetic derivative of a Skr. * Kusndri, medial

s becoming always h under a phonetic law common to Kasmir! and the related

dialects ; for the change hn > nh compare Grierson, Phonology of Indo-Aryan

Vernaculars, Z. D. M. G., 1896, p. 33.

8 I am unable to account for the name Mahwi. Could it be the corruption of

an Apabhramsa derivative of Madhumatl ? This name, though properly applied

to an affluent of the Kisangahga, is used in a Mahatmya also for the latter river

itself ; see Note B, Rajat. i. 37, § 16.

4 Compare Prof. Sacha, u’s note, India, ii. p. 316. Alberuni values his farsaM.

at 4 Arabian miles or approximately 4x2186 yards. Hence 1 farsakh =

English miles.

& See Drew, loc. ext.

8 See G. Forster, Journey from Bengal to England , 1808, ii; p. 46.

1899.]

MUHAMMADAN NOTICES.

25

river Jailam comes,’ that is, of the gorge through which the river flows

immediately below Baramfila. This estimate agrees closely with the

actual road distance between Muzaffarabad and Baramula which is

given by Drew as 84 miles.1 At the other or Kasmir end of the ravine

Alberuni places quite correctly ‘ the watch-station Dvdr ’ (Skr. Dvara )

the position of which, as we shall see below, is marked to this day by

the site of the old gate known as Drang.

“ Thence leaving the ravine you enter the plain, and reach in two

more days Addishtan , the capital of Kashmir,

tion of the Valley passing on the road the village Ushkara. All

this is perfectly accurate. Adhistbana ‘the

capital ’ is, of course, meant for S'rinagara2 and Ushkara for Uskur ,

opposite Baramula, the ancient Huskapura already mentioned by Hiuen

Tsiang.3 Alberuni’s mention of Uskur which is on the left river bank,

shows that then as now the ordinary road from the ‘ Grate of Varaha-

mula’ to S'rlnagara passed on the left or southern side of the Valley.

Two marches are still counted for this part of the journey.

The capital is correctly described as “ being built along both banks

of the river Jailam which are connected with each other by bridges

and ferry boats.” It is said to cover ‘ a space of four farsakh .’ This if

interpreted to mean ‘ a space of four Farsakh in circumference,’ would

not be too far from the truth, assuming that all suburban areas around

the city are included in the estimate. The course of the river above

and below the capital is traced rightly enough as far as the Valley is

concerned. “ When the Jailam has left the mountains and has flowed

two days’ journey, it passes through Addishtan. Four Farsakh farther

on it enters a swamp of one square Farsakh.” Here, of course, the Volur

lake (Mahapadma) is meant. “ The people have their plantations on

the borders of this swamp, and on such parts of it as they manage to

1 See loc. cit. According to Drew’s table six marches are counted, but one

of them is very short. On the modern route following the opposite side of the

river five marches are now reckoned from Domel, opposite to Muzaffarabad, to

Baramula.

2 Adhisthana, used again ii. p. 181, is a term which indicates that Alberuni’s

informant was a Sanskrit-speaking person. The common designation of the capital

was S'rinagara or simply Nagara ; see § 91 below.

S The text as rendered by Prof. Sachau, speaks of “ UshTcdra which lies on

both sides of the Yalley, in the same manner as Baramula .” There is either some

corruption in the text here or Alberuni’s informant had not made himself sufficiently

clear. What he must have meant, is that Ushkara lay on the opposite side of the

river in the same manner as Baramula, that is at the entrance of the ravine.

Baramula as the text spells the name, reproduces an earlier form of the Kasmiri

Varahmul, from Skr. Vardhamula.

J. i. 4

26

ANCIENT GEOGRAPHY OF KASHMIR. [Extra No. 2,

reclaim. Leaving this swamp, the Jailam passes the town of U.^hkara,

and then enters the above-mentioned ravine.”

The only mistake and this one easily explained is contained in the

account of the river’s origin. It is described as rising “ in the moun¬

tains Haramakot where also the Ganges rises ; cold, impenetrable regions

where the snow never melts nor disappears.” It is easy to recognize

here the reference to Mount Haramukuta and the sacred Gatiga-lake

at the foot of its glacier in which Kasim rian tradition places the source

of the Siudhu river.1 The latter is the greatest tributary of the Vitasta

within Kasmir and is traditionally identified with the Gaiiga, as on the

other hand the Vitasta with the Vamuna.2 The special sanctity of

the Sindu ( ‘Uttaragahga ’) and the popularity of its supposed source

as a pilgrimage place sufficiently account for the substitution in Albe-

runi’s notice.

Entering the open plain of the Kasmir Valley by the Baramula

gorge “you have for a march of two more days, on your left the moun¬

tains of Bolor and Shamilan, Turkish tribes who are called Bliattavaryan.

Their king has the title of Bhattn-Shah .” It is clear that Alberuni’s

informant here means the mountain ranges to the north and north-west

of the Valley which form its borders towards the Dard country and

Baltistan. The latter has been known by the name of Bolor for many

centuries.3 4 I am unable to trace in Kasmirian or other sources the

names of the ‘ Shamilan’ and ‘ Bhatta.’41 But as a subsequent remark

mentions ‘ Gilgit , Aswlra, and Shiitas ,’ that is the modern Gilgit, Has5r

(Astor) and Cilas as their chief places, there can be no doubt that the

inhabitants of the Dard territory to the north-west of Kasmir are

meant together with the Baltis.

“Marching on the right side [of the river], you pass through

_ . , . „ villages, one close to the other, south of the

Description of Pir

Pantsal capital and thence you reach the mountain

Kuldrjak, which is like a cupola, similar to the

1 See below, § 57, and Rajat. note i. 57.

8 See Rdjat. note i. 57. In Haracar. iv. 54 the Vitasta itself is designated as

the ‘ Gahga of the north ’ (Uttaragahga). This renders the location of its source

in the lake of Haramukuta still more intelligible from a traditional point of view.

8 Compare Yule, Marco Polo, i. pp. 187, sq. ; Cunningham, Anc. Geogr., p. 83.

4 Alberuni’s Bhatta may possibly represent the term Bhutta or Bhautta (the

modern Ks. Buia) which is applied in the Sanskrit Chronicles to the population of

Tibetan descent generally, from Ladakh to Baltistan. (See Rdjat. note i. 312).

Alberuni calls their language Turkish, but it must be remembered that he has

spoken previously (i. p. 206) of ‘ the Turks of Tibet ’ as holding the country to

the east of Kasmir. There the Tibetans in Ladakh and adjacent districts are clearly

intended.

1899.]

MUHAMMADAN NOTICES.

27

mountain DunbRwand (Damawand). The snow there never melts. It

is always visible from the region of Takeshar and Lauhawar (Lahore).”

I have already elsewhere shown that the mountain here described

is the Tatakuti peak (33d 45' lat. 74° 33' long.).1 It rises to a height of

15,500 feet in the central part of the Pir Pantsal range and is the

loftiest as well as the most conspicuous point of the mountain chain to

the south of Kasmir. It has the shape described by Alberuni, is sur¬

rounded by extensive snow-fields and can be seen through the greatest

part of the year from the Panjab districts of Sialk5t and Gujranwala

corresponding to the old Takeshar (Takkadesa). Alberuni puts the

distance between this peak and the Kasmir plain at two farsakh. This

estimate is somewhat too low, inasmuch as the direct distance on the

map between the peak and the nearest point of the open Valley is about

15 miles.

He is, however, quite exact in placing the fortress Lauhur to the

west of it as we have already seen that this stronghold is identical with

the Loharakotta of the Chronicle, the present Loh^rin. The entrance to

the Loli^rin Valley lies almost due west of Tatakutl. To the south of

the peak he places ‘the fortress Raj a girl ’ which is also mentioned by

Kalhana, vii. 1270, and must be looked for somewhere in the Upper

Suran Valley. Alberuni speaks of these two hill fortresses as “the

strongest places ” he had ever seen.

He had personally had an opportunity of judging of their strength

when accompanying Mahmud’s expedition against Kasmir. On that

occasion he had made the observation of the latitude of Lauhur (Lohara)

to which he refers in another chapter of his work.2 The result of this

observation, 33° 40' lat. as shown in the author’s Canon Masudicus,

very closely approaches the real one, which is 33° 48' according to the

Survey map. It is very probable that he obtained at the same occasion

the very accurate information regarding the distance from Lauhur to

the Kasmir capital. He gives it as 56 miles, “ half the way being

rugged country, the other half plain.” Alberuni’s measurement accor¬

ding to the previously stated valuation represents about 69 English

miles. This is but little in excess of the actual road distance via the

Tos^maidan pass as estimated by me on the tour referred to in the

above-quoted paper. The description of the road, too, corresponds

closely to the actual character of the route.

Alberuni closes his account of Kasmir geography with a reference

to the town of Rajawari which is the Rajapurl of the Chronicles, the

1 See my paper ‘ The Castle of Lohara Ind. Ant., 1897, § 12.

2 See India, i. p. 317, with Prof. Saohau’s note ii. p. 341. In the same passage

he quotes the latitude of Srinagar as 34° 9' from the Karanasdra.

28

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

modern Rajauri. In Hindu times it was the capital of a small hill-state

situated immediately to the south of the Pir Pants al range arid often

tributary to Kasmir. AlberunI distinctly names it as the farthest place

to which Muhammadan merchants of his time traded and beyond which

they never passed. We have already seen what the connection was

which enabled him to collect reliable and detailed information of the

region beyond that barrier. As another proof of the accurate know¬

ledge thus acquired, we may finally mention his description of the

Kasmir climate which is far more exact than any account available to

us previous to the second quarter of this century.1

Section IV. — Indian notices.

Deficiency of non-

Kasmlrian texts.

15. Nothing, perhaps, can illustrate better the lamentable lack of

exact geographical information in general

Sanskrit literature than to turn from the

accounts of the Chinese pilgrims and AlberunI

to what Indian authors, not Kasmlrians themselves, can tell us of the

Valley.

Were we to judge merely from the extreme scantiness of the data

to be gleaned from their extant works, we might easily be led to assume

that Kasmir was to them a country foreign and remote in every way.

However, we observe the same vagueuess and insufficiency of local

references in the case of territories immediately adjoining the old

centres of literary activity. It is hence evident that the conspicuous

absence of useful information on Kasmir may equally well be attributed

to the general character of that literature.

The name Kasmira , with its derivative Kasmira , as the designation

of the country and its inhabitants, respectively, is found already in the

Ganas to Panini’s grammar and in Patanjali’s comments thereon.3 The

Mahabharata too refers in several passages to the Kasmiras and their

rulers, but in a fashion so general and vague that nothing more but

the situation of the country in the hill region to the north can be

concluded therefrom.3

The Puranas enumerate the Kasmiras accordingly in their lists of

northern nations. But none of the tribal names, partly semi-mythical,

1 See India, i. p. 211, and below, § 77.

2 See the references in the Thesaurus of Bohtlingk-Roth, s. v. Kasmira, and in

supplement V,, p. 1273. The references to other texts in this paragraph have also

been taken from that work except where otherwise specified.

3 Compare in particular Mahdbh. II. xxvii. 17.

1899.]

THE KAS'MIR CHRONICLES.

29

which are mentioned along with them in the Puranas examined by me,

indicate any more distinct location of the country.1 2

Varahamihira (circ. 500 a.d.) in his Brhatsamhita includes the

Kasmiras curiously enough in the north-eastern division. Among the

regions and peoples named under the same heading there are a number

of purely legendary character like 4 the kingdom of the dead ’ ( nasta -

rdjya), the 4 gold region,’ 4 * the one-footed people,’ etc. But besides

these names and others of a different type which cannot be clearly

identified, we recognize the names of tribes which undoubtedly must be

located in the immediate neighbourhood of Kasmir. Thus we have the

Abhisdras , Daradas , Darvas, Khas'as , Kiras , and somewhat more distant

the country of Kuluta (Kulu) and the Kaunindas or Kaulindras

(Ptolemy’s KvAirS/nV^).8

Perhaps the most specific piece of information regarding Kasmir

that Sanskrit literature outside the Valley can convey to us, is con¬

tained in the term Kasmlra or Kdsmiraja which designates the saffron

and according to the lexicographers also the root of the hustha or costus

speciosus. Both the saffron and the Kustha have since early times

been famous products of Kasmir.3

Section V. — The Kas'mir Chronicles.

Abundance of

Kasmlrian sources.

16. The want of detailed and exact geographical information just

noticed in old Indian literature generally stands

in striking contrast to the abundance of data

supplied for our knowledge of old Kasmir by

the indigenous sources. The explanation is surely not to be found in

the mere fact that Kasmlrian authors naturally knew more of their own

country than others for whom that alpine territory was a distant, more or

less inaccessible region. For were it so, we might reasonably expect to

find ourselves equally well informed about the early topography of other

1 Compare Vdyupur. xlv. 120 ; xlii. 45 ; Padmapur. I. vi. 48, 62 ; Bhdgavatapur.

XII. i. 39; Visnupur. IY. xxiv. 18.

2 See Brhatsamhita, xiv. 29 sqq., and Ind. Ant., 1893, pp. 172, 181 ; also Alberuni

India, i. p. 303,

s Regarding the saffron cultivation of Kasmir, compare Lawrence, Valley,

p. 342, and below, § 78.

The kustha, now known in Kasmir by the name of kuth, is the aromatic root of

the Saussurea Lappa which grows in abundance on the mountains of Kasmir ; see

Lawrence, p. 77. The kuth is still largely exported to China and might be hence

one of the medicinal plants which Hiuen Tsiang particularly notices among Kasmir

products ; se& Si-yu-ki, i. p. 148.

30

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

parts of India which have furnished their contingent to the phalanx of

Sanskrit authors. Yet unfortunately this is by no means the case.

The advantageous position we enjoy in Kasmir is due to a combina¬

tion of causes- of which the most important ones may at once be here

indicated. In the first place we owe it to the preservation of connected

historical records from a comparatively early date which acquaint us

with a large number of particular localities and permit us to trace their

connection with the country’s history.

Another important advantage results from the fact that Kasmir,

thanks chiefly to its geographical position and the isolation resulting

from it, has escaped those great ethnic and political changes which have

from time to time swept over the largest portion of India. Local

tradition has thus remained undisturbed and still clings to all prominent

sites with that tenacity which is characteristic of alpine tracts all over

the world. The information preserved by this local tradition in Kasmir

has often proved for our written records a most welcome supplement

and commentary.

Finally it must be remembered that in a small mountain country

like Kasmir, where the natural topographical features are so strongly

marked and so permanent, the changes possible in historical times as

regards routes of communication, sites for important settlements, cul¬

tivated area, etc., are necessarily restricted. The clear and detailed

evidence which the facts of the country’s actual topography thus fur¬

nish, enables us to elucidate and to utilize our earlier data, even where

they are scanty, with far greater certainty and accuracy than would be

possible on another ground. The observations here briefly indicated

will be in part illustrated by the review of our Kasmlrian sources.

17. Epigraphical records on stone or copper such as elsewhere in

India form the safest basis for the study of

local topography, have not yet come to light

in Kasmir. The few fragmentary inscriptions

hitherto found are all of a late date and do not furnish any topographi¬

cal information. In their absence Kalhana’s RajataranginI is not only

the amplest but also the most authentic of our sources for the historical

geography of Kasmir. The questions connected with the historical

value of the work, its scope and sources, have been fully discussed

in the introduction to my translation. Here we have only to consider

its character as our chief source of information on the old topography

of Kasmir.

Kalhana’s work, composed in the years 1148-49 a.d., is our oldest

record of the history of the various dynasties which ruled Kasmir from

the earliest period to the time of the author. The earlier Chronicles

Kalhana’s Raja-

taranginl.

THE KAS'MlR CHRONICLES.

31

1899.]

which Kalhana has used and quoted, have all been lost. We are hence

unable to judge what he took from each, and how he worked up their

contents. Largely legendary in the first three Books, his narrative

reaches firm historical ground with the Karkota dynasty in the Fourth

Book. From Avantivarman’s reign (a.d. 855-883) onwards which

opens the Fifth Taraiiga, the Chronicle may be considered an accurate

and reliable historical record.. As the author approaches his own time,

his narrative grows more and more detailed.

In illustration of the latter fact it may be mentioned that of the

whole work comprising nearly eight thousand S'lokas, more than one-half

is devoted to the relation of the reigns which fill the century and a half

immediately preceding the date of composition. We have certainly no

reason to regret the fulness with which Books vii. and viii. relate the

events of the author’s own time and of the period that lay near it.

From a historical point of view, Kalhana’s detailed account of contem¬

porary history and the near past must always retain its value. We

can appreciate its advantages also with special regard to the elucida¬

tion of the old topography of the country. This will become at once

clear by a brief analysis of the topographical information contained in

the Chronicle.

It is doubtful whether Kalhana writing for readers of his own

country and time, would have deemed it necessary to give us a connec¬

ted and matter-of-fact description of the land, even if the literature

which he knew and which was his gnide, had furnished him with a

model or suggestion for such a description. The nearest approach to it

is contained in a brief passage of his introduction, i. 25-38. This ac¬

quaints us in a poetical form with the legends concerning the creation

of Kasmir and its sacred river, the Vitasta, and enumerates besides the

most famous of the many Tirthas of which Kasmir has ever boasted in

abundance. The few panegyric remarks which are added in praise of

the land’s spiritual and material comforts, i. 39-43, do credit to the

author’s love of his native soil. But they can scarcely be held to raise

the above to a real description of the country.

18. Notwithstanding the absence of such a description Kalhana’s

Chronicle yet proves by far our richest source

of information for the historical geography of

Kasmir. This is due to the mass of incidental

notices of topographical interest which are spread through the whole

length of the narrative. They group themselves conveniently under

three main heads.

Considering the great attention which the worship of holy places

has at all times claimed in Kasmir, we may well speak first of the

Kalhana’s notices

of Tirthas.

32

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

notices which appertain to the Topographia sacra of the Valley.

Kasmir lias from early times to the present day been a land abundantly

endowed with holy sites and objects of pilgrimages. Kalhana duly

emphasizes this fact when he speaks, in the above-quoted introductory

passage, of Kasmir as a country ‘ where there is not a space as large

as a grain of sesamum without a Tlrtlia.’1

Time and even the conversion to Islam of the great majority of

the population has changed but little in this respect. For besides the

great Tirthas which still retain a fair share of their former renown

and popularity, there is scarcely a village which has not its sacred

spring or grove for the Hindu and its Ziarat for the Muhammadan.

Established as the latter shrines almost invariably are, by the side

of the Hindu places of worship and often with the very stones taken

from them, they plainly attest the abiding nature of local worship in

Kasmir.

This cannot be the place to examine in detail the origin and

character of these Tirthas and their importance for the religious history

of the country. It will be enough to note that the most frequent

objects of such ancient local worship are the springs or Nagas , the

sacred streams and rivers, and finally the so-called svayambhu or ‘ self-

created ’ images of gods which are recognized by the eye of the pious in

various natural formations. These several classes of Tirthas can be

traced throughout India wherever Hindu religious notions prevail, and

particularly in the sub-Himalayan regions (Nepal, Kumaon, Kangra,

Udyana or Swat). Still there can be no doubt that Kasmir has, from

old times claimed an exceptionally large share in such manifestations of

divine favour.

Nature has indeed endowed the Valley and the neighbouring moun¬

tains with an abundance of fine springs. As each of these has its

tutelary deity in the form of a Naga ,2 * we can easily realize why popular

tradition looks upon Kasmir as the favourite residence of these deities.5

Hiuen Tsiang already had ascribed the superiority of Kasmir over other

countries to the protection it received from a Naga.4 Kalhana, too, in

his introduction gives due prominence to the distinction which the land

1 i. 38.

2 Compare my note i. 30 on the Nagas and their worship.

S The Nilamatcrpurdna, 900-972, gives a long list of Kasmir Nagas and puts

their number at thousands, nay Arbudas (see 971 ).

4 Si-yu-ki , i. p. 148. Hiuen Tsiang, like other Chinese pilgrims, calls the Nagas

by the term of ‘ dragon ; ’ no doubt because the popular conception represents

them under the form of snakes living in the water of the springs or lakes they

protect.

1899.]

THE K AS' Ml It CHRONICLES.

33

enjoys as f lie dwelling-place of Nila, king of Nagas, and of many other

of his tribe.1

Kalhana’s frequent references to sacred springs and other Tirthas

are of topographical interest, because they enable us to trace with

certainty the earlier history of most of the popular pilgrimage places

still visited to the present day. The list already mentioned acquaints

us with the miraculous springs of Papasudana and Tri-8amdhycl,

Sarasvatfs lake on the Bheda hill, the ‘ Self-created Fire’ ( Svayambhii ),

and the holy sites of Nandiksetra, S'drada, Cakradhara and Vijaye£a. It

shows which were tlie Tirthas most famous in Kal liana’s time. The

legends connected with the early semi-mythical kings give the chronicler

frequent occasion in the first three Books to speak in detail of particular

sacred sites. Almost each one of the stories furnishes evidence for the

safe location of the latter.8 But also in the subsequent and purely histo¬

rical portions of the work we read often of pilgrimages to such sacred

places or of events which occurred at them.

Kalhana shows more than once so accurate a knowledge of the

topography of particular Tirthas that his personal visits to them

may be assumed with great probability. This presumption is parti¬

cularly strong in the case of Nandiksetra which his father Canpaka is

said to have often visited as a pilgrim and to have richly endowed,

and of the neighbouring shrine of Bhute&varas* Also the distant Tirtha

of Sarada, in the Kisanganga Valley seems to have been known person-

nally to the Chronicler.41 Considering the popularity which pilgrimages

to sacred sites have always enjoyed among Kasmirians, the conclusion

seems justified that Kalhana owed perhaps no small part of his practical

acquaintance with his country’s topography, to the tours he had made

as a pilgrim.

19. A second fruitful source of valuable topographical notices is

contained in those very numerous references

Kalhana’s references

to foundations.

by particular kings. If we leave aside the curious list, i. 86-100, taken by

1 Rajat. i. 28-31. The Nagas are supposed to have come to Ka6mlr when

Kasyapa, their father, had drained ‘ the lake of Sati,’ and to have found there a

refuge from Garuda ; comp. Nilamata, 59 sqq.

2 Compare the legends of the Sodara spring, i. 123 sqq. ; of the Krtyasrama

Vihara, i. 131 sqq. ; of the Jyestharudra at Nandiksetra and S'rinagari, i. 113, 124 ;

the story of the Susravas Naga, i. 203 sqq. ; the description of the pilgrimage to the

Taksaka Naga, i. 220 sqq . ; the story of the Isesvara temple, ii. 134; of Ranasvamin,

iii. 439 sqq., etc.

3 See vii. 954 ; viii. 2365 and note v. 55 sqq. Compare also below, § 57.

4 See Note L, viii. 2492, § 4.

J. i. 5

which Kalhana makes to the foundation of

towns, villages, estates, shrines, and buildings

34

ANCIENT GEOGRAPHY OF KAS'MlR.

[Extra No. 2,

Kalhana from Padmamihira in which certain local names are by fanciful

etymologies connected with seven of the ‘lost kings,’1 * * it may be safely

assumed that these attributions are based either on historical fact or at

least on genuine local tradition. Kalhana specially informs us in his

introduction 8 that among the documents he had consulted for his work,

there were ‘ the inscriptions recording the consecration of temples and

grants [of land] by former kings.’ Such records no doubt supplied a

great portion of the numerous notices above referred to. Often such

notices may have been taken from less authentic sources. But we may

always claim for them the merit of acquainting us with the names of

the respective localities and buildings, as used in the official language

of Kalhana’ s time, aud with the traditions then current regarding their

origin and date.

The system of nomenclature which was regularly followed in Kasmir

in naming new foundations, must have helped to preserve a genuine

tradition regarding the founder. In the vast majority of cases the names

of new towns and villages are formed by the addition of - pura to the

name of the founder, either in its full or abbreviated form.5 * * Similarly

the names of temples, monasteries, Mathas and other religious structures

show the name of their builder followed by terms indicating the deity

or the religious objects to which the building was dedicated.4 Many of

l See regarding this nnhistorical list note i. 86. The local names, like Khona -

mu$a, Oodhard, S'amdngdsd, etc., are all genuine enough. What Padmamihira did was

to evolve fictitious names of kings out of these by means of popular etymology.

» i. 15.

5 Thus we have, e.g., the well-known localities of Huskapura, Kaniskapura,

Juskapura (which retain the memory of their Indo- Scythian founders) ; Fravara-

purob (for Pravarasenapura), the old official designation of the pi*esent capital ;

Padmapura , Avantipura, Jayapura (for Jaya.pich.pnra) and a host of others. The

custom of naming new localities in this fashion, or of renaming earlier ones in

honour of the actual ruler, can be traced through successive periods of Muhammadan

and Sikh rule down to the present day ; comp, eg., Zaiv^por (named after Zainu-1-

‘abidin) ; Shahabuddlnpur (now Shadipur) ; Miihammadpur ; Ranbirsinghpur (in¬

tended to replace Shahabad ), etc.

4 Thus in the case of S'iva-temples -Isa or -isvara is invariably added (comp., e.g.,

Pravaresvara, Amptesvara, etc.), as in that of Visnu-shrines with equal regularity

-svainin (-kesava) ; comp, e.g., Muktasvdmin (built by Muktapida), Avantisvdmin ,

Bhhnakesava (erected by Bhimapala S'ahi), etc.

Buddhist monasteries receive the name of their founder with the addition of

- vihdra or -bhavana ; comp. Jayendravihdra, Cankunavihara, Amrtabhavana (fonnded

A

by Queen Amrtaprabha, the present Antabavan), Skandabhavana (for Skandagup-

tabhavana), and many more, as shown in my Notes on Ou-k'ong, p. 4.

For Mathas compare e. g. Diddamntha (Didamar) ; Subhat dmatha, Nanddmatha ,

Lothikdmatha, Cakraviatha, etc.

1899.]

THE KAS'MiR CHRONICLES.

35

these religious structures left their names to the sites at which they

were erected. They can thus be traced to the present day in the

designations of villages or city quarters.1

The topographical interest which Kalhana’s notices of town-founda¬

tions possess is considerably enhanced by the fact that in more than

one case they are accompanied by accurate descriptions of the site

chosen and the buildings connected with them. Thus Kalhana’s

detailed accounts of the foundation of Pravarapura , iii. 336-363, is

curiously instructive even in its legendary particulai's. It enables us

to trace with great precision the original position and limits of the city

which was destined to remain thereafter the capital of Kasmir.2 3 Simi¬

larly the description given of Parihasapura and its great shrines has

made it possible for me to fix with accuracy the site of the town which

Lalitaditya’s fancy elevated for a short time to the rank of a capital,

and to identify the remains of the great buildings which once adorned

it.8 Not less valuable from an antiquarian point of view is the account

given to us of the twin towns Jayapura and Dvavavati which King

Jayapi da founded as his royal residence near the marshes of And^rkoth.4

We shall see below to what extent the correct identification of the extant

ruins of Kasmir has been facilitated by these and similar accounts of the

Rajataranginl.

20. Valuable as the data are which we gather from the two

groups of notices just discussed, it may yet

be doubted whether by themselves, that is,

unsupported by other information, they can

throw as much light on the old topography

of Kasmir as the notices which we have yet to consider. I mean the

whole mass of incidental references to topographical points which we

find interwoven with the historical narrative of the Chronicle.

It is evident that where localities are mentioned in the course

of a connected relation of events, the context if studied with due

regard to the facts of the actual topography, must help us towards a

correct identification of the places meant. In the case of the previous

notices the Chronicler has but rarely occasion to give us distinct indica¬

tions as to the position of the sites or shrines he intended. In our

Topographical data

in historical

narrative.

1 The name of the Amrtabhavana, iii. 9, survives in the present Antabavan ;

Didddmatha and Skandabhavana in the Didamar and Khandabavan quarters of

S'rinagar ; similarly Lalitaditya’s great temple of Mdr.tdn.da left its name to the

village and district of Matan.

8 See note iii. 339-349 and below, § 92.

3 Compare Note F, iv. 194-204, and below, § 121.

4 See note iv. 506-511 ; also below, § 122.

36

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

attempts to identify the latter we have therefore only too often to depend

either on the accidental fact of other texts furnishing the required

evidence, or to fall back solely on the comparison of the old with modern

local names. That the latter course if not guided and controlled by

other evidence, is likely to lead us into mistakes, is a fact which re¬

quires no demonstration for the critical student.

It is different with the notices the consideration of which we have

left to the last. Here the narrative itself, in the great majority of

cases, becomes our guide and either directly points out to us the real

locality meant or at least restricts to very narrow limits the area within

which our search must proceed. The final identification can then

be safely effected with the help of local tradition, by tracing the modern

derivative of the old local name, or by other additional evidence of this

kind.

For the purpose of such a systematic search it is, of course, a very

great advantage if the narrative is closely connected and detailed. And

it is on this account that, as already stated above (§ 17), Kalhana’s

lengthy relation of what was to him recent history, in Books vii. and

viii., is for us so valuable. An examination of the topographical notes

in ray commentary on the Chronicle will show that the correct identifica¬

tion of many of the localities mentioned in the detached notices of the

first six Books has become possible only by means of the evidence fur¬

nished by the more detailed narrative of the last two.

In this respect the accounts of the endless rebellions and other

internal troubles which fill the greater portion of the reigns af the

Lohara dynasty, have proved particularly useful. The description of

the many campaigns, frontier-expeditions and sieges connected with

these risings supplies us with a great amount of topographical details

mutually illustrating each other. By following up these operations on

the map, — or better still on the actual ground, as I was often able to

do, — it is possible to fix with precision the site of many old localities

which would otherwise never have emerged from the haze of doubt and

conjecture.

In order to illustrate these general remarks it will be sufficient to

refer to a few typical examples among the many identifications thus

arrived at. As the corresponding notes of my commentary fully in¬

dicate the evidence on which these identifications are based, as well as

the process of reasoning by which they were arrived at, it will not be

necessary here to go ‘into details. A very characteristic example is

furnished by the important stronghold and territory of Lohara, which

was formerly supposed to bo Lahore. Its correct location at the

present Loh^rwi and the identification of the several places and routes

1899.]

THE KAS'MIR CHRONICLES.

37

mentioned in the same neighbourhood became possible only, as Note

E, iv. 177, shows,1 * 3 through the indications contained in Kalhana’s

description of the several sieges which this mountain fastness underwent

in his own time. Similar instances are the identifications of the

Gopadri hill (the present Takht-i Sulaiman), and of the streams

Mahasarit and Ksiptika (Mar and Kut^kul). Though prominent

features in the topography of the capital itself, they could not have

been correctly located but for the evidence supplied by the narrative

of the last Book.8 The same is the case, e.g., with the name of the

district Holada (Vular) and the important ethnic designation of Khasa .s

21. It is impossible to read attentively Kalhana’s Chronicle and

in particular those portions which give fuller

occasion for the notice of localities, without

being struck with the exactness of his state¬

ments regarding the latter and with, what I

may call, his eye for matters topographical.

We must appreciate these qualities all the more if we compare

Kalhana’s local references with that vague and loose treatment which

topographical points receive at the hands of Sanskrit authors gener¬

ally.4 * * * If it has been possible to trace with accuracy the great majority

of localities mentioned in the Chronicle, this is largely due to the

precision which Kalhana displays in his topographical terminology. It

is evident that he had taken care to acquaint himself with the localities

which formed the scene of the events he described. Here too I may

refer for more detailed evidence to my translation of the work and the

notes which accompany it. A few characteristic points may, however,

be specified as examples.

Striking evidence for the care with which Kalhana indicates topo-

1 Compare also my paper on the ‘ Castle of Lohara,’ Ind Ant. 1897, p. 225 sqq.

below, § 49.

3 Compare for Gopadri , notes i. 341 ; viii. 1104-10 ; for the Mahasarit, note

iii. 339-349 ; for the Ksiptika, note viii. 732.

S See notes i. 306 and i. 317.

4 Nor should we forget the difficulty which Kalhana had to face by writing in

metrical form. True indeed it is what Alberuni says of this form as adopted by

Hindu scientific writers : “ Now it is well-known that in all metrical compositions

there is much misty and constrained phraseology merely intended to fill up the

metre and serving as a kind of patchwork, and this necessitates a certain kind of

verbosity. This is also one of the reasons why a word has sometimes one meaning

and sometimes another” ( India , i. p. 19).

Fortunately Kalhana has managed to escape these dangers as far as the topogra¬

phical notices of his work are concerned. We find in his local terminology neither

that mistiness nor multiplicity of meaning Alberuni so justly complains of.

Accuracy of

Kalhana’s

topograhy.

38

ANCIENT GEOGRAPHY OP KAS'MlR. [Extra No. 2,

graphical details, is furnished by his description of the great operations

which were carried out under Avantivarman with a view to regulating

the course of the Vitasta and draining the Valley.1 Thanks to the

exactness with which the relative position of the old and new confluence

of the Vitasta and Sindhu is described, before and after the regulation,

respectively, it has been possible even after so many centuries to trace

in detail the objects and results of an important change in the hydro¬

graphy of the Valley.2

Equal attention to the topographical details we find in numerous

accounts of military operations. Of these it will suffice to quote here

the descriptions of the several sieges of S'rinagar, under Sussala ;3 the

battle on the Gopadri hill in the same reign ;4 * the blockade of Lohara ,

with the disastrous retreat through the mountains that followed,6 and,

last but not least, the siege of the Sii'cihsila castle. The topographical

accuracy of the latter account as proved in Note L, viii. 2492, almost

presupposes on Kalhana’s part a personal examination of the site. It is

all the more noteworthy, because the scene of the events there recorded

was a region outside Kasmir proper, distant and difficult of access.

There are also smaller points that help to raise our estimate of

Kalhana’s reliability in topographical matters. Of such I may men¬

tion for example the close agreement we can trace everywhere between

Kalhana’s statements regarding distances, whether given in road or

time-measure, and the actual facts. The number of marches reckoned

by him is thus always easily verified by a reference to the stages ob¬

served on the corresponding modern routes.6 Not less gratifying is it

to find how careful Kalhana is to distinguish between homonymous

localities.7 In addition we must give credit to our author for the just

observation of many characteristic features in the climate, ethnography,

and economical condition of Kasmir and the neighbouring regions.8 All

these notices help to invest with additional interest the data furnished

for the old topography of the country.

1 Compare v. 84-121.

2 Compare Note I, v. 97-100, on the Vitasta sindhusaihgama, and below, §§ 69-72.

3 See viii. 729 sqq ; 1060 sqq.

4 Compare viii. 1099—1115.

& See viii. 1842-80 and Note E, iv. 177, § 10.

8 Compare for distance measurements note i. 264 ; v. 103 ; vii. 393 ; for the

reckoning of marches on the Vitasta Valley route, v. 225 ; on the T5s?maidan pass,

vii. 140 ; on the route to the PIr Pantsiil Pass, vii. 558 ; on the way to Martanda,

vii. 715, etc.

^ Compare notes i. 113 ; i. 124 ; v. 123 on the several Jyestharudras and the

way in which Kalhana specifies them.

8 Compare below, §§ 77-79.

1899.]

THE KAS'MIR CHRONICLES.

39

If the advantages thus accorded to us are duly weighed there

seems every reason to congratulate ourselves on the fact that the

earliest and fullest record of Kasmir history that has come down to us

was written by a scholar of Kal liana’s type. Whatever the short¬

comings of his work from a historical point of view may be, we may

well claim for him the merit that he has provided us with a sound and

ample basis for the study of the historical geography of his country.

22. Another point still remains to be considered here in connection

with Kalhana’s Chronicle, viz., to what extent

can we accept the Sanskrit forms found in his

text as the genuine local names of the period.

This question deserves attention, because the popular language actually

spoken in Kasmir in Kalhana’s time and for many centuries earlier,

was not Sanskrit but undoubtedly an Apabhramsa dialect derived from

it, which has gradually developed into the modern Kasmiri.

Notwithstanding this circumstance I think that Kalhana’s local

Sanskrit form of

local names.

names can on the whole safely be taken as the genuine designations of

the localities, i.e ., those originally given to them. My grounds for this

belief are the following.

We have ample evidence to show that Sanskrit was the official and

sole literary language of the country, not only in Kalhana’s own time

but also in those earlier periods from which the records used by

him may have dated. This official use of Sanskrit we know to have

continued in Kasmir even into Muhammadan times. It assures us at

once that the vast majority of village and town names must from the

beginning have been given in Sanskrit. A detailed examination of

Kalhana’s local names will easily demonstrate, on the one hand that

these names are of genuinely Sanskrit formation, and on the other,

that their modern Kasmiri representatives are derived from them by a

regular process of phonetic conversion. We look in vain among this

class of old local names for any which would show a foreign, i.e., non-

Aryan origin and might be suspected of having only subsequently been

pressed into a Sanskritic garb.

As Sanskrit was used as the language of all official records for

many centuries previous to Kalhana’s time, the Sanskrit names origin¬

ally intended for the great mass of inhabited places could be preserved,

in official documents anjdiow, without any difficulty or break of tradi¬

tion. And from such documents most of Kalhana’s notices of places

were undoubtedly derived, directly or indirectly.

Only in rare cases can we suppose that the original form of a local

name of this kind had been lost sight of, and that accordingly the Chro¬

nicler, or his authority, had to fall back on the expedient of sanskriti-

40

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

zing in its stead the Apabhramsa or Kasmirl form, as well as be could.

There are in fact a few instances in which we have indications of such

a metamorphosis. Thus we find the same local name spelt either

Bhaleraka or Baleraka in the Chronicle, and a village which Kalhana

calls Ghoramulaka , referred to by Abhinanda, the author of the Kadam-

barlkathasara (first half of 9th century), as Gauramulaka.1 It is difficult

to avoid the conclusion that we have here varying attempts to reproduce

in a Sanskritic garb original Apabhramsa names. But these cases are

very rare indeed, and even in them other explanations of the different

spellings are possible.

These observations apply with nearly the same force also to other

local names recorded in the Chronicle, such as those of mountains,

streams, passes, etc. The great majority of these names must have

very early found their place in official documents or, as we shall see

below, in the Sanskrit legendaries or Mahatmyas of the numerous

Tirthas. If any of them are in reality adaptations of Prakrit or Apa-

bhram^a forms, their quasi-official use is yet likely to have originated

a long time before the date of Kalhana.

Even to the present day the local nomenclature of Kasmir, whether

in the Valley or in the mountains, shows throughout an unmistakeably

Sanskritic character. This is most clearly illustrated by the constant

recurrence of such terms as - pur or par ( < pura), - mar (< matha),

-9hdm (< a&rama), -koth (< kotta), -gam or gom ( < grama), -kundal

(< kundala), -vor (< vata), in village names; of -sar (< saras),

-nambal (< nadvala), nag (< naga) in names of lakes, marshes, etc, ;

of -van ( < vana), - nar , (< nada), -marg (< mathika), -gal (< galika),

brgr (< bhattarika), -vath ( < patha) in designations of alpine localities,

peaks, passes, etc. ; -kul (< kulya), -khan (< khani) in names of streams

and canals.

The Sanskrit etymology of the specific names preceding these terms,

is even in their modern phonetic form very often equally transparent.

At an earlier stage of the language the Apabhramsa names must have

approached the corresponding Sanskrit forms much more closely.

The reproduction of the popular names in a Sanskrit form could have

then but rarely been attended with much difficulty or doubt. We may

hence safely assume that the Sanskrit forms recorded by Kalhana

represent in most cases correctly the original local names, and in the

remainder cannot differ much from them.

23. The later Sanskrit Chronicles which were composed with the

distinct object of continuing Kalhana’s work,

furnish valuable supplements to the topogra¬

phical information contained in the latter.

Later Sanskrit

Chronicles.

I Compare notes viii. 1861, and vii. 1239 ; viii. 2410.

1899.]

THE KAS'MIR CHRONICLES.

41

These Chronicles are the Rdj atarangini of Jonaraja who continued the

narrative down to the reign of Sultan Zainu-l-‘abidin and died over his

work, a.d. 1459 ; 1 the Jciina-Rajatarangini composed by Jonaraja’s

pupil S'rivara which deals in four Books with the period a.d. 1459-86 ;2

and finally, the Fourth Chronicle which was begun under the name

Rajavalipatakd by Prajyabhatta and completed by his pupil S'uka some

years after the annexation of Kasmir by Akbar, a.d. 1586. 3

It will be seen from the above dates that the narrative of the last

two works falls entirely beyond the period of Hindu rule to which our

enquiry is limited, and which may be considered to close finally with

the usurpation of Shah Mir, a.d. 1339. The same holds good of the

greater portion of Jonaraja’s Chronicle. The reigns of the late Hindu

rulers, from Jayasimha to Queen Kota, are there disposed of with a

brevity corresponding more to their own insignificance than to the

intrinsic historical interest of the epoch.4 Notwithstanding this dif¬

ference in date the materials supplied by these later Chronicles have

often proved of great use in clearing up points of the old topography

of Kasmir. For the mass of localities mentioned in them goes back

to the Hindu period, and the names by which they are referred to, are

also still mostly the old ones.

Yet on the whole the inferiority of these later Chronicles when

compared with Kalhana’s work, is as maiked in the matter of topo¬

graphical information as it is in other respects. In the first place it

must be noted that the whole text of these three distinct works does

not amount to more than about one-half of Kalhana’s work. For re¬

ferences to sacred sites and buildings and other places of religious

interest the account of Muhammadan reigns offers naturally but little

opportunity. The incidental notices of other localities are also in

proportion less numerous and instructive. For these later authors

allow considerably more room to episodic descriptions and do by no

means show that care for accuracy in topographical statements which

wTe have noticed in Kalhana.

It is curious to note how the gradual decline of Hindu learning in

Kasmir during the period of troubles and oppression which lasted with

short interruptions for two and a half centuries previous to Akbar’s

conquest, is marked also in the character and contents of these later

1 See S'rlv. i. 6.

2 See Fourth Chron. 6.

3 Compare Fourth Chron. 8 sqq. Prajyabhatta’s composition ended with the year

A.n. 1513-14 and the reign of Fatah Shah (verses 14-64).

4* The narrative of the period 1149-1339 a.d. fills only 305 verses in Jonaraja’s

Chronicle (347 according to the Bombay edition).

J. 1. 6

42

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Chronicles. Jonauaja was a scholar of considerable attainments, but

apparently without any originality. He shows himself yet well-

acquainted with the old local nomenclature of the Valley, though outside

it he too commits himself to forms like Purusavira (for Peshawar, recte

Purusapura), etc.

S'rIvara is a slavish imitator of Kalhana, not above reproducing

whole verses of his predecessor. His text looks often more like a cento

from the Rajatarangim than an original composition. Notwithstanding

the thorough study of Kalhana’s work which this kind of exploitation

presupposes, we find S'rIvara more than once betraying ignorance of the

old names for well-known Kasrnir localities. Thus we have the name of

the Mahasarit stream transformed into Mori , an evident adaptation of

the modern Mar;1 Siddhapatha , the modern Sidau, represented as

Siddhadesci ;2 the Tirtha of Martanda regularly referred to by its modern

name Bhavana (Bavan), etc.3

The work of Prajyabhatta and S'uka is inferior in composition

even to S'rlvara’s Chronicle, and by the increased number of modern

local names proves its authors’ scant familiarity with the old topography

of Kasrnir. Thus the ancient Krtyasrama , the scene of Kalhana’s

Buddhist legend, i. 131 sqq ., figures repeatedly in their narrative as

Ktcasrama , i.e ., by its modern name Kits^hom.4 * Even the well-known

liajapuri is metamorphosed into Rajavlra (! ), a queer reproduction of

the modern Rajauri.b The old castle of Lohara reappears as Luhara , an

evident approach to the present Loh^rin ;6 the ancient site of Cakra-

dliara is. turned into Cakradhara , etc.7

It is evident that when Sanskrit ceased to be the language used for

official purposes, the knowledge of the ancient names of localities and

of the traditions connected with the latter must have become gradually

more and more restricted. In view of this decrease of traditional

knowledge we have to exercise some caution when utilizing the evidence

of the later historical texts for the elucidation of the old topographical

data. At the same time it is easy to realize that their help is often of

considerable value when connecting links have to be traced between

those earlier data and the facts of modern topography.

1 See 8'riv. i. 440 ; iii. 278 ; comp, note on Rdjat. iii. 339.

2 S'riv. iii. 354; iv. 203, 661.

3 S’riv. i. 376 ; iii. 372.

* See Fourth Chron. 234, 240, 384; compare also note on Rdjat . i, 147,

& Fourth Chron. 542, sqq.

6 lb., 134, 143, sqq.

T lb., 330.

1899.]

THE KAS'MIR CHRONICLES.

43

Persian Tarlkhs.

24. It is convenient to refer here briefly to the Persian Tariklis

of Kasmir which to some extent may be looked

upon as continuing the works of Kalhana and

his Pandit successors. Unfortunately they furnish no material assistance

for the study of the old topography of the country.

All these works give in their initial portion an account of the

Hindu dynasties which pretends to be translated from the Raja-

tarangini. Yet the abstract so given is in each case very brief and

chiefly devoted to a reproduction of the legendary and anecdotal parts

of Kalhana’s narrative. We thus look in vain in these abstracts for

the modern equivalents of those local names, the identification of which

is attended with any difficulty.

In illustration of this it may be mentioned that even the Tarikh

of Haidar Malik Cadura (TsadV),1 which is the earliest work of this

class accessible to me and the fullest in its account of the Hindu period,

compresses the narrative of Jayasimha’s reign, filling about two thousand

verses in the Rajatarangini, into two quarto pages. Of the localities

mentioned in the original account of this reign not a single one is

indicated by the Muhammadan Chronicler.

The later works which all belong to the 18th or the present

century, are still more reticent on the Hindu period and seem to have

largely copied Haidar Malik’s abstract. Taking into account the endless

corruptions to which local names written in Persian characters are

exposed, it will be readily understood why reference to these texts

on points of topographical interest yields only in the rarest cases some

tangible result.

25. It is a fortunate circumstance that several of the older Kasmir

poets whose works have been preserved for us,

have had the good sense to let us know some¬

thing about their own persons and homes. The topographical details

which can be gleaned from these authors, though comparatively few in

number, are yet of distinct value. They enable us to check by independ¬

ent evidence Kalhana’s local nomenclature, and in some instances

acquaint us with localities of which we find no notice in the Chronicles.

The first and most helpful of these Kasmir ian authors is the well-

known polyhistor Ksemendra. His works, composed in the second and

third quarter of the 11th century, form important landmarks in various

fields of Indian literature. Ksemendra seems to have felt a genuine

Kasmir poets.

1 Written a.h. 1027, i.e., a.d. 1617, in the twelfth year of Jahangir’s reign.

Haidar Malik takes his epithet Cadura, recte Tsad^r, from the Kasmir village of that

name situated in the Nagam Pargana, some 10 miles south of S'rlnagar, close to the

village of Yah^tdr.

44

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

interest, rare enough among Indian scholars, for tlie realities of his

country and the life around him. He does not content himself with

informing us of his family, the date of his works and the places where

he wrote them.1

In the SamayamdtrJcd , one of his most original poems, which is

intended to describe the snares of courtezans, he gives among other

stories an amusing account of the wanderings of his chief heroine,

Kaiikall, through the length and breadth of Kasmir.2 * The numerous

places which form the scene of her exploits, can all easily enough be

traced on the map. More thau once curious touches of true local colour

impart additional interest to these references. To Ksemendra’s poem

we owe, e.g ., the earliest mention of the Pir Pantsal Pass ( Pahcala -

dhara) and its hospice (mat ha)? There too we get a glimpse of the

ancient salt trade which still follows that route with preference. Else¬

where we see the heroiue smuggling herself as a Buddhist nun into the

ancient Vihara of Krtydsrama , etc.4 *

A different sketch of topographical interest we owe to the poet

Bilhana. He left his native land early in the reign of King Kalasa

(1063-89 a.d.), and after long wanderings became famous as the court

poet of the Calukya king Tribhuvauamalla Parmadi in the Dekhan.

In the last canto of his historical poem, the Vikramcihkadevacarita ,

Bilhana gives us a glowing picture of the beauties of the Kasmir

capital. Notwithstanding its panegyrical character, this account is laud¬

ably exact in its local details.6 * 8 In another passage the poet describes

to us his rural home and its surroundings at the village of Khonamusa ,

south-east of S'rlnagar. His touching verses attest as much his year¬

ning for his distant home as the faithfulness of his local recollections.6

1 Compare the colophons of the various works first discovered and noticed by

Prof. Buhler, Report, pp. 45 sqq. and Appendix.

2 This hnmorous peregrination fills the ii. Samaya of the work ; see Kdvyamdld

edition, pp. 6-16. The abundance of curious local details makes a commentated

translation of the little Krivya very desirable, notwithstanding the risky nature

of part of its contents. A personal knowledge of Kasmir would certainly be

required for the task.

8 See Samayam. ii. 90 sqq. The matha on the pass corresponds to the present

‘ Alt lab dd Scirai, a short distance below the top of the pass on the Kasmir side

see below, § 44.

4 Samayam. ii. 61 sqq.

& Prof. Buhler to whom we owe the discovery of Bilhana’s chief work, has

given in his Introduction an admirable analysis of the contents of Sarga xviii. as

illustrating the poet’s biography. For his description of contemporary S'rlnagara,

see pp. 7 sqq.

8 See Viler am. xviii. 70 sqq. Prof. Buhler during his Kasmir tour, 1875, had

the satisfaction of visiting the poet’s native place, the present village of KhunQmoh.

1899.]

THE KAS'MIR CHRONICLES.

45

Similar in character though less ample in detail, is the description

of Kasmir and its capital Pravarapura which Mankha, Kalhana’s

contemporary, inserts in the iii. Canto of his Kavya Srlkanthacaritn.1

Here we have the advantages of a commentary written by Jonaraja, the

Chronicler, which duly notices and explains the points of local interest.

26. To complete our review of those Kasmirian texts of topo¬

graphical interest which may be distinguished

The Lokapraka§a. ,

as secular, we must reler briefly to the curious

glossary and manual which goes by the name of Ksemendra’ s Lokapra-

knsa. Professor A. Weber has recently published valuable extracts

from this text.2 I myself have had occasion to refer to it frequently

in the notes on the Rajatarangini.3 The work represents a strange

mixture of the usual Kosa and a practical handbook dealing with various

topics of administration and private life in Kasmir.

A great deal of the information contained in it is decidedly old, and

probably from the hand of our well-known Ksemendra. But there are

unmistakeable proofs, both in the form and contents of the book, showing

that it has undergone considerable alterations and additions down even

to the 17th century. This is exactly what we must expect in a work

which had remained in the practical use of the Kasmirian ‘ Karkuns ’

long after the time when Sanskrit had ceased to be the official language

of the country.

The Lokaprakasa supplies us with the earliest list of Kasmir

Parganas. It gives besides the names of numerous localities inserted

in the forms for bonds, ‘ Hundis,’ contracts, official reports, and the like

which form the bulk of Prakasas ii. and iv. The Pargana list as well

as these forms exhibit local names of undoubtedly ancient date side by

side with comparatively modern ones. Some of the latter belong to

places which were only founded during the Muhammadan rule.4

He conlcl thus verify on the spot every point of the description which Bilhana gives

of that “coquettish embellishment of the bosom of Mount Himalaya;” see Report,

pp. 4 sqq.

1 See S'rikanthac. iii. 10—24, 68 sqq.

2 See Zu Ksemendra' s Lokaprakasa , in Indische Studien, xviii. pp. 289-412.

3 See particularly Note H (iv. 495), on the Kasmir monetary system, § 10.

* Compare, e.g., in Prakasa ii. Jainanagara, founded by Zainu-l-‘abidin (see

Jonar 1153); Aldbhadenapura ( S'riv . iv. 318), etc.

46

ANCIENT GEOGRAPHY OF KASHMIR. [Extra No. 2,

Section VI. — The Nilamata and Mahatmyas.

The Nilamata-

purana.

27. We Lave already above drawn attention to the fact that

Kasmir has since early times been pre-emi¬

nently a country of holy sites and places of

pilgrimage of all kinds. These objects of

ancient local worship have always played an important part in the

historical topography of the Valley and the adjacent mountain regions.

It is hence no small advantage that there are abundant materials at

our disposal for the special study of this Topographia sacra of Kasmir.

The oldest extant text which deals in detail with Kasmirian Tlrthas,

is the Nilamatapurana. This work which Kalhana used as one of his

sources, J claims to give the sacred legends regarding the origin of the

country and the special ordinances which Nila, the lord of Kasmir

Nagas, had revealed for the worship and rites to be observed in it.8

It is unnecessary to refer here to the legends which are related at the

commencement of work, and to ‘ the rites proclaimed by Nila’ which

together with the former occupy about two-thirds of the extant text.1 2 3

These parts have been fully discussed by Prof. Biihler in his lucid

analysis of the Nilamata.4 The remaining portions, however, deserve

here special notice as forming, — to use Prof. Biihler’s words — “a real

mine of information, regarding the sacred places of Kasmir and their

legends.”

In the first place we find there a list of the principal Nagas or

sacred springs of Kasmir (vv. 900-975). This is followed by the in¬

teresting legend regarding the Mahapadma lake, the present Volur,

which is supposed to occupy the place of the submerged city of Candra-

pura (vv. 976-1008). 5 The Purana then proceeds to an enumeration

of miscellaneous Tlrthas chiefly connected with S'iva’s worship (vv.

1009-48). To this is attached a very detailed account, designated as

Bhutesvaramahatmya , of the legends connected with the sacred lakes

and sites on Mount Haramukuta (vv. 1049-1 148).6 Of a similar

Mahatmya relating to the Kapatesvara Tirtha, the present K5ther,7

only a fragment is found in our extant text (vv. 1149-68). The list of

1 See Raj at. i. 14.

2 Compare Rdjat. i. 178-184.

3 Nilamata, vv. 1-366, contain the legends, v. 367-899 the rites above referred to.

4 See Report, pp 38 sqq.

8 Compare below, § 74, and Report, p. 10.

8 Compare below, § 57, and Rdjat. notes i, 36, 107, 113.

T See below, § 112, and Rdjat. i. 32 note.

THE NILAMATA AND MAHATMYAS.

47

1899].

Visnu-Tirthas which succeeds it (vv. 1169-1248), is comparatively

short, as indeed the position of this god is a secondary one in the

popular worship of Kasmir.

After a miscellaneous list of sacred Samgamas or river-confluences,

Kagas and lakes (vv. 1249-78) we are treated to a somewhat more

detailed synopsis of the chief Tirthas of Kasmir (vv. 127J-1371).

This is of special interest, because an attempt is made here to describe

the Tirthas in something like topographical order, and to group with

them such localities as are visited on the same pilgrimage. It is thus

possible to determine, with more certainty than in the case of other

Tirtha lists, the particular holy sites intended by the author.

This synopsis starts in the east with the fountain of the Kilanaga

(Vernag), and follows with more or less accuracy the course of the

Vitasta and its affluents down to the gorge of Varahamula. A short

Vitastamahatmya, describing the origin and miraculous powers of this

the holiest of Kasmir rivers (vv. 1371-1404), closes the text of

Nilamata, such as it is found in our Manuscripts.

This text is unfortunately in a very bad condition owing to numer¬

ous lacunae and textual corruptions of all kinds. Prof. Biihler held

that the Nilamata in its present form could not be older than the 6th

or 7th century of our era.1 It appears to me by no means improbable

that the text has undergone changes and possibly additions at later

periods. On the whole, however, the local names found in it bear an

ancient look and agree closely with the forms used by Kalhana. The

difference in this respect between the Kilamata and the Mahatmyas, in

their extant recensions, is very marked and helps to prove the compara¬

tively late date of most of the latter. On the other hand it deserves

to be noted that without the more systematic and detailed accounts of

the various Tirthas as found in the Mahatmyas, the identification of

many of the sacred places referred to in the Kilamata would probably

have been impossible.

The fact of all extant copies of the work showing practically

the same defective text, seems to indicate that the changes and

additions to which I alluded above, cannot be quite recent. If such

a revision had been made at a time comparatively near to the date of

our oldest MS. we could, after the analogy of other instances, expect

an outwardly far more correct, i.e. ‘ cooked,’ text. The operation here

suggested was actually performed some thirty years ago by the late

Pandit Sahibram. Receiving the orders of Maharaja Ranbir Singh to

1 Compare Report , p. 40. The oldest and best MS. of the Nilamata which I

was able to secure and collate, is dated in the Laukika year 81. This date judging

from the appearance of the MS. probably corresponds to a.d, 1705-6.

48

ANCIENT GEOGRAPHY OF KAS'MlR.

[Extra No. 2,

The Haracarita-

cintamani.

prepare the text of the Nilamata for edition, he ‘revised’ the work with

scant respect for its sacred character by filling up the lacunte, expanding

obscure passages, removing ungrammatical forms, etc.1 Fortunately

Prof. Biililer reached Kasmir early enough to learn the origin of this

‘ cooked ’ text, and to give due warning as to its true character.

The Nilamata seems thus to have escaped in recent times that

process of continual adaptation which, as we shall see, must be assumed

to have greatly affected all extant Mahatmyas. The reason probably

is that it could never have been used, like the latter, as a practical

pilgrims’ manual and itinerary by the Purohitas of the various Tirthas.

28. Among the texts dealing specially with the sacred sites of

Kasmir the Haracaritacintamani can be placed,

perhaps, nearest in date to the Nilamata-

purana. It is not like the latter and the

Mahatmyas, an anonymous composition, claiming recognition in the

wide folds of canonical Purana literature. It owns as its author the poet

Jayadratha, of the Kasmhian family of the Rajanakas, and a brother of

Jayaratha. The pedigree of the family as given iu Jayaratha’s

Tautrdlokaviveka , a S'aiva treatise, shows that Jayadratha must have lived

about the end of the 12tli or beginning of the 13th century.2

His work which is written in a simple Kavya style, relates in

thirty-two Cantos as many legends concerning S'iva and his various

Avataras.3 Eight of these legends are localized at well-known Kas-

mirian Tirthas. They give the author ample opportunity of mentioning

other sacred sites of Kasmir directly .or indirectly connected with the

former.4

Jayadratlia’s detailed exposition helps to fix clearly the form which

the legends regarding some of the most popular of Kasmirian Tirthas

had assumed in the time immediately following Kalhana. The local

names as recorded by Jayadratha, agree closely with those of the

Rajataraiiginl.6 They prove clearly that the forms employed by Kalhana

must have been those generally current in the Sauskrit usage of the

period. For the interpretation of Nllamata’s brief notices the Hara-

1 See Report, pp. 33, 38.

2 Compare Buhler, Report, pp. 61, 81, cliii.

3 The Haracaritacintamani has recently been printed as No. 61 of the Kdvyamdld

Series, Bombay, (1897), chiefly from the text as contained in my MS. No. 206.

4 The cantos containing these legends are i. Jvalalihgavatai’a, iv. Nandirudra-

vatara, vii. Cakrapradana ; x.-xiv. Vijayesvara-, Pihgalesvara-, Vitasta-, Svayambhu-

natha-, Kapatesvara Avataras.

6 An index of the Kasmir local names in the Haracaritacintamani, with explan¬

atory notes, has been prepared under my supervision by P. Govind Kaul and printed

as an Appendix to the Kavyamala edition.

1899.]

THE N TEAM AT A AND MAHATMYAS.

49

The Mahatmyas.

caritacintamani is of great value. Its plain and authentic narrative

enables us often to trace the numerous modifications which the various

local legends as well as the names of the localities connected with

them have undergone in the extant Mahatmyas.

29. Reference has already been made above to the numerous texts

known as Mahatmyas which we possess of

all the more important Tirthas of Kasmir.

They claim with few exceptions to be extracted from Puranas or

Puranic collections ( Samhitcis).1 Ordinarily they set forth in detail

the legends relating to the particular pilgrimage place, the spiritual

and other benefits to be derived from its visit, and the special rites

to be gone through by the pilgrims at the various stages of the itinerary.

The abstract given of the S'aradamahatmya in Note B, on Rajat. i. 37,

may serve to indicate the manner in which these subjects are usually

treated in the average texts of this class.

Prof. Biihler was the first to recognize the value of the Mahatmyas

for a systematic study of the old topography of Kasmir. Among the

Sanskrit Manuscripts which he acquired during his tour in Kasmir,

there are sixteen distinct texts of this kind.2 My own search in this

direction, facilitated by successive visits to the various Tirthas them¬

selves, has enabled me to collect altogether fifty-one separate Mahatmya

texts. The list of my collection which has been given in a supple¬

mentary Note,3 may be considered fairly to exhaust the present range

of this literature.

In extent the Mahatmyas vary greatly. By the side of texts like the

Vitastamdlidtmya with its fifteen hundred STokas, we have legendaries

of more modest dimensions amounting only to a few dozens of verses.

Equally marked differences in the matter of age become apparent on

closer examination.

Unmistakeable indications prove that many of the Mahatmyas now

in actual use are of late composition or redaction. Among the texts so

characterized, the Mahatmyas of some of the most popular pilgrimage

places, like the Haramukuta lakes, the cave of Amaranatha, Isesvara

(Isobar), are particularly conspicuous. The indications here referred to

are furnished chiefly by the local names which in their very form often

betray a modern origin. This may conveniently be illustrated by a

1 Most of the Kasmir Mahatmyas allege to be portions of the Bhrnglsasamhitd.

Others claim special authority by representing themselves as parts of the Adi,

Brahma, Brahmavaivarta, Varaha and Bhavisyat Puranas.

2 See Report, pp. iv. sqq. Nos. 48, 51, 52, 55, 62, 75, 82, 84, 99, 100 there quoted

as separate texts are only chapters of the Amarandthamdhdtmya.

3 See Supplementary Note AA.

J. i. 7

50 ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

brief analysis of the most instructive of sucli names found in the

Vitastamfthatmya.

This text claims to furnish an account of all the Tirthas along the

course of the holy river and is designated as a portion of the Bhrnglsa-

samhita. Notwithstanding this pretended antiquity we find the famous

Nilanaga introduced to us by the name of Viranaga (i. 58; ii. 33).

This form is wholly unknown to the Nilamata, Rajatarangini or any

old text. It is nothing but a clumsy rendering of the modern name of

the village Vernag near which this fine spring is situated.1 The

ancient site of Jayavana, mentioned by Bilhana and Kalhana, the pre¬

sent Zevan , is metamorphosed into Yavanl (vi. 4). 2 The village of

Pandrethan which derives its name from Puranadhisthana , 3 ‘the old

capital,’ and bears the latter designation even in S' rivara’s Chronicle,

figures as Pddadrstika (!), xii. 24. That Maksikasvamin (May^sum)4 *

and the Mahasarit (Mar) 6 7 appear as Mayasimd and Marl , can after this

specimen of fancy nomenclature scarcely surprise us.

But we must all the same feel somewhat startled when we find that

this text which claims to be revealed by S'iva, refers repeatedly to the

modern village of Shad1 pur, at the confluence of the Vitasta and Sind,

by the name of Sdradapura. Shad* pur, an abbreviation for Shahabuddln -

pur , was, as Jonaraja’s Chronicle shows, founded only in the J4th

century by Sultan Shahabu-d-din.6 Quite on a level with the knowledge

of old topography here displayed are many other references to loca¬

lities, e.g ., the mention of the modern garden Shdlimdv, a creation of

the Mughals (S'alamara), xxi. 39 ; of the ancient Huskapura as TJsah-

karanal (for Uskiir !), xxix. 103, etc.

In several cases these fancy renderings of modern local names

are explained by whimsical etymologies which again in due turn give

rise to new-fangled legends quite in the style of the old niddnakathds.

Similar proofs of modern origin can be traced in several other

popular Mahatmyas, though perhaps not with equal frequency. Thus

we find in the Haramukutagangdmahdtmya the name of the sacred

mountain itself transformed from Haramukuta into Haramukha (the

1 The name Vernag is probably derived from the name of the Pargana Ver, men¬

tioned by Abu-1-Fazl, ii. p. 370.

2 See below, § 105.

3 See Rdjat. iii. 99 note and below, § 89 ; also Srlv. iv. 290.

4 See Rajat. iv. 88 note and below, § 99.

6 Compare Rajat. iii. 339-349 note and below, § 65.

8 See Jonar. 409. A popular etymology accepted in good faith by more than

one European writer, sees in Shad^pur the * village of the marriage,’ soil, between

the Vitasta and Sind Rivers !

7 Compare Rajat . i. 168 note and below, § 124.

1899.]

THE NILAMATA AND MAHATMYAS,

51

present Haramukh),1 the ancient site of Bhutesvara (ButlPser) so well-

known to the Rajatarangini and all old texts, turned into Bodhesvara ,

etc. In the Amaranathamdhatmya of which there is a comparatively

old copy in the Poona collection, we are also treated to Padrsti as the

Sanskrit name of Pandrethan, to Susramanaga (for Ks. Susramnag )

as the name of the lake where the Naga Susravas of the old legend

took up his abode,2 and the like. Examples of local names similarly

perverted in other Mahatmyas will have to be mentioned passim in

our account below.

It is important to note that by the side of texts like those just

mentioned, there are others which on the whole show close conformity

with our genuine old sources both in matter of legend and local names.3

And even lin the Mahatmyas which in their present form we have

every reason to consider as recent compositions, there is often abundant

evidence of the use of earlier materials and traditions.4 It will be

easier to understand the singular discrepancies in the value and charac¬

ter of these texts on examining the peculiar conditions under which

they have originated.

30. The Mahatmyas are in the first place hand-books for the

Purohitas of the particular Tirthas who have

Origin ^nd purpose privnege of taking charge of the pilgrims,

of Mahatmyas. They serve the priests as chief authorities for

the claims they put forth on behalf of the holiness of their Tirtha,

and for the rewards they promise for its visit. They are also intended

to support their directions as to the rites to be observed by the pilgrim,

and the route to be taken by him on the journey. It is usual for the

Purohitas to recite the Mahatmya for the benefit of their clients in the

course of the pilgrimage tour. At the same time its contents are

expounded to them by a free verbal rendering in Kasmiri.

1 See below, § 57. The kli at the end of the modern name is due to a phonetic

law of Kasmiri which requires the aspiration of every final tenuis; see J. A. S. B.,

1897, p 183.

2 Compare Raj at. i. 267 note. The modern Ks. form Susramnag is the regular

phonetic derivative of Susravanaga by which name the lake is designated in the

Nllamata, Haracaritacintamani, etc.

8 Among such the Mahatmya collection known as the S'arvavatdra (No. 213

in my list of MSS.), the Mdrtdndamahdtmya (No. 219), the Vijayesvaramdhdtmya

(No. 220), may be particularly mentioned. None of these, however, are now known

to the local Pnrobitas, more recent and inferior texts having taken their place.

4 Thus e.g., the Mahatmya of the present Isobar ( Isesvara ; see Hajat. ii. 134)

shows plainly its very recent origin by calling the Tirtha Tsavihdra (a garbled

reproduction of I&bror < Isesvara), and by similar blunders. Yet it knows correctly

the sacred spring of S'atadhdrd already mentioned by Ksemendra.

52

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

As but very few of the priests have enough knowledge of Sanskrit

to follow the text intelligently, these translations are more or less

] earned by heart. Often as my manuscripts show, interlinear Kasmiri

glosses are resorted to in order to assist the reader’s memory.

These local priests known now in Kasmir as thdnapatl (Skr. sthana-

pati), are as a rule quite as ignorant and grasping as their confreres,

the Pujaris, Bhojkis, etc., of India proper. They are held deservedly

in very low estimation by the rest of the Brahman community. That

their condition was more or less the same in earlier times too, though

their influence and numbers may have been greater, can be safely con¬

cluded from more than one ironical allusion of Kalhana.1 These are

the people to whose keeping the Mahatmya texts have always been

entrusted. Their peculiar position and calling explain, I think, most

of the curious changes which the latter have undergone.

Tenacious as local worship is, there is the evidence of concrete

cases to show that not only the route of pilgrimage, but the very site of

a Tlrtha has sometimes been changed in comparatively recent times.

In proof of this it will suffice to refer to the detailed account I have

given of the transfers that have taken place in the case of the ancient

Tirthas of Bhedd and Sarada .2 Minor modifications must naturally have

been yet far more frequent. The visit of a principal Tirtha is regularly

coupled with bathings, S'raddhas and other sacrificial functions at a

series of other sacred spots. The choice of these subsidiary places of

worship must from the beginning have depended on local considerations.

As these changed in the course of time, variations in the pilgrimage

route must have unavoidably followed.

To bring the text of the Mahatmya into accord with these succes¬

sive changes was a task which devolved upon the local Purohitas. The

texts we have discussed above bear, in fact, only too manifestly the traces

of their handiwork. Sound knowledge of Sanskrit and literary culture

are likely to have been always as foreign to this class of men as

they are at present. When it became necessary for them to introduce

the names of new localities into the text of the Mahatmya there was

every risk of these names being shown not in their genuine old forms,

but in hybrid adaptations of their modern Kasmiri equivalents. This

risk naturally increased when Sanskrit ceased to be the official

language of Kasmir, and the knowledge of the old local names was

gradually lost even among those maintaining scholarly traditions in the

country.

1 Compare Rajat. ii. 132 note and v. 465 sqq. ; vii. 13 sqq.-, viii. 709, 900 s qq-, 939.

* Compare Notes A (Rajat. i. 35) and B (Rajat. i. 37).

THE NILAMATA AND M AH ATM Y AS.

53

1899.]

31. Another potent cause seems to have co-operated in this vitiation

of the local nomenclature of the Mahatmyas.

Popular etymology j mean ‘ popular etymology.’ We have already

m local names of , , 5 , j j. . , ,, , ,

Mahatmyas referred to the tendency displayed throughout

these tracts of making the names of localities,

rivers, springs, etc., the starting-point for legendary anecdotes. For men

of such very scant knowledge of Sanskrit as the thanapatls invariably

are, it was naturally far easier to explain such etymological stories when

they were based on the modern local names.

It is undoubtedly this reason which has, e.g„ led the compiler of the

present Haramukutagangamahatmya to substitude the name Karanka-

nadi for the old Kanakavdhinl. By the latter name the stream coming

from the Haramukuta lakes is designated in all our old texts, as ex-

plained in my note on Rajat. i. 149-150. By turning Kankanai, the

modern derivative of this old name, into Karankanadi, ‘ the skeleton-

stream/ the compiler of the Mahatmya gets an occasion to treat his

readers to a legend likely to appeal to their imagination. The river

is supposed to have received this appellation, because Garuda had

dropped at its Samgama with the Sindliu the skeleton ( karanka ) of

the Rsi Dadliici which indra before had used as his weapon, etc.1

This story, it is true, is wholly unknown to the Nilamata or any other

old text. But, on the other hand, it has got the great merit of being

easily explained and proved to any Kasmiii pilgrim. He cannot fail

to realize the manifest connection between Karanka and his familiar

karanz , ‘skeleton.’

An exactly similar case of ‘ popular etymology ’ has been noticed

in the analysis of the S'aradamahatmya as contained in my Note B

(i. 37). There the name of the village SunQ-Drang is reproduced as

Suvarncirdlidhgaka and explained by a legend, how the Muni S'andilya

had at that spot half his body ( ardhangaka ) turned into gold ( suvarna ),

etc. In reality the village name is derived from the old term Dranga,

‘ watch-station,’ by which the place is mentioned by Kalhana.2 The

distinguishing prefix Sun ?-, meaning ‘ gold ’ in Kasiniri, was given to it,

because it lay on the route to the old gold-washing settlements in the

Kisangaiiga Valley.3

1 The story is spun out at great length in Patala iii. of the Haramukutaganga-

mahdtmya , MS. No. 221.

2 See viii. 2507, 2702.

& For other examples of local names in Mahatmyas metamorphosed for the

above reason, compare my notes ltajat. vi. 177 ( Bhhnudvipa in the Martandamah., for

Bum^zu); i. 267 (S'esandga in the Amaresvaramah., for the older Susramandga , recte

Susravondga ) ; Note 0, i. 124 ( Jyesthesvara , the present Jyether, turned into a site

of Jyestha), etc.

54

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

It would be easy to multiply examples showing the strange vicissi¬

tudes to which old topographical names are exposed at the hands of the

local Purohita. But the explanations already given will suffice to prove

that the topographical data found in Mahatmyas can only then be used

safely when they are critically sifted and supported by our more reliable

sources.

A critical examination of these data is, however, much impeded

by the difficulty we experience in fixing the exact age of particular

Mahatmyas and their component portions.1 Even in the case of ap¬

parently old texts modern additions and changes may be suspected,

while again the most recent concoctions may preserve fragments of

genuine tradition.2 In view of these considerations I have not- thought

it safe to crowd my maps with hundreds of names of petty Tirthas

as found in the Mahatmyas, but have marked only those pilgrimage

sites the ancient names of which can be established with certainty.

1 The difficulty here indicated is increased by the fact that no really old

manuscripts of Mahatmyas seem to be preserved in Kasmir. MSS. written on

birch-bark, i.e., earlier than the 17th century, are quite unknown at present. Of the

numerous paper MSS. I have examined, none seem to me older than two cen¬

turies at the utmost. It is probable that this absence of older copies is due to the

rough usage to which Mahatmya MSS. are exposed when carried about on the

pilgrimage tours.

2 I am glad that chance gave me an opportunity of gaining some personal ex¬

perience of the mauner in which Mahatmyas are occasionally produced. Some ten

years ago the Purohitas or Bachbattas of the Ganapatyar quarter in SYinagar reco¬

vered an ancient Lihga from a Mosque and began to erect a small shrine for it' near

the river Ghat of Malayar. Guided by a local tradition which, as far as I can judge,

may be genuine, they believed this to have been the site of the shrine of S'iva

Vardhamanesa mentioned already in the Ra j ataranginl (see note ii. 123), The Lihga

was re-consecrated accordingly by this name.

In 1891, when examining old sites in this part of the city, I also visited the

temple of Vardhamanesa then under construction. The interest I showed in the

old Lihga and in the tradition regarding it, coupled with an appropriate Daksina,

soon secured me the confidence of the head-Purohita of the little shrine. ‘ Pandit 5

T.R., a man more intelligent than the average of his fraternity, was not slow to

confess to me that the Mahatmya of the Tirtha in spe was as yet under preparation.

Some weeks later when in camp near S'rlnagar, I received the visit of my Purohita

from Vardhamiinesa’s shrine. He brought me the draft of the new Mahatmya

and asked my assistance in revising it.

I found it to consist chiefly of extracts from the Vitastamahatmya. The passages

dealing with Vardhamanesa and the neighbouring Tirthns within the city had been

suitably amplified with laudatory verses in the usual Mahatmya style culled from

other texts. The vested interests of other local shrines had received due recogni¬

tion by being included in the Ydtrd of Vardhamanesa. I did what I could to indicate

the genuine names of these localities. This quasi-antiquarian co-operation does

not seem to have detracted from the popularity of the new Mahatmya among the

Bachbattas of Ganapatyar.

1899.]

THE NILAMATA AND MAHATMYAS.

55

Abu-l-FazPs account

of TIrthas.

32. It is a curious fact that among our authorities for the Topogra-

pliia sacra of Kasmir we must allow a con¬

spicuous place to a Muhammadan writer. It

is Abu-l-Fazl, the minister of Akbar, who in

the chapter of his Ain-i Akbarl dealing with the ‘Sarkar of Kashmir ’

has left us a very accurate account of many of the holy places in the

Valley.1 Abu-l-Fazl’s detailed description of Kasmir is valuable in

many respects to the historical student. But it is particularly in con¬

nection with our topographical search that we must feel grateful to the

author for having like his great master “ caught some of the enthusiasm

of the Valley ” (Rennell).

Abu-l-Fazl tells us that “ the whole country is regarded as holy

ground by the Hindu sages.” He also refers in general terms to the

numerous shrines dedicated to the various deities and to the popular

worship of ‘ snakes,’ i.e ., the Nagas, “ of whom wonderful stories are

told.” He then proceeds to describe in detail the most notable sites,

giving among these particular prominence to what Dr. Bernier aptly

called ‘ les merveilles ’ of the country.

This account of Abu-l-Fazl represents for us an authentic survey

of all the Kasmirian Tirthas that were well-known and popular at the

end of the 16th century. It serves as a most useful link between

our older texts dealing with these pilgrimage places and the modern

tradition. It helps us to check the data of the Mahatmyas in many

particulars of topographical interest. Abu-l-Fazl’s notes have enabled

me to trace in more than one instance the position of ancient Tirthas

or particular features regarding them which have since his time been

wholly forgotten.2 It cannot be doubted that Abu-l-Fazl’s list of

sacred sites to which we have to refer so frequently in our subsequent

notes, was supplied by competent Brahman informants just as his

abstract of the Sanskrit Chronicles.

1 Yol. i. pp. 564-570 in Prof. Blochmann’s edition of the Aln-i Akbarl j vol. ii.

pp. 354-366 in the Bibliotheca Indica translation of the work (Col. H. S. Jarrett).

Abii-l-Fazl’s account of Kasmir would well deserve a fuller commentary than the

one which the translator, in the absence of special local studies, was able to give.

The account of Mirza Haidar (in the Tarikh-i Rashidi) and Bernier’s notes could

conveniently be discussed on the same occasion.

2 Compare my notes on Bhedagiri (i. 35), the S'dradatirtha (i, 37), the

Tdksaicandga (i. 220) ; also supplementary note to i. 107.

56

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Section VII. — Local tradition.

Local tradition of

the learned.

33. It now remains for us only to indicate briefly what help

surviving tradition offers for the study of

the ancient topography of Kasmir. The tradi¬

tion with which we are here concerned, presents

itself in two forms. One is the tradition of the ‘ learned,’ regarding the

ancient sites of the country in general, kept up more or less in connec¬

tion with written records. The other is that genuine local tradition

which is strictly confined in its limits but is kept up equally among

literate and illiterate of particular places.

Among those who represent in Kasmir learned tradition of the

former type there must again be distinguished the few Pandit families

of S'rinagar in which the serious study of Sanskrit S'astras has been

maintained, and the great host of ‘ Bachbattas.’ With the latter class

we have already become partially acquainted in the course of our

examination of the Mahatmyas. We have had occasion to note the

conspicuous absence of genuine knowledge as regards the ancient

topography of the country in those texts which form the character¬

istic products of this class’ literary activity.

The Purohitas’ knowledge of Sanskrit is ordinarily of the scantiest

kind, and their ‘ reading ’ confined to Mahatmyas and devotional texts

learned by heart without proper comprehension. We can hence scarcely

expect them to have preserved genuine traditions regarding those

historically interesting localities which are mentioned only in the

Chronicles. It is only in the matter of those sacred sites, pilgrimage

routes and the like which form as it were, their own particular profes¬

sional domains, that their testimony can claim special attention. Yet

even in this limited field the Purohitas’ traditions are, as we have seen,

often of a very modern growth. Their statements, therefore, require

under all circumstances to be tested with critical caution.

34. ‘ Learned ’ tradition as represented by the S'rlnagar Pandits

of modern times, is best guaged by an ex-

Sahibram s TIrtha amination of what the late Pandit Sahibram

sanigra a ^ 1872) has specially recorded on the sub¬

ject of ancient sites.

P. Sahibram who was undoubtedly the foremost among Kasmirian

Sanskrit scholars of the last few generations, had been commissioned by

the late Maharaja Ranbir Singh to prepare a descriptive survey of all

ancient Tirthas of Kasmir. For this purpose a staff of Pandits was

placed at his disposal whose business it was to collect the necessary

1899.]

LOCAL TRADITION.

57

materials in the various parts of the country. The large work which

was to he prepared on the basis of these materials, was never completed,

and of the latter themselves I was able to recover only small portions.1

But some time before his death Pandit Sahibram had drawn up

abstracts of the information he had collected under the title of

Kasmlratirthasamgraha , and of these I have been also able to obtain

copies. The most detailed and apparently latest recension of this

Tirthasaihgraha is the one contained in No. 61 of Prof. Biihler’s col¬

lection of MSS. now at Poona.

This little work gives a list of numerous Tirthas with brief indica¬

tions of their special features and position, arranged in the topographi¬

cal order of Parganas. It is useful enough as a comprehensive synopsis

of such sacred sites as were known at the time to local worship. The

references to many obscure little shrines, Nagas, etc., show that the

enquiries of Pandit Saliibram’s assistants had been extensive. But the

work proves at the same time how little help traditional learning in Kasmir

could offer in our days for the serious study of the old topography of the

Valley.

Pandit Sahibram’s plan is to indicate each Tirtha’s position by

mentioning the territorial division in which it is situated, as well as the

nearest village or other well-known locality. It was undoubtedly the

learned author’s desire to give all local names in their old Sanskrit forms

as far as they were known to him. Accordingly we find a number of

localities correctly mentioned by their genuine old designations. But

unfortunately the number of the latter is truly insignificant when

compared with those local names which are plainly recognizable as new

fabrications, as worthless as those already mentioned in connections

with the topography of the modern Mahatmyas.

In consideration of the fact that P. Sahibram deserves to be looked

upon as the best representative of modern Kasmirian scholarship,2 it is

only just to illustrate the above remarks by a few examples. I take

them only from among those local names the genuine forms of which

can be easily ascertained from the Bajatarangini. The lake of the

Naga Susravas ,3 the present Susram Nag, is named Susramanaga in one

1 The papers acquired by me refer to some of the north-eastern Parganas

and contain descriptions (in Sanskrit) of the various Nagas, Lingas, etc., the

miraculous stories relating to them, together with the devotional texts which are

supposed to be used at their worship. Quaint illustrations and maps accompany

the text. The whole forms a large-sized folio. The critical value of these records

is very slight.

2 See Prof. Buhler’s Report, pp. 4, 38.

S See Rajat. i. 267 note, and below, § 59.

J. I. 8

58

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

recension and Susumnandga ( / ) in the other. The old Parganas of

Holacla, Laulaha , Khuyasrama are turned on account of their modern

names Vular, Lolau, Khuy^hom, into the 1 Rastras ’ of Volara, Lalava ,

Khoyahama. Banahal, the old Bdnasdld ,l figures as Bhanusala ; Khruv,

the ancient Khaduvl ,2 * 4 * known correctly even to so late a text as the

Lokaprakasa, as Khrdva . The well-known Klionamusa (Khun^moh)

appears as Ksusnamosagrama (!) The name of the ancient village Jaya-

vana'6 which fares badly too, as we have seen, in the Mahatmyas, is

metamorphosed into Jlvana ; Bgnyil , the old PLiranyapura^ is with a

flight of historical fancy turned into a foundation of king Ranaditya (!).

Even the sacred Tirtha of Tulamulya (TuHmul) does not escape a

renaming as Stliulamula , though in this case the local Mahatmya, with

its Tulamula , keeps close enough to the old name.6 After this, village

names like JJskara , Ramasrama , Kicakasrama , as designations of the

old Huskapura, Rdmusa , Krtydsrcima can scarcely surprise us.6 The

number of districts, towns, villages, streams, lakes and other topogra¬

phical features (exclusive of Tirthas) mentioned by Pandit Sahibram

amounts to nearly three hundred. But scarcely two dozens of the

names given for them are in accord with our old authorities.

Pandit Sahibram was one of the few modern Kasmirian scholars

who have seriously occupied themselves with the Rajatarangini and the

later Chronicles. This is shown by the elaborate abstracts he had

prepared of these works.7 Hence the indifferent knowledge of ancient

topography as displayed in his Tirthasamgraha, must appear all the

more striking. Yet in reality it is easily enough accounted for.

What knowledge learned tradition in Kasmlr has retained of

ancient sites as distinct from Tirthas and the like, is confined to a

few prominent localities which, for one reason or the other, were of

special interest to the Pandits. Thus the capital Pravarapura-S'rinagara

with several of its quarters, Vijayesvara , Suyyapura , Vardliamula ,

Padmapura , and some other places of importance in the Valley have

continued to be known by their ancient names. This was probably

because these names never ceased to be employed in colophons of Sans¬

krit manuscripts, in horoscopes, and similar records. In the case of a

1 See note viii. 1665, and below, § 41.

3 See note viii. 733 ; also § 105 below.

8 Compare note vii. 607, and § 105 below.

4 See note i. 287, and § 104 below.

6 Compare note iv. 638.

' 6 See notes i. 168; ii. 55 ; i. 147.

7 These abstracts, called Rdjatarangimsaihgraha, were acquired by Prof. Buhler ;

see Nos. 176-8 of th'e Poona collection. It deserves to be noted that in them no

attempt whatever is made to explain points of topographical interest.

1899.]

LOCAL TRADITION.

59

few other localities again like Jayapura, Damodara’s Udar , CaJcradhara ,

there were well-known popular legends which plainly indicated their

identity with sites mentioned in the Ttajatarahgini. But for the great

mass of ancient places there were no special reasons of this kind to

assure a recollection of their old names. It is hence only natural that

all genuine knowledge of their identity and earlier history has gradu¬

ally disappeared from the Pandits’ tradition.

Nothing but systematic enquiry on the lines of modern historical

research could help towards a recovery of the knowledge thus lost.

But such an enquiry could not be expected either from P. Saliibram

or any other indigenous scholar uninfluenced by Western critical

methods.

35. Popular local tradition has fortunately in Kasmir proved far

more tenacious than the tradition of the

Popu.laiM.oeal tradi learned. I have often derived from it valuable

aid in my local search for particular sites. My

antiquarian tours have given me ample opportunity to convince myself

that when collected with caution and critically sifted, such local tradi¬

tions can safely be accepted as supplements to the topographical infor¬

mation of our written records. In illustration of this statement I may

refer to the evidence gathered from local tradition in reference to the

sites of LoJiara,1 Hastivanja ,2 Kramavarta ,3 Jayapura ,4 Skandabhavana ,5

etc.

In more than one instance it can be shown that local legends which

Kalhana heard, still cling unchanged to the same sites. As striking

examples may be mentioned here the legends concerning Damodara’s

Udar ,6 the burned city of King Nara ,7 the temple of Pravaresa .8

It cannot be doubted that this tenacity of local tradition in Kasmir

is due largely to the isolation secured for the country by its alpine

position. Nothing is more instructive in this respect than a comparison

with the territories of ancient Grandhara and Udyana, or with the Panjab

plains. These regions so rich in ancient Hindu sites are particularly

devoid of local traditions connected with them. This fact is easily

understood if we think of the many and great ethnic changes which

1 See Raj at. Note E (iv. 177), § 15.

2 See Rajat. note i. 302, and J. A. S. B., 1895, pp. 379 sq .

8 Compare Note D (iii. 227) ; J. A. S. B., 1895, pp. 381 sq ; also below, § 43.

4 See note iv. 506 sqq., and below, § 122.

& See Note K (vi. 137).

6 See note i. 156; below, § 119.

7 See note i. 202 ; below, § 108.

8 See note iii. 350; below, § 96.

60

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

have passed over the land. Kasmir, fortunately for antiquarian research,

throughout its known history has escaped such great convulsions and

the breaks of tradition usually connected with them.

The influence of the geographical position of Kasmir can be traced

here also in another direction. Mountainous surroundings and conse¬

quent isolation tend everywhere in alpine countries to develop and

foster conservative habits of life and thought. We find these habits

most strongly marked in the population of the valley, and may safely

ascribe to them a great share in the preservation of local traditions.

1899.] POSITION AND CONFIGURATION OF KAS'MIR VALLEY.

61

CHAPTER III.

GENERAL GEOGRAPHY.

Section I. — Position and configuration of KAAmIr Valley.

36. Nature itself when creating the great Valley of Kasmir and

its enclosing wall of mountains, seems to have

The name KaSmlra. agsured to this territory not only a distinct

geographical character but also a historical existence of marked indivi¬

duality. We see both these facts illustrated by the clearly defined

and constant use of the name which the territory has borne from the

earliest accessible period.

This name, Ka^mira in its original Sanskrit form, has been used as

the sole designation of the country throughout its known history. It

has uniformly been applied both by the inhabitants and by foreigners.

We can trace back its continued use through an unbroken chain of

documents for more than twenty-three centuries, while the name itself

undoubtedly is far more ancient. Yet notwithstanding this long history

the current form of the name down to the present day has changed but

slightly in the country itself and scarcely at all outside it.

The Sanskrit Kasmira still lives as Kasmir (in Persian spelling

Kashmir) all through India and wherever to the West the fame of the

Valley has spread. In the language of the inhabitants themselves the

name is now pronounced as Kasir.1 This form is the direct phonetic

derivative of Kasmir , with regular loss of the final vowel and assimila-

1 The adjective Kasur ‘ Kasmirian ’ corresponds to Skr. Kasmir a. The u of the

last syllable is probably due to the v of an intermediate form * Kasyira ; see below.

62

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

tion of m to the preceding sibilant. With reference to a phonetic rule,

prevalent through all Indo-Aryan Vernaculars, which favours the change

of medial Skr. m into i>,1 we are led to assume an intermediate Prakrit

form * Ka6vir[ a]. In support of this we may point to the striking

analogy of the Kasnrir local name Sang as which, as shown in my note on

Rajat. i. 100, goes back through an older recorded form S'vangas to

* Smangcisa , the Samahgasa of the Chronicle. It has already been

shown above that we have to recognize in this *Kasvira the original

Prakrit form which Ptolemy’s Kacr7r€ipa, Kacrneipia (pronounced Kaspira,

Kaspiria) are intended to transcribe.2

Linguistic science can furnish no clue to the origin of the name

Kasmlra , nor even analyze its formation.3

Etymologies of This fact, however, has not saved the name

from being subjected to various etymological

guesses which for curiosity’s sake may receive here a passing notice.

It must be held to the credit of Kasmirian Sanskrit authors that their

extant writings are wholly innocent of this display of etymological fancy.

No less illustrious a person than the Emperor Babar opens the list.

His suggestion was that the name may be derived from the hill-tribe

* Kas ’ living in the neighbourhood of Kasmir.4 We easily recognize

here the reference to the Khasas of the lower hills. Their name,

however, in its true form has, of course, no connection with Kasmir.

Another etymology, first traceable in the Haidar Malik’s Chronicle and

hence reproduced by other Muhammadan writers,5 derives the first part

of the name from ‘ Kasha p,* i.e., Ka^yapa, and the second either from

1 Compare Dr. Grierson’s remarks, Z. D. M. G., 1. p. 16.

2 See above, § 5.

S If the Unadisutra, 472, Kaser mut ca is to be applied to the word Kasmir a, the

latter would have to be dissolved into kas-m-ira according to the traditional gram¬

matical system.

4 See Memoirs of Baber, transl. by Leyden and Erskine, p. 313. A Persian MS. :

of the text adds that mir signifies mountain, Erskine, Introduction, p. xxvii., im¬

proves upon this etymology by extending it to Kashgar, the Casia regio and

Casii Montes of Ptolemy. Ritter, Erdkunde, ii. p. 1127, from whom I take this

reference, not unjustly queries why the learned editor should have stopped short of

the Caspium mare and other equally manifest affinities.

Babar’s conjecture figures still seriously in a note of the latest translation of

the Ain-i Akbari, ii. p. 381.

Regarding the name and habitation of the Khasas, compare Rajat. i. 317 note.

5 It was first introduced to the European reader by Tieffenthaler’s extract

from Haidar Malik’s Chronicle compare Description historique et gdographique de

V Inde, ed. Bernouilli (1786), i. p. 79 (also p. 89 as to source). Compare also Wilson,

Essay, p. 94, for a similar note from the Waqi‘dt-i Kashmir of Muhammad ‘Azirn ; here

j ** is a clerical error for j f;*3 v-aAS'.

1899.] POSITION AND CONFIGURATION OF KAS'MIR VALLEY.

63

Ks. mar , i.e.f Skr. matha 1 habitation,’ or a word mir, supposed to mean

* mountain.’ *

It was, perhaps, a belief that this whimsical etymology represented

some local tradition, which induced even so great a scholar as Burnouf

to risk the conjectural explanation of Kasmira as * Kasyapamira, i.e.,

4 the sea of Kasyapa.’ 8 There is neither linguistic nor any other

evidence to support this conjecture. It would hence scarcely have been

necessary to refer to it had it not on the authority of a great name

found its way also into numerous works of a more general character.1 2 3

37. Just as the name Kasmir has practically remained unchanged

through the course of so many centuries, so

Extent^ml^position aiso pag territorial extent of the country

which it designated. This has always been

confined to the great valley drained by the headwaters of the Vitasta

and to the inner slopes of the ring of mountains that surround it.

The natural limits of the territory here indicated are so sharply

marked that we have no difficulty in tracing them through all our

historical records, whether indigenous or foreign. Hiuen Tsiang,

Ou-k'ong and Alberunl’s accounts, as we have seen, show them clearly

enough. Kalhana’s and his successors’ Chronicles prove still more in

detail that the Kasmir of Kasmirian tradition never extended materi¬

ally beyond the summit-ridges of those great ranges which encircle and

protect the Valley.

A detailed description of the geographical position of Kasmir does

not come within the scope of this paper. Nor is it needed since there is

an abundant modern literature dealing with the various aspects of the

geography of the country. For an accurate and comprehensive account

I may refer to the corresponding portion of Mr. Drew’s work and to the

graphic chapter which Mr. Lawrence devotes to the description of the

Valley.4 It will, however, be useful to allude here briefly to some of

the characteristic features in the configuration of the country which

have an important bearing on its ancient topography.

Kasmir owes its historical unity and isolation to the same facts

which give to its geographical position a distinct and in some respects

1 The Ks. word mar < Skr. matha , is in common nse in the country as the

designation of Sarais, shelter-huts on passes, etc. Mir might have been connected by

Haidar Malik’s Pandit informants with the name of Mount Meru or with mira ,

meaning according to a Kosa parvataikadesa, see B. R., s. v.

2 Compare his note in Humboldt, L'Asie centrale , i. p. 92.

3 See, e.g., Lassen, Ind. Alt., i. p. 54 note ; McCrindle, Ancient India as des¬

cribed by Ptolemy, p. 108 ; V. de St. Martin, Mem. de V Acad, des Inscript ., Sav.

E’trang ., v., ii. p. 83; Kiepert, Alte Geographie , 1878, p. 36.

4* See F. Drew, The Jnmmoo and Kashmir Territories , 1875, Chapters viii.-x.;

W. Lawrence, The Valley of Kasmir, 1895, pp, 12-39.

64

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

almost unique character. We have here a fertile plain embedded among

high mountain ranges, a single valley large enough to form a kingdom

for itself and capable of supporting a highly developed civilization.

Its height above the sea, nowhere less than 5000 feet, and its peculiar

position assure to it a climate equally free from the heat of India and

the rigours of cold, peculiar to the higher mountain regions in the north

and east.

The form of the country has been justly likened to a great irregular

oval, consisting of a similarly shaped level vale in the centre and

a ring of mountains around it. The low and more or less flat part

of the country measures about 84 miles in length, from south-east to

north-west, while its width varies from 20 to 25 miles. The area com¬

prised in this part lias been estimated at 1800 or 1900 square miles.1

Around this great plain rise mountain ranges which enclose it in an

almost unbroken ring. Their summit lines are everywhere but for a

short distance at the southernmost point of the oval, more than 10,000

feet above the sea. For the greatest part they rise above 13,000 feet,

while the peaks crowning them tower up to altitudes close on 18,000

feet. Reckoned from the summit lines of these ranges, the length of

the irregular oval enclosed by them is 'about 116 miles, with a varying

width from 40 to 75 miles. The whole area within these mountain

boundaries may be estimated at about 3,900 square miles.

The slopes of the mountains descending towards the central plain

are drained by numerous rivers and streams all of which join the

Vitasta within the Kasmir plain. The side-valleys in which these

tributaries flow, add much ground to the cultivated area of the country

several of them being of considerable length and width. But even

the higher zones of the mountain-slopes where cultivation ceases, add

their share to the economical wealth of the country. They are clothed

with a belt of magnificent forests, and above this extend rich alpine

pastures, close up to the line of perpetual snow.

In the great mountain-chain which encircles the country, there is

but one narrow gap left, near to the north-west end of the Valley.

There the Vitasta after uniting the whole drainage of Kasmir flows

out by the gorge of Baramula (Varahamula) on its course towards the

sea. For a distance of nearly 200 miles further this course lies through

a very contracted valley which forms a sort of natural gate to Kasmir.

It is here that we find the old political frontier ot Kasmir extending

beyond the mountain-barriers already described. For about 50 miles

below the Varahamula gorge the narrow valley of the Vitasta was held

in Hindu times as an outlying frontier tract of Kasmir.8

1 Compare Drew, Jummoo , p, 162, for this and subsequent statements.

8 See below § 53.

1899.] POSITION AND CONFIGURATION OF KAS'MIR VALLEY.

. 65

38. The general configuration of the country here indicated in

its broadest outlines may be held to account

Legend of Satisaras. , . , . , , , , . ,

tor the ancient legend which represents

Kasmir to have been originally a lake. This legend is mentioned by

Kalhana in the Introduction of his Chronicle and is related at great

length in the Nllamata.1

According to this earliest traditional account the lake called

Satisaras , ‘the lake of Satl (Durga),’ occupied the place of Kasmir

from the beginning of the Kalpa. In the period of the seventh Manu

the demon Jalodbhava (‘ water-born ’) who resided in this lake, caused

great distress to all neighbouring countries by his devastations. The

Muni Kasyapa, the father of all Nagas, while engaged in a pilgrimage

to the Tirthas in the north of India, heard of the cause of this distress

from his son Nila, the king of the Kasmir Nagas. The sage thereupon

promised to punish the evil-doer and proceeded to the seat of Brahman

to implore his and the other gods’ help for the purpose. His prayer

was granted. The whole host of gods by Brahman’s command started

for Satisaras and took up their position on the lofty peaks of the

Naubandhana Tirtha above the lake Kramasaras (Konsar Nag). The

demon who was invincible in his own element, refused to come forth

from the lake. Visnu thereupon called upon his brother Balabhadra

to drain the lake. This he effected by piercing the mountains with his

weapon, the ploughshare. When the lake had become dry, Jalodbhava

was attacked by Visnu and after a fierce combat slain with the god’s

war- disc.

Kasyapa then settled the land of Kasmir which had thus been pro¬

duced. The gods took up their abodes in it as well as the Nagas, while

the various goddesses adorned the land in the shape of rivers. At first

men dwelt in it for six months only in the year. This was owing to a

curse of Kasyapa, who angered by the Nagas had condemned them to

dwell for the other six months together with the Pisacas. Accordingly

men left Kasmir for the six months of winter and returned annually in

Caitra when the Pisacas withdrew. Ultimately after four Yugas had

passed, the Brahman Candradeva through the NUanaga’s favour acquired

a number of rites which freed the country from the Pisacas and exces¬

sive cold. Henceforth Kasmir became inhabitable throughout the year.

The legend of the desiccation of the lake is alluded to also by

Hiuen Tsiang, though in another, Buddhistic form.2 Its main features

as related in the Nllamata, live to this day iu popular tradition. They

1 See Bdjat. i. 25-27 ; Nllamata, vv. 26-237. A detailed extract of the Nila-

mata’s story has been given by Prof. Buhler, Report, p. 39,

2 See Si-yu-Tci, transl. Beal, i. p. 149.

J. I. 9

60

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

are also reproduced in all Muhammadan abstracts of the Chronicle.1

From Haidar Malik’s Tarikh the legend became known to Dr. Bernier

who prefaces with it liis description of the ‘ Paradis terrestre des Indes.’ 2

It has since found its way into almost every European account of

Kasmir.

It is probable that this legend had much to do with drawing from

the first the attention of European travellers

Lacustrme^featuies certain physical facts apparently supporting

the belief that Kasmir was in comparatively

late geological times wholly or in great part occupied by a vast lake.

But few seem to have recognized so clearly as the late Mr. Drew, the true

relation between the legend and the above facts. I cannot put his view

which from a critical point of view appears to be self-evident, more

clearly than by quoting his words: “The traditions of the natives —

traditions that can be historically traced as having existed for ages — tend

in the same direction, [viz., of the Yale having been occupied by a lake,]

and these have usually been considered to corroborate the conclusions

drawn from the observed phenomena. Agreeing, as I do, with the con¬

clusion, I cannot count the traditions as perceptibly strengthening it ;

I have little doubt that they themselves originated in the same physical

evidence that later travellers have examined.”3

The geological observations upon which modern scientific enquirers

like Mr. Drew and Colonel Godwin Austin, have based their belief as

to the former existence of a great lake, are mainly concerned with the

undoubted ‘ lacustrine deposits ’ found in the so-called Udars or Karewa

plateaus to be noticed below. But it seems to me very doubtful whether

we can reasonably credit the early Kasmirians with a correct scientific

interpretation of such geological records. It appears far more probable

that the legend was suggested by an observation of the general form of

the valley and by a kind of natural inference from the historical changes

in the country’s hydrography.

We shall see below that great drainage operations took place at

various periods of the country’s history which extended the cultivated

ground and reduced the area covered by lakes and marshes. To any one,

however ignorant of geology, but acquainted with the latter fact, the

picture of a vast lake originally covering the whole Yalley might natur¬

ally suggest itself. It would be onough for him to stand on a hill-side

somewhere near the Yolur, to look down on the great lake and the

adjoining marshes, and to glance then beyond towards that narrow gorge

1 Compare, e.g., Ain-i Akb., ii. p. 380 ; Wilson, Essay , p. 93.

2 Seo Bernier, Travels in the Mogul Empire, ed, Constable, p. 393.

3 See Jummoo, p. 207.

67

1899.] POSITION AND CONFIGURATION OF KASAI III YALLEY.

of Baramula where the mountains scarcely seem to leave an opening.

It is necessary to bear in mind here the singular flights of Hindu imagin¬

ation as displayed in the Puranas, Mahatmyas and similar texts. Those

acquainted with them, will, I think, be ready to allow that the fact of

that remarkable gorge being the single exit for the drainage of the

country, might alone have sufficed as a starting-point for the legend.

In respect of the geological theory above referred to it may yet

be mentioned that in the opinion of a recent authority “ even the pre¬

sence of true lacustrine deposits does not prove that the whole of the

Kasmir lake basin was ever occupied by a lake.” 1 At the present day

true lacustrine deposits are still being formed in the hollows of the rock

basin represented by the lakes of the north-west portion of the Valley.

It is held probable “ that the conditions have been much the same as at

present, throughout the geological history of the Kasmir Valley,” only a

minor area of the latter having at various periods been occupied by lakes.

Whatever view may ultimately recommend itself to geologists, it is

certain that the lacustrine deposits of Kasmir, though of no remote date,

speaking by a geological standard, are far older than any monuments

of man that have yet been discovered.2 Mr. Drew was undoubtedly

right in denying the existence of lacustrine deposits round any known

ancient buildings or other works of man in the Valley.

39. None of the natural features of Kasmir geography have had

a more direct bearing on the history of the

^"afS^mouEd:aDis^ ^ country than the great mountain-barriers

that surround it. They may hence rightly

claim our first consideration.

The importance of the mountains as the country’s great protecting

wall has at all times been duly recognized both by the inhabitants and

foreign observers. Since an early time Kasmirians have been wont to

pride themselves on their country’s immunity from foreign invasion, a

feeling justified only by the strength of these natural defences. We

find it alluded to by Kalhana who speaks of Kasmir as unconquerable

by the force of soldiers and of the protection afforded by its mountain

walls.3 The feeling is very clearly reflected in all foreign records. We

have already seen what special notice is taken by Hiuen Tsiang and

Ou-k'ong of the mountains enclosing the kingdom and of the difficulty

of the passes leading through them.4 The statements of the early Arab

1 See Oldham’s Manual of Indian Geology (1893), quoted by Mr. Lawrence,

Valley , p. 50.

S See Drew, Jummoo, pp. 207 sq,

3 See Raj at. i. 31, 39.

* Compare above, §§ 9, 11.

68

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

geographers brief as they are, lay due stress on the inaccessible cha¬

racter of the mountains. Alberuni does the same and shows us besides

the anxious care taken in old days to maintain this natural strength of

the country by keeping strict watch over the passes.1

Even when Kasmir had suffered a partial conquest from the north

and had become Muhammadanized, the belief in the invincibility of

its bulwarks continued as strong as before. Thus Sharlfu-d-dln, the

historian of Timur, writing apparently from materials collected during

the great conqueror’s passage through the Panjab Kohistan (circ. a.d.

1397), says of Kasmir: “This country is protected naturally by its

mountains on every side, so that the inhabitants, without the trouble

of fortifying themselves, are safe from the attacks of enemies.” The

subsequent account of the routes into Kasmir and other exact details

suggest that the author of the Zafarnama had access to genuine

Kasmlrian information.2

40. It is this defensive character of the mountain ranges to which

we owe most of our detailed information

^uoimt^n routes.11 regarding their ancient topography. We have

already in connection with the accounts of

Alberuni and the Chinese pilgrims had occasion to note the system of

frontier watch-stations by which a careful guard was kept on the passes

leading through the mountains. These fortified posts and the passes they

guarded, play an important part in the narrative of Kalhana and his

successors. As most of the Chronicle’s references to Kasmir orography are

directly connected with these watch-stations it will be useful to promise

here a few general remarks regarding their character and purpose.3

The small forts which since ancient times closed all regularly used

passes leading into the Valley, are designated in the Chronicles by the

word dvara ‘ gate ’ or by the more specific terms dranga or dliakka.

Numerous passages show that they served at the same time the

purposes of defence, customs and police administration. They were

garrisoned by troops under special commanders, designated as drangesa

or drahgadhipa. The control over all these frontier stations and the

command of the ‘ Marches ’ generally was vested in Hindu times in one

high state officer, known by the title of dvarajpati , ‘ lord of the Gate,’

or equivalent terms.4

1 See above, § § 12, 14.

2 See the extract from Sharifu-d-din’s Zafarnama in Tdrikh-i-Rashidz. transl.

by N. Elias and E. D. Ross, p. 432 ; compare also Ritter, Asien, ii. pp. 1122, sq.

8 For detailed references regarding these stations see my notes, J, A. S . B.,

1895, pp. 382 sqq. ; Raj at. i. 122 ; iii. 227 (D).

4 Compare Rdjat, note v. 214.

69

1899.] POSITION AND CONFIGURATION OF KAS'MlR VALLEY.

The organization of the system was somewhat changed in Muham¬

madan times when the guarding of the several routes through the

mountains was entrusted to feudal chiefs known as Maliks (Skr. margesci) .l

These held hereditary charge of specific passes and enjoyed certain privi¬

leges in return for this duty. In other respects the system underwent

scarcely any change. The fortified posts with their small garrisons

survived on all important routes almost to our own days being known as

rahddrl in the official Persian.2

It may be noted that apart from their character as military

defences against foreign inroads the Drangas were also in another

respect true ‘gates’ to the country. Nobody was allowed to pass

outside them coming from the Valley without a special permit or pass.

The system thus provided an important check on unauthorized emi¬

gration which was withdrawn only after the last Kasmir famine

(1878). 3

In order to appreciate fully the importance of these frontier watch-

stations it should be remembered that the mountain regions immediately

outside Kasmir were almost in every direction held by turbulent hill-

tribes. To the hardy Dards ( Darod ) in the north and the restless

Khakhas ( Iihasa ) in the south and west the rich Kasmir with its weak

population has always appeared as a tempting prey. The last inroad of

plundering Khakhas occurred not more than half a century ago and will

not soon be forgotten.4 At the same time it is certain that the valour

of these hardy mountain clans on the confines of Kasmir has at all times

contributed greatly to the natural strength of the mountain defences.

Without this protective belt the latter themselves would scarcely have

remained so long proof against foreign invasion.

I A detailed and interesting account of the Maliks and the routes held by

them is given by Baron Hugel, Kaschmir, ii., pp, 167 sqq. ; i., p. 347.

® See J. A. 8. B , 1895, p. 385; also below, § 49, 52.

3 For an early reference to this system of passports at the Dvaras, see Jonar.

654. For a description of the cruel exactions often connected with ‘ Rahdari,’

compare Lawrence, Valley , p. 215. I have never been able to visit the sites of the

old watch-stations at the several passes without thinking of the scenes of human

suffering they must have witnessed for centuries.

* Compare Rajat. i. 317 note.

70

ANCIENT GEOGRAPHY OF KASHMIR. [Extra No. 2,

Section II. — The PIr Pantsal Range.

41. In order to understand correctly the data relating to the an-

, cient topography of the mountains around

Kasmir orography.

Kasmir, it is necessary to acquaint ourselves

with their actual configuration and character. In the following account

it will be possible only to indicate the most prominent features of this

mountain system, and those directly connected with the historical

data under discussion. For detailed information on Kasmir orography

a reference to the lucid and instructive account in Mr. Drew’s work

may be specially recommended.1

The mountain ring enclosing Kasmir is divided into three main

ranges. One of these, usually designated as the Pir Pantsal Range,

forms the boundary of the Kasmir Yalley to the south and southwest.

It may be considered to begin from the southernmost part of the

Valley where the Banahal Pass, 9200 feet above the sea, marks the

lowest depression in the chain of mountains. After running for about

35 miles from east to west the range turns to the north-northwest.

In this direction it continues for about fifty miles more, and after attain¬

ing its greatest elevation in the Tatakuti Peak {15,524 feet above the

sea), gradually descends towards the Valley of the Vitasta. All im¬

portant old routes towards the Panjab cross this great mountain barrier,

and this circumstance enables us to trace some interesting information

regarding its ancient topography.

The Banahal Pass at the eastern extremity of the range must owing

to its small elevation have always been a con¬

venient route of communication towards the

Upper Cinab Valley and the eastern of the

Panjab hill-states. It takes its modern name from a village at the south

foot of the pass which itself is mentioned in Kalhana’s Chronicle by the

name of Banahal!.2 3 The castle of Banasala was in Kalhana’s own time

the scene of a memorable siege (a.d. 1130) in which the pretender

Bhiksacara was captured and killed. Coming from the Cinab Valley

he had entered Visalata ,s the hill district immediately south of the

Banahal Pass with the view to an invasion of Kasmir. As his move-

E astern portion of

Pir Pantsal Range.

1 See Jummoo , pp. 192-206.

8 See Rdjat. viii. 1665 sqq. and note. Banahal is the direct phonetic derivative

of Skr. BanaGald, medial Skr. s being regularly changed into h in Kasmiri.

3 See Rdjat. viii. 177. The name of Visalata is probably preserved in that of

the river Bichldri. Visalata more than once served as a safe retreat for Kasmirian

refugees ; comp. Rdjat. viii. 177, 697, 1074.

1899.]

THE PXR PANTSAL RANGE.

71

ment fell in the commencement of the winter, he could not have selected

a more convenient route. The Banahal Pass is the only one across

the Pir Pantsal Range on which communication is never entirely

stopped by snowfall. Kalhana’s narrative shows that the political and

ethnographic frontier of Kasmir ran here as elsewhere on the water¬

shed of the range. For the castle of Banasala, though so near as to be

visible already from the top of the pass ( sanihata ), was already held by

a Khasa chief.1

Proceeding westwards from Banahal we come to a group of three

snowy peaks reaching above 15,000 feet. With their bold pyramidal

summits they form conspicuous objects in the panorama of the range

as seen from the Valley.2 Kasmir tradition locates on them the seats

from which Visnu, S'iva and Brahman, according to the legend already

related, fought Jalodbhava and desiccated the Satisaras. The western¬

most and highest of these peaks (15,523 feet) forms the famous

Naubandhana Tirtha. According to the legend related in the Nilamata

and other texts and connected with the Indian deluge story, Visnu in

his fish Avatara had bound to this peak the ship (nan) into which

Durga had converted herself to save the seeds of the beings from des¬

truction.3 At the foot of this peak and to the northwest of it, lies a

mountain lake over two miles long known now as Konsar Nag , the

Kramasaras or Kramasara of the Nilamata and Mahatmyas.4 * It is

supposed to mark a footstep ( 'Jcrama ) of Visnu, and is the proper object

of the Naubandhana pilgrimage.

About 8 miles straight to the west of this lake, the range is crossed

hJ a pass, over 14,000 feet high, known now by the name of Sidau or

Budil. It lies on a route which in an almost straight line connects

Srinagar with Akhnur and Sialkot in the Panjab plain. Running up

and down high ridges it is adapted only for foot traffic, but owing to

its shortness was formerly a favourite route with Kasmlrls.6 The name

Sidau is given to the pass from the first village reached by it on the

1 Rajat. viii. 1674, 1683. Savikata is the regular term for ‘pass.’

2 Marked on maps as ‘ Brama Sakai,’ perhaps a corruption for Brahmatsikhara

* Brahman’s peak.’

8 See Nilamata, 33 sqq . ; Haracar. iv. 27; S'riv. i. 474 sqq. ; S'arvdvatdra iii. 4, 12 ;

v. 43, etc.

4 See S'riv. i. 482 sqq. where a visit of Sultan Zainu-l-‘ahidln to this lake is

related at length ; Nilamata, 121, 1272 ; Naubandhanamdhdtmya, passim ; S'arvavatdra

iii. 10; v. 174, etc.

6 According to Drew, Jummoo, p. 524, the distance from Jammu to S'rinagar by

the Sidau route is reckoned at 129 miles while vid the Banahal it is 177 miles.

The name Budil is given to the pass from the hill-district adjoiuing it on the

south ; compare my note Rajat. vi. 318.

72

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Kasmir side. It is by this name, in its original form Sipdhapatha,

that the pass is mentioned in Kalhana’s Chronicle as the route chosen

for a pretender’s irruption in Sussala’s reign.1

A snowy peak close to the west of the pass of Siddhapatha marks

the point where the main range changes its direction towards north-

northwest. From the same point there branches oft in a westerly

direction the lower Ratan Pir Range to which we shall have to refer

below. Beyond this lie the passes of Rupri and Darhal, both above

13,000 feet in height. They are not distinctly named in the Chronicles.

But as they give most direct access to Rajauri, the ancient Rajapuri , and

are crossed without much trouble during the summer months they are

likely to have been used from an early time. Near the Darhal Pass

lies the Nandan Sar , one of the numerous tarns which along this portion

of the chain mark the rock-ground beds of old glaciers. It is probably

the Nandana Naga of the Nilamata.

42. About five miles due north of the Nandan Sar we reach the

_ , _ lowest dip in the central part of the whole

Pir Pantsal Route. . , , . . .

range. It is marked by the pass known as

Pir Pantsal , 11,400 feet high. The route which crosses it has from

early days to the present time been the most frequented line of com¬

munication from Kasmir to the central part of the Panjab. The

frequent references which the Chronicles make to this route, permit us

to follow it with accuracy from the point where it enters the mountains.

This is in the valley of the Rembyar^ River (Ramanyafavi) , a little

below the village of HurVpur.

This place, the ancient S'uuapura, is often referred to as the entrance

station for those reaching Kasmir from Rajapuri and the neighbouring

places, or vice versa as the point of departure for those travelling in the

opposite direction.2 S'urapura was founded by S'ura, the minister of

Avantivarman, in the 9th century evidently with the intention of estab¬

lishing a convenient emporium on this important trade-route.3 He

transferred to this locality the watch-station ( drahga ) of the pass. Its

site, as I have shown in my Notes on the Ancient Topography of the Pir

Pantsal Route* can still be traced at the place known as Ildhi Darwaza

(‘ the gate of God ’), a short distance above the village. We find the

1 See Rdjat. viii. 557. In the Chronicles of S'rivara and his successors the tract

about Sidau is repeatedly referred to as Siddhddesa, an evident adaptation of the

Ks. form of the name.

2 See Rdjat. iii. 227, Note D, § 1,

3 Compare Rajat. v. 39 note.

* See J. A. S. B., 1895, p. 385. This paper should be compared for all details

regarding the other sites along this route.

1899. J

THE PIR PANTSAL RANGE.

73

Hastivafija.

commanders of this frontier-station more than once engaged in military

operations against intending invaders from the other side of the moun¬

tains.

Ascending the valley of the Rembyar?' or Ramanyatavi for about

7 miles we reach the point where the streams coming from the Pir

Pantsal and Rupri Passes unite. In the angle formed by them rises

a steep rocky hillock which bears on its top a small ruined fort

known as Kamelankoth. These ruins probably go back only to the

time of 1 Ata Muhammad Khan,’ the Afghan Governor of Kasmir,

who, about 1812, fortified the Pir Pantsal Route against the Sikh

invasion then threatening. But I have proved in the above-quoted

paper that they mark the original position of the ancient watch-station

on this route before its transfer to S'urapura.1 Kalhana, iii. 227, calls

this site Kramavarta. This name is rendered by a glossator of the 17th

century as Kamelanakotta and still survives in the present Kamelankoth

(* Kramavartanam kofta).

43. The old ‘ Imperial Road’ constructed in early Mughal times

then ascends the narrow valley, keeping on

its left side high above the Pir Pantsal stream.

At a distance of about four miles above Kamelankoth and close to the

Mughal Sarai of ‘ Allabad, a high mountain- ridge slopes down from

the south and falls off towards the valley in a wall of precipitous cliffs.

The ridge is known as Hqst'vafij. This name and the surviving local

tradition makes it quite certain that we have here the spot at which a

curious legend told by Kalhana was localized from early times.2

The Chronicle, i. 302 sqq. relates of King Mihirakula whose identity

with the White Hun ruler of that name (circ. 515-550 a.d.) is not

doubtful, that when on his return from a tour of conquest through India

he reached the ‘ Gate of Kasmir,’ he heard the death-cry of an elephant

which had fallen over the precipice. The gruesome sound so delighted

the cruel king that he had a hundred more elephants rolled down at the

same spot. The old glossator on the passage informs us that “ since

that occurrence the route by which Mihirakula returned, is called

TIastivahja .” The Persian Chroniclers too in reproducing the anecdote

give Hastivanj as the name of the locality.

The local tradition of the neighbouring hill tracts still knows the

story of a king’s elephants having fallen down here into the gorge

below. It also maintains that the old route to the Pass, in the times

before the construction of the ‘ Imperial Road ’, crossed the Hast'vanj

ridge and followed throughout the right bank of the Pir Panigal

1 J. A. S. B.y 1895, pp. 384 sq.

* Compare J. A. S. B„ 1895, pp. 378 sqq.

J. i. 10

74

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

stream. This is fully borne out by a statement of Abii-l-Fazl.1 De¬

scribing the several routes available on the march from Bhimbhar

to Kasmir, he clearly distinguishes “ the route of Hastivanj (MSS.

Hastivatar) which was the former route for the march of troops,” from

the ‘ Pir Pantsal route ’ which Akbar used on his visits to Kasmir.

The name HasPvanj contains in its first part undoubtedly hast1, the

Ks. derivative of Skr. hastin, ‘ elephant.’ The second part is connected

by the Persian compilators with the root vanj meaning ‘to go’ in Wes¬

tern Panjabi. The close connection between the name and the local

legend already heard by Kalb ana is evident enough. But whether the

latter had any foundation in fact or merely arose from some ‘ popular

etymology ’ of the name, cannot be decided.

The story helps in any case to make it quite clear that the ancient

route from the Pir Pantsal Pass kept to the right or southern side of

the valley. My enquiries on the spot showed that this route though

neglected for many centuries is passable for laden animals and not un-

frequently used by smugglers.2

44. ‘Aliabad Sarai is a Mughal hosjrice erected for the shelter of

travellers about half a mile above Hast1 vanj.

It is about the highest point on the ascent to

the p>ass where fuel can conveniently be obtained. I think it hence

probable that the Matha or hospice which Ksemendra mentions on

the Pir Pantsal Pass, must have been situated somewhere in this

neighbourhood.

Pafiealadharamatha.

1 See Aln-i Alcb ., ii. pp. 347 sq . The form Hastivatar in the text is a clerical

error for Hastivanj, easily explained in Persian characters.

2 Dr. Bernier who in the summer of 1665 accompanied Aurangzeb’s court to

Kasmir, has left us, in his Ninth Letter to M. de Merveilles, an accurate and graphic

account of the Pir Pantsal Route. While ascending the Pass from the Pan jab side

he passed the spot where two days earlier an accident had happened curiously

resembling Mihirakula’s story. Fifteen of the elephants carrying ladies of the

Imperial seraglio, owing to some confusion in the line of march, fell over the

precipice and were lost ; see Bernier’s Travels, ed. Constable, p. 407. The curious

Map of Kasmir given in the Amsterdam edition of 1672 shows accordingly the

‘ Pire Penjale ’ mountain with a troop of elephants rolling in picturesque confusion

over its side.

Former editions of Ince’s ‘ Hand-book ’ placed the scene of this accident at a

spot called Lai Ghulam just opposite Hastivanj on the ‘ Imperial Road.’ It is evident

that this wrong location was due to the original compiler having somehow confused

Bernier’s account and the local tradition referring to Hastivanj. The edition of

1888, p. 64, rectifies this mistake, but still indicates Lai Ghulam as the site “ of many

a dreadful accident” before the causeway of the ‘ Imperial Road ’ was made. Asa

matter of fact, the left side of the valley was not used at all as a route before the

construction of the ‘ Imperial Road ’ along its cliffs.

Of the accident on Aurangzeb’s march no recollection survives.

1899.]

THE PIR PANTSAL RANGE.

75

Ksemendra makes this interesting reference in that curious portion

of the Samayamatrka already alluded to, which describes the wander¬

ings of the courtezan Kaiikali.1 * The heroine of his story after effect¬

ing some petty thefts in Kasmir proceeds to S'urapura. There she

passes herself off as the wife of a load-carrier ( bharika ) engaged on the

‘salt road.’8 By this term the Pir Pantsal route is quite correctly

designated. It has remained to the present day the chief route by which

the produce of the Panjab salt-mines coming via Jehlam and Bhimbhar

enters Kasmir.3 4 * * * She keeps up the disguise which is evidently intended

to help her through the clutches of the officials at the frontier watch-

station, by taking next morning a load on her head and starting

with it towards the pass ( samhata ). On the way she passes along high

mountains by precipitous paths deeply covered with snow. By night¬

fall she reaches the PaNcaladharamatha after having in the meantime

assumed the guise of a respectable housewife and apparently disposed

of her load. It being late in the season, she passes the night there

shivering with cold. Thence she finds her way open to India where a

career of successful adventures awaits her.

45- Ksemendra’s itinerary is of particular value because it sup¬

plies us with the only mention of the old

The name Pancala. e , , T , T, • , •

name or the pass I can trace. It is certain

that with him Pancaladhara designates the highest portion of the route,

i.e ., the Pass of the Pir Pantsal. It is equally obvious that Pancala

is the original of the modern Ks. Pantsal which is in fact identical

with the earlier form except for the regular change of Skr. c into Ks.

ts. In the Pahari dialect of the population inhabiting the valleys to

the south the name is still pronounced Pancala

1 See Samayam, ii. 90 sqq., and above, § 25.

8 Professional load-carriers or Coolies are found to this day in numbers in

Hiir^por, Pasiana and other places near the Pir Pantsal Pass. Of Zainu-l-‘abidin it

is specially reported that he settled a colony of load-carriers from Abhisdra (i.e., the

country about Bhimbhar) at the customs-station of S'urapura ; see S'riv. i. 408.

Coolies are the only means of transport on the Pir Pantsal and other passes when

the snow lies to any depth.

3 Salt is a considerable article of import into Kasmir where it is wholly wanting j

see Lawrence, Valley, p. 393. I remember vividly the long strings of salt-laden

bullocks which I used to meet daily when marching into Kasmir by the Pir Pantgal

route.

4 I am not certain of the origin of the pronunciation of the name as Pir Panjal

now accepted by Anglo-Indian usage. It is known neither on the Kasmir nor on

the Panjab side of the range itself. It meets us first in Bernier’s ‘ Pire Penjale.’

Tieffenthaler, however writes more correctly Pensal ; see Description de Vlnd 1786,

pp. 87 sq.

76

ANCIENT GEOGRAPHY OF KASHMIR.

[Extra ]tfo. 2,

The term dhdrd which is added to Pahcala, represents in all pro¬

bability the equivalent of our ‘ pass.’ Skr. dhdrd means generally the

sharp edge of some object. According to Wilson’s Dictionary, as quoted

by Bothlingk-Roth, the word also carries the specific meaning of ‘ edge

of a mountain.’ It is probable that this meaning was taken by Wilson’s

Pandits from some Kosa. In any case it agrees closely with the

use of the word dhar in the modern Pahari dialects south of Kasmir.

There it is well-known as the designation of any high mountain ridge

above the region of alpine pasture.

We are tempted to see in Pancdla a distinct local name, either of the

Pass itself or of the whole mountain chain. But the use of the modern

derivative Pantsal presents difficulties in the way of a certain conclu¬

sion. The word Pantsal is applied in Kasmir chiefly to the great

mountain chain which forms the boundary of the country to the south,

i.e., the range to which conventional European usage gives the name of

‘ Pir Pantsal.’ Yet the meaning now conveyed to a Kasmirl by the

term Pantsal, is scarcely more than that of ‘ high mountain range.’

The word is used in combination with specific names for the desig¬

nation of subordinate branches of the great range towards the Panjab.

Thus the range crossed on the way from the Pir Pantsal Pass to Rajauri,

is known as ‘ Eatan Pantsal ,’ and the one crossed by the Haji Pir Pass

between Uri and Prunts (Punch) as ‘ Haji Pantsal .’ Sometimes, but

not so generally, the term is applied also to mountains wholly uncon¬

nected with the Pir Pantsal system.

On the whole I am inclined to believe that Pancdla > Pantsal had

originally the character of a specific local name. It may have been

applied either to the whole of the great southern chain of mountains or

its central portion about the Pir Pantsal Pass. Subsequent usage may

then have extended the application of the term just as it has that of

the name ‘Alps’ in Europe. Our materials, however, are not sufficient

to enable us to trace the history of the word with certainty. 1

46. In this connection it will be useful briefly to notice also the

word Pir which forms the first part of the

Pir, a term for pass. modern designation of the Pass. This word

is now used more or less frequently for ‘ Pass’ both in Kasmir and the

hill-tracts south of it. Mr. Drew who seems to have given more

attention to local nomenclature in these hills than other travellers,

in his explanation of the term starts from the well-known meaning

of Pir in Persian, an 1 old man ’ and thence a ‘ saint or Eaqir.’ 2

1 The main facts regarding the modern use of the word Panifial have been quite

correctly recognized already by Drew, Jummoo , p. 157.

2 See Jummoo, p. 157 note.

1899.]

THE FIR PANTSAL RANGE.

77

He refers to the common practice of Faqirs establishing themselves

on Passes for the sake of refreshing travellers and of receiving their alms.

“ When any noted holy Faqir died on a Pass, the place became sacred to

his memory, and was often called after him, his title of Pir being prefix¬

ed ; at last it became so common for every important Pass to have a

name beginning with Pir that the word acquired the secondary meaning

of Mountain Pass.” Mr. Drew refers to the fact that Dr. Bernier already

found an aged hermit established on the Pass who had resided there

since the time of Jahangir. He was supposed “ to work miracles, cause

strange thunders, and raise storms of wind, hail, snow and rain.” From

this ‘ Pir,’ Mr. Drew thinks, the Pass acquired the first part of its

present name.

I agree with the above explanation as far as the use of the

Persian word Pir is concerned. But I suspect that the custom of

connecting mountain passes with holy personages rests on a far older

foundation. Superstitious belief has at all times and in all mountainous

regions peopled the solitary summits and high ridges with spirits and

other supernatural beings. To this day Kasmirian Brahmans fully

believe in the presence of Devatas and ‘ Bhutas’ of all sorts on high

mountain passes. In those parts of the Himalaya where Hinduism has

survived among all classes, this superstition can, no doubt, be found

still more fully developed.

On all Kasmir Passes, however rarely visited, stone-heaps are found

marking the supposed graves of imaginary ‘ Pirs.’ Every pious Muham¬

madan on passing adds his stone to them. Yet these little cairns existed

there in all probability long before Islam reached the country. Exactly

the same custom is observed, e.g., by the Hindu Pilgrims to Amaranatha

on crossing the Vavajan Pass above the lake of Susravonaga, ‘to

please the Devas ’ as the Maliatmya says.1

We can show that almost all famous Ziarats in Kasmir, whether of

real or imaginary Muhammadan saints, occupy sites which were sacred

in earlier times to one or the other Hindu divinity. We can scarcely

go far wrong in concluding by their analogy that the ‘ Pirs ’ of the

Muhammadan wayfarers have only taken the place of the older Hindu

‘ Devas.’

This surmise is strikingly corroborated by the only passage of the

1 See Amarandthamahdtmya, vii. 1 sqq. The stones placed are supposed to

represent mathikds, 1 shelter-huts in which the gods can find refuge from the evil

wind blowing on the pass (hence its alleged Sanskrit name Vdyuvarjana). The duty

of making these Mathikas is enjoined in vii. 19. Hathikam ye na kurvanti tatraiva

Vayuvarjanc I ddrunam narakam ydnti satakalpam na samsayah II krtvd tu mathikdrh

devi pujayed vidhipdrvakam I arpayed devaprityartharh daksindbhih samanvitam II.

78

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Pass of Pir Pantsal.

Sanskrit Chronicles which mentions the Pir Pantsal Pass by its proper

name. Sri vara iii. 433, when relating the return of a Kasmir refugee ‘by

the route of S'urapura ’ in the time of Hasan Shah (circ. a.d. 1472-84),

tells us of a fatal chill he caught “ on the top of the Pancaladeva .”

It is clear that the name here used corresponds exactly to the modern

Pir Pantsal, ‘Pir’ being the nearest Muhammadan equivalent for ‘Deva.’

Dr. Bernier’s account has already shown us that popular superstition had

not failed to transfer also the supernatural powers of the ‘ Deva’ to the

Pir who acted as his representative on the Pass.

47. We may now return to the description of the old route where

we left it at ‘Aliabad Sarai and resume our

journey towards the Pass. From the Mughal

hospice the road ascends in a gently sloping valley westwards

until at at a distance of about 4 \ miles the height of the Pass is

reached. Close to the point where the descent towards the Panjab

begins, stands the hut of a Faqir. He has inherited the post of

Bernier’s Pir, but little of his spiritual powers and his emoluments.

An octagonal watch-tower close by, occupied by a Sepoy post till a few

years ago, may mark the site of an earlier outpost.

The descent is here as on all Passes of the range far steeper on the

Panjab side than towards Kasmir, Pusidnct , the next stage, which is

reached by zigzag paths along the rocky slope of the mountain, lies

already more than 3000 feet below the Pass. The little village is an

ancient place. It is undoubtedly the Pusyananada of Kalhana who

mentions it repeatedly in connection with the civil wars of his own

time.1 Pusyananada served often as a refuge for rebel leaders for whom

Kasmir had become too hot. They could thence conveniently resume

their inroads. We see here again clearly that the Kasmir frontier ran

on the watershed of the range ; for of Pusyananada it is distinctly said

that it belonged already to the territory of Rajapuri.

From Pusiana the road descends in a westerly direction along the

bed of a stream which belongs to the headwaters of the Tausi (Tohi)

of Prunts. The next stage is the hill- village of Bahramgala, a consi¬

derable place which is mentioned already by S'rivara under the name

of Bhairavagala.2 From Bahramgala the route turns to the south and

crosses, by the Pass known as Ratan Pir (8200 feet), the range which

has already been mentioned as a branch from the Pir Pantsal chain.

There the route enters tbe region of the middle mountains and descends in

an open valley to Rajauri, the ancient Rajapuri, where we may leave it.

1 Compared Bdjat. viii. 959 note. The ending ndda is identical with ndla ,

Anglo-Indice ‘ Nullah,’ i.e., ‘ valley, ravine.’

2 See Srlv. iv. 529, 589.

1899.]

THE PlR PANTSAL RANGE.

79

48. Beyond the Pir Pantsal Pass the summit-line of the main

range rises again considerably. The Tang -

Centi_al^pait^of Pir ^ a pags which is about five miles due north

of the Pir Pantsal Pass and is mentioned by

Abu-1-Fazl, 1 is already far higher. The track crossing it is scarcely

practicable for animals.

The same is the case, as personal experience showed me, with

the next two Passes, known by the Pahari names of Gittapani and

Coti Gall ; they are both over 14,000 feet high. The first one was

probably used on occasion of the inroad related by Shi vara, iv. 589

sqq. We are told there of a rebel force which coming from Rajauri

evaded the troops of Sultan Muhammad Shah posted at Surapura, by

crossing the mountains in the direction of Kacagala. This place, as

shown on the map, corresponds undoubtedly to the alpine plateau or

* Marg ’ of Kactgul on the northern slope of the Pir Pantsal range.

A short distance to the northwest of the Coti Gall Pass the range

culminates in its greatest snowy peak, Mount Tatakuti , which rises to a

height of 15,524 feet. Owing to its bold shape and central position

this peak is the most conspicuous object in the panorama of the whole

range, whether seen from the Kasmir Valley or from the Panjab

plains. To the north it presents a precipitous face of unscaleable rocks.

On the south it is surrounded by snowfields which on the occasion of an

ascent made late in the season I found still of considerable extent.

We have already seen that it is this peak which Alberuni describes

under the name of Kularjak .2 For an observer from the Panjab

plain about Gujrat the appearance of the peak, with its glittering dome

of snow, is very striking, notwithstanding the great distance (about

87 miles as the crow flies). I have sighted it on very clear days even

from Lahore Minars.

From Tatakuti the chain continues at a great elevation for a con¬

siderable distance, the summit ridge keeping an average height between

14,000 and 15,000 feet. We find it crossed first by the Passes of Sangsa -

fed , Nicrpur and Gorgall , all difficult routes leading down into the valley

of Loh^rin, the ancient Lohara. It is only at the Tos^maidan Pass

that we meet again with an important and ancient line of communica¬

tion.

49. This Pass being on the most direct route between the Kasmir

capital and Lohara, was of special importance

Tos^ maidan Routs. , . , , . £ , , r/ ,

during the reigns of the later Kasminan

kings whose original home and safest stronghold was in Lohara. We

1 See Ain-i Ahb., ii. p. 348.

8 Compare above, § 14.

80

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

fiad accordingly the route leading over the Tos^maidan Pass often

referred to in the last two Books of Kalhana’s Chronicle.

But apart from this historical connection the Tos^maidan route must

have always been prominent among the old lines of communication from

Kasmlr owing to its natural advantages. It was the shortest route into

the Valley of Punch (Parnotsa) and hence to that portion of the western

Pan jab which lies between the Jehlam and Indus. It was besides

under the old conditions of road and travel probably the easiest and

safest route in that direction.1

This old route started from the present village of Drang, situated at

the foot of the mountains in the Biru Pargana, circ. 33° 57' lat. 74° 36'

long. The name of the village is, of course, nothing but the old term

of dranga , ‘ watch-station.’ In old times the place was distinguished

as Karkotadranga.2 It may have received the distinctive first part of

its name, Karkota, from the mountain-ridge now known as Kakodar ,

which is passed higher up on the route. Ks. Kakodar could well be

derived form an earlier Skr. form like * Kdrkotadhara. The Tirtlia-

samgraha also mentions a Kdrkotandga somewhere in this direction.

From Drang where a customs-station exists to this day, the road

ascends over an easy forest-clad slope to the edge of the Tos^maidan.

This is, as the name indicates, a large upland plateau of undulating

grazing grounds, rising very gradually from a level of about 10,000 feet.

At the point where the road strikes the northern edge of the plateau,

there are several ruined towers. They seem to have been last repaired

on occasion of the Sikh invasion of 1814 to be referred to below, but are

probably far older. The spot is known to this day as Barbal which

in Ks. means ‘the place of the Gate’ (Ks. bar < Skr. dvara). In view

of this designation and the commanding position of the place we can

safely locate here the proper Dvara or ‘ Gate ’ of this route.3

The route after crossing the Tos^maidan plateau ascends over

gently sloping grassy ridges to the Kakddar spur and passing along the

south foot of the latter reaches the Pass. The ascent is so gradual and

easy that though the elevation of the latter is over 13,000 feet, the con¬

struction of a cart-road would so far meet with little difficulty. The

Pass itself is equally easy.

On its west side two routes are available. One descends in the

1 The historical references to this route will be found collected in Note E

{Raj at. iv. 177) on Lohara, §§ 5-14.

2 Compare Rdjat. vii. 140 ; viii. 1596 notes.

3 The term dvara is actually used by Kalhana, vii. 140, 1301, for a fortified

post on this route. The village Drang is a suitable enough position for a customs

and police station ; the point for military defence, however, is higher np at ‘ Barbal/

1899.]

THE PIR PANTSAL RANGE.

81

GagrI Valley past the village of Ghdmbar mentioned in the Rajatararigini

by the name of S'arambara.1 The other leads over a cross-spur in a

south-westerly direction straight down into the valley now known as

LofHrin. The position of the ancient castle of Lohara which I was

able to trace in the centre of this great and fertile valley, has been fully

discussed by me in a separate note.2 About 8 miles further down the

valley and at the point where its waters meet the stream coming from

GagrI, lies the large village of Mandi. It marks the site of the old

‘ market of Attalika,’ repeatedly mentioned by Kalhana.3 From Mandi

onwards the route passes into the open valley of the Tohi (Tausi) of

Prunts which offers an easy line of communication down to the plains.

The historical importance of the T5s?maidan route is best illustra¬

ted by the fact that it was chosen on two occasions for expeditions

aiming at the invasion of Kasmir. We have already referred to Mah¬

mud of Ghazna’s expedition, probably of a.d. 1021, which Alberuni

accompanied, and to which we owe the valuable information recorded

by him.4 This attempt at invasion* perhaps the most serious of

which we know during Hindu times, was frustrated by the valorous

defence of the castle of Lohara and a timely fall of snow. Nor was

Maharaja Ranjit Singh more successful when in 1814 he first attempted

to invade Kasmir by this route.6 The portion of the Sikh army led by

him in person, safely reached the Tos^maidan plateau where the Afghan

defenders were posted near the towers above mentioned. Difficulties of

supplies, however, and the news of a reverse sustained by the column

marching by the Pir Pantsal route forced on a retreat. This ended in

a complete rout in the mountain defiles about LolHrin.

It may yet be mentioned that the route over the Tos?maidan was

already in all probability followed by Hiuen Tsiang on his way to

Parnotsa or Prunts.6 It remained a favourite trade route until the recent

• "

Jehlam Valley cart-road was constructed. Owing to the elevation of

the Pass, however, this route is always closed by snow longer than, e.g.,

that of the Pir Pantsal. During the winter, therefore, the road from

Lohara to Kasmir lay by the lower passes in the west leading into the

Vitasta Valley below Baramula.7

1 See Rajat. viii. 1875-77 note.

2 See Note E, iv. 177 ; also Ind. Ant . 1897, pp. 225 sqq.

3 See Rajat . viii. 581 note.

4 See above, § 14.

3 For a more detailed account of this expedition, see Note E, iv. 177, § 14.

6 Compare above, § 9.

1 See Note E (Rajat. iv. 177), §§ 7, 8, for Kalhana’s references to the ocoasions

when this more circuitous route was used.

J. I. II

ANCIENT GEOGRAPHY OF KASHMIR.. [Extra No. 2,

12

Not far to the north of the Tos^maidan Pass the range still attains

a height of over 15,000 feet in a group of bold snowy peaks. Its

summit ridge then gradually descends and is crossed by some lower

passes from the neighbourhood of the well-known alpine plateau of

Gulmarg. From a summit behind Gulmarg (marked ‘ Sallar ’ on the

map) several spurs radiate. They form the northern end of the range,

and descend very steeply and with faces of rugged cliffs into the narrow

valley of the Vitasta.

Section III — The Vitasta Valley.

50. We have already spoken of the Vitasta Valley as the single

outlet for the waters of Kasmir and as the

Valley ot Vitasta. great gate of the country. We may now cast

a glance at the old route leading through it and at the defences by

which nature has fortified it.

The Vitasta Valley below Baramiila is confined between two ranges

of mountains. The one to the south is a branch of the Pir Pantsal

Range separating from the main chain at a point behind Gulmarg. The

range to the north belongs to a mountain-system which culminates in

the Kajnag Peak (14,400 feet) and is usually designated by the name

of the latter. These two ranges accompany the course of the river for

some eighty miles westwards down to the point near Muzaffarabad

where the Vitasta makes its sudden bend to the south.

Along the whole length of the Valley, cross-ridges, more or less

steep and rugged, run from both sides down to the river-bed. This

consists from below Baramiila of an almost unbroken succession of

rapids, the fall in level being nearly 3000 feet in the above distance.

The Valley is throughout narrow and wanting in level ground. But

for about 50 miles, down to the old Kasmir frontier line, it may more

fitly be described as a narrow ravine. Only occasional alluvial terraces

high above the river afford room here for settlement and cultivation.

Owing to this extremely confined nature of the Valley, communica¬

tion on the route leading along it must have always been troublesome

and risky in old times. The natural difficulties of this long defile were

no doubt considerably increased by the restless disposition of the Khasa

tribe which has held it since ancient times. The Sikhs who were the

last to fight their way through these passes, suffered more than one

disaster at the hand of the hill-men. The line of forts erected by

them along the valley attests to this day the trouble they experienced

1899.]

THE V1TASTA VALLEY.

83

in holding the passage.1 The military difficulties of a march through

such a succession of dangerous defiles must have been even greater in

old times which knew no fire-arms. The protection of the route against

an active enemy who could easily seize and hold all commanding posi¬

tions, was then, no doubt, a still more difficult task.

51. It is probably on account of the circumstances here briefly

T,., __ _ , indicated that we hear in the Chronicles com-

Vitasta Valley Route. of fte rQute following the

Vitasta. Being the shortest line of communication to the present

Hazara District and the Indus, it was certainly used from early times.

We have seen that Hiuen Tsiang and Ou-k'ong coming from the

ancient Gandhara and Urasa followed it on their way to Kasmir, and

that it was well-known to Alberunl.

But it seems probable that its importance, military and commerciab

was then far smaller than that of the Plr Pantsal and Tos^maidan

routes. It is only in modern times that this western route has attained

real prominence. This originated in the time of the Afghan rule over

Kasmir when the route along the Vitasta to Muzaffarabad and hence

though Hazara afforded the shortest and least exposed line of com¬

munication between Kasmir and Peshawar.2 Subsequently after the

annexation of the Pan jab, the establishment of the hill- station of

Murree naturally drew traffic in this direction. The construction of the

Tonga Road from Murree to Baramula in our own time finally assured

to this route its present supremacy.

There is at present a road on each side of the Valley leading down

to Muzaffarabad. But only the route along the right bank of the river

can claim any antiquity. The one on the opposite bank has come into

general use only within the last few decades since traffic towards

Murree and Rawalpindi sprung up. The track chosen for the old road

is easily accounted for by topographical facts. We have already noticed

that the Vitasta Valley route was of importance chiefly as leading to

Hazara (Urasa) and hence to the old Gandhara. A glance at the map

will show that the open central portion of Hazara is most easily gained by

crossing the Kisanganga just above Muzaffarabad and then passing the

comparatively low ridge which separates this river from the Kunhar

stream. The route here indicated finds its natural continuation towards

1 Moorcroft’s account of his attempt :o use the Muzaffarabad route in 1823

gives a graphic picture of the obstacles created by the rapacious hill-tribes ; see

Travels , ii. pp. 281 sqq. Compare also Lawrence, Valley , p. 200.

2 Baron Hugel quite correctly notes a Kasmir tradition that the Baramula

route was properly opened up only about 80 years before his own visit (1835) on

the arrival of the Pathans; see Kaschmir , ii. p. 174.

84

ANCIENT GEOGRAPHY OE KAS'MIR.

[Extra Vo. 2,

Kasmlr on the right bank of the Vitasta, the crossing of the latter being

wholly avoided. It has already been shown above that this route, now

marked by the stages of Abbottabad, Garlii Habibullah, Muzaffarabad

and Baramula, is directly indicated in Alberuni’s itinerary.1

52. We may now proceed to examine the old notices regarding

this route. It started in Kasmir from the twin

G-ate of Varahamu a. 0f Yarahamfila- IDiskapura which occu¬

pied the sites of the present Baramula and Uskiir, respectively. Huska-

pura on the left river bank, though the more important of the two

places in ancient times, has dwindled down to a mere village. Varaha-

mula- Baramula, however, on the opposite bank is still a flourishing

place and an emporium of trade. It occupies a narrow strip of open

ground between the river and the foot of a steep mountain side.

Close to the western end of the town a rocky ridge with a precipi¬

tous slope runs down into the river-bed. Only a few yards’ space is

left open for the road. At this point there stood till last year (1897) an

old ruined gateway known to the people as the Drang or ‘ watch-station.’

It had been occupied as a military police post; until the ‘Bahdari ’

system was abolished, watch was kept here over those who entered or

left the Valley. I had examiued the gateway in 1892. When revisiting

the spot in May, 1898, I could scarcely trace its foundations. The

decayed walls had meanwhile been sold by auction, and its materials

carried away by a contractor.

Though the structure I had seen, was scarcely older than the time

of Sikh rule, there can be little doubt that it marked the site of the

ancient ‘Gate’ of VarahamCila. This is clearly indicated by the situa¬

tion of the spot which is by far the most convenient in the neighbourhood

for the purpose of a watch-station. Moorcroft does not mention the

name Drang , but describes the gateway itself accurately enough. Here

then, we may assume, stood in ancient times “ the stone gate, the western

entrance of the kingdom”, through which Hiuen Tsiang had passed before

he reached Huskapura ( Hu-se-kia-lo ), his first night’s quarter in the

Valley. Ou-k'ong too and Alberuni, as we have seen, knew well this

watch-station which is also mentioned by Kalhana under the general

designation of Dvara .2

The road keeps close by the bank of the river as it winds in rapid

fall through the rock-bound gorge. About two and a half miles below

‘ Drang’ the hill sides recede slightly, leaving room for a small village

1 See above, § 14 [The construction of a Tonga road between Abbottabad and

Muzaffarabad, recently sanctioned (1899), is sure to make the old route through

Hazara again popular.]

8 See Rdjat. viii. 413 note.

THE VITASTA VALLEY.

85

1899.]

Old frontier in

Vitasta Valley.

called Naran Thai. Near it stands a little temple, with a spring close by

which is visited by pilgrims and is probably identical with the Narayana-

sthana of the Nilamata.1

About a mile below this point and close to the village of Kiiad?'-

niyar,2 the river turns sharply round a steep and narrow spur project¬

ing into the valley from the north-west. A ledge of rocks continues

the spur below the river-bed and forms the first serious rapid of the

Vitasta below which boats cannot pass (see map). The road crosses the

spur by a narrow and deep cut, known as DyarVgul. Kalhana’s Chronicle

knows this curious cutting by the appropriate name of Yaksadara, ‘ the

demon’s cleft.’ According to the tradition there recorded the operations

by which Suyya, Avantivarman’s engineer, lowered the level of the

Vitasta, extended to this point of the river bed.3

53. Two miles below Dyar^gul we pass near the village of Zehenpor

some ancient sites vaguely described by Vigne

and Hiigel. Still further down near the

village of Gingal the map marks the ruins of a

temple which I have not been able to visit. But no localities on this

route are known to us from our old sources until after about three and a

half marches we reach the side valley marked on the map as ‘ Peliasa.’

This valley and the large village at its entrance are known indeed to

the Pahari population by the name of Peliasa. But the Kasmlris

settled at several places along the Vitasta Valley call them Buliasa.

This form of the name which I ascertained by local enquiries, enables

us to identify this locality with the BolyIsaka of the Rajatarangini.

Kalhana in his account of S'amkaravannan’s ill-fated expedition

towards the Indus (a.d. 902) mentions Bolydsalca as the place where the

Kasmir army retreating from Urasa reached the border of their own

territory.4 This reference is of special interest as it shows that Kasmir

authority extended in Hindu times down to this point of the Valley.

We can easily reconcile this fact with the existence of the ‘ Dvara ’ at

Varahamula.

The gorge at the latter place offered a convenient position for

establishing a watch-station which was to secure control over the traffic

and the collection of customs. But in regard to military defence

a frontier-line in the immediate vicinity of the Kasmir Valley would

have been very unsafe. I believe, therefore, that the Vitasta Valley

1 See Nilamata , 1179, 1315, 1349. The name occurs also repeatedly in the several

Y ara h aksetrama ha tiny a s .

2 Perhaps the Khddandvihdru of Rdjat. iii. 14.

3 Compare Rdjat. v. 87 note.

4 See Rdjat. v. 225 note. *

86

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2

below Varahamula was held as an outlying frontier- tract as far as the

present Buliasa. It is exactly a few miles below this place that ascend¬

ing the Valley the first serious difficulties are encountered on the road.

An advanced frontier-post could scarcely have occupied a strategically

more advantageous position.

The conclusion here indicated is fully supported by what Kalhana’s

narrative tells us of a locality almost exactly opposite to Buliasa.

Kalhana mentions in two places a place called Viranaka in connection

with events which make it clear that it lay in the Vitasta Valley and

just on the border of Kasmir territory.1 2 I have been able to trace the

position of Viranaka at the modern hill- village of Viran , near the left

bank of the Vitasta and only a short distance above Buliasa. The

valley below the old frontier thus marked is now known as Bvarbidi.

Its ancient name is given by an old gloss of the RajatarahginI which

speaks of Bolyasaka as situated in DvaravatI. Local enquiries have

shown me that even to the present day popular tradition indicates a

ridge a short distance above Buliasa as the eastern limit of Dvarbidl.3

In the account of S'amkaravar man’s above-mentioned expedition six

marches are reckoned from the capital of Urasa to Bolyasaka, This

agrees exactly with the present reckoning which also counts six marches

from the vicinity of Buliasa to Abbottabad.3 Near this place, the

modern head-quarter of the Hazara District, the old capital of Urasa

was in all probability situated.

54. It remains for us to notice briefly what is known of ancient

localities on the left side of the Valley. As

already explained there was no great line of

communication on this side corresponding to the present Murree-

Baramula Road. Yet for two marches down the Valley, as far as UrI,

the route of the left bank is likely to have been much frequented.

From UrI a convenient route leads over the easy Haji Pir Pass to

Prunts or Parnotsa. This pass owing to its small elevation, only 8500

feet, is never completely closed by snow. It is hence much used

during the winter-months when the more direct routes to Kasmir via

the Pir Pantsal, Tos^maidan or other high Passes are rendered imprac¬

ticable.

Left bank of Vitasta.

1 See Rdjat. v. 214 and viii. 409. In the first passage we hear of an attack

made on Viranaka by the chief commander of the frontier posts ( dvdresa ). In the

second Viranaka is referred to as a settlement of Khasas which offered the first safe

refuge to Sussala when defeated before Varahamula, A.D. 1111.

2 See Rdjat. v. 225 and note v. 214.

8 Compare Rdjat. v. 217 note ; Cunningham, Anc. Geogr ., p. 104, and Drew,

Jummoo, p. 528.

I

1899.] the vitasta valley. 87

Marching down the valley from Uskiir: Huskapura, we first cross

the spur which bounds the gorge of Varahamula from the south. We

then reach a fertile little plain, about two miles broad, charmingly

situated in an amphitheatre of high pine-clad mountains and facing the

Dyai ^gul ridge. It is known as Naravav and contains at the village of

S'ir and Fattegarh considerable remains of ancient temples. On a

small plateau which forms the western boundary of this plain by the

river bank, lies the village of Kitsthom It marks the site of the an¬

cient Buddhist convent of Krtyasrama , the foundation of which a curi¬

ous legend related by Kalhana attributes to the son of Asoka.1 Ou-

k'ong refers to it as the ‘ monastere du mont Kitche

At Buniar, near the end of the first day’s march we pass the well-

preserved ruins of an ancient temple which are of considerable antiqua¬

rian interest. Its name and date cannot be traced in our extant records.

Another similar ruin, but far more decayed, flanks the road about mid¬

way between Buniar and Uri.

From near the latter place the Vitasta Valley is held on the left

bank chiefly by the Khakha tribe, on the right by the closely related

Bombas. In the former we recognize the ancient KJiasas whose settle¬

ments lower down the Valley, at Viranaka, are distinctly mentioned by

Kalhana.2 The predatory habits and restless ways of the Khasas form

a frequent theme in the Chronicle. The modern Kliakhas and Bombas

have up to the middle of the present century done their best to main¬

tain this ancient reputation, just as their seats have remained the old

ones.

T See Rdjat. i. 147 note; also my Notes on Ou-k'ong, pp. 13 sqq « Krtyasrama is

mentioned already by Ksemendra, Samaijam, ii. 61.

2 Rdjat. viii. 409.

88

ANCIENT GEOGRAPHY OF K AS'M T R. [Extra No. 2,

Section IV. — Northern mountain range.

55. The mountains which enclose the Kasmir Valley in the north¬

west and north, may be looked upon as one

Range towards great range. Their chain nowhere shows any

Karnau and Sardi. s ° , X1 . .. . *

marked break though its direction changes

considerably. The routes leading through these mountains have never

been of such importance in the history of Kasmir as the routes towards

India and the west. Hence our information regarding the old topogra¬

phy of this mountain range is also less detailed.

We are least informed about that portion of the range which joins on

to the Kajnag Peak north-west of Baramiila and then continues in the

direction of south to north towards the upper Kisanganga. The water¬

shed of this portion forms the western boundary of Kasmir towards

Karnau , the ancient Karnaha.1 This territory which may be roughly

described as lying between the Kisanganga and the Kajnag Range, seems

at times to have been tributary to Kasmir. Yet wre hear of it only in

the concluding portion of Ival liana’s Chronicle, and there too no details

are given regarding the routes leading to it. These routes as the

map shows, start from the ancient districts of S'amald (Hamal) and

Uttara (Uttar).

At the point where the summit of the range comes nearest to the

Kisan^an^a, it takes a turn to the east and continues in this direction

for more than 100 miles. The summit ridge keeps after this turn at a

fairly uniform height of 12,000 to 13,000 feet for a long distance.

From the northern parts of the Uttar and Lolau Parganas several

routes cross the range in the direction of the Kisanganga.

Kalhana has occasion to refer to these in connection with the

expedition which took place in his own time against the Sirahsila castle.

This stood on the Kisanganga close to the ancient Tirtha of the goddess

S'arada still extant at the present S'ardi.2 One of these routes leads past

the village of Drang , situated at 74° 18' 45" long. 34° 33" 30" lat. It is

certain that the place took its name from an ancient watch-station here

located and is identical with the Drang A mentioned by Kalhana in con¬

nection with the above expedition.3 I have not been able to visit the

place in person but was informed in the neighbourhood that remains of

1 Compare Rajat. viii. 2485 note.

2 Compare regarding the S'araddtirtha and the castle of S'irahiilu , notes i. 36

( B) and viii. 2492 (L), respectively ; also below, § 127.

8 See Rajat, viii. 2507 note.

1899.]

NORTHERN AND EASTERN MOUNTAIN RANGES.

89

old watch-towers are still found on the path which leads up to the pass

behind the village of Drang.

Besides the route marked by this old frontier-station there are

others leading in the same direction. One is to the west over the

Sitalvan Pass ; the other lies in the west and passing through the

valley of Kroras descends directly to S'ardi along the Madhumati

stream. The portion of the Kisanganga Valley into which these routes

lead, can never have been of much importance itself though there are

indications of gold-washing having been carried on in it.1 2 But from

S'ardi starts a route leading very directly, by the Kankatori (Sarasvati)

River and over a high pass, into Cilas on the Indus ;a this line of cornu-

nication may already in old times have brought some traffic to S'ardi.

Owing to the inroads made by Cilasis and the restless Bomba

chiefs of the Kisanganga Valley, the Pathan Governors found it neces¬

sary to settle Afridis at Drang and the neighbouring villages to guard

the passes. The presence of these Afghan colonies shows that the con¬

ditions which necessitated the maintenance of the old frontier watch-

station at Draiiga, had altered little in the course oE centuries.

56. Above S'ardi the course of the Kisanganga lies for a long

distance through an almost inaccessible and

Pass of^Dugdha- uninhabited gorge. Hence for over 30 miles

eastwards we find no proper route across the

mountain range. Kalhana gives us a vivid and interesting account of

the difficulties offered by a winter-march along the latter when he

describes the flight of the pretender Bhoja from S'irahsila castle to the

Darads on the Upper Kisanganga.3

The line of communication we meet next is, however, an important

one. It leads from the north shore of the Volur lake into that part

of the Upper Kisanganga Valley which is known as Gurez, and connects

with the routes leading to Astor and the Baltl territory on the Indus.

The road used in recent years, and now improved by British engineers

into the ‘ Gilgit Transport Road,’ crosses the range by the Tragabal or

Razdiangan Pass, nearly 12,000 feet high. But the route frequented in

ancient times lay some eight miles further to the east.

Kalhana refers in several places to the hill fort of Dugdhaghata

which guarded the mountain-route leading into Kasrnir territory from

inroads of the Darads. The latter can easily be shown to have held

1 Compare Note B on S'arada ( Rajat . i. 36), §§ 2, 16. To this circumstance the

of Drang owes probably the distinguishing designation of Sun^-Drang ‘ the Gold

Drang,’ by which it is popularly known.

2 See Bates, Gazetteer , p. 490.

8 See Rajat. viii. 2710 sqq,

J. i. 12

90

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

then as now the Kisanganga Valley about Gurez and the neighbouring

territories to the north. From Kalhana’s description it is evident that

this frontier fort stood on, or close to, the summit of a pass. Thanks to

the indications of the Chronicle I was able to identify its site on the

top of the Dudvlchut Pass.1 2 The Pass (shown on the map by its ancient

name Dugdhaghata ) is approached on the Kasmir side from the valley

of the BandSpor stream, still known to the Brahmans by its old name

Madhumatl. At the small village of Atavutli (map ‘ Atawat ’) a side

valley is entered which is narrow and somewhat difficult below, but

higher up widens. Its highest portion which forms the immediate ap¬

proach to the pass, is an open alpine valley kuown to the mountain

shepherds as Vijje Marg.

The term Marg which denotes any high alpine grazing ground fre¬

quented in the summer by herdsmen, is the modern Kasmiri equivalent,

and direct derivative, of Skr. mathika. It designated originally the

small shelter-lmts of stone or wood usually erected on such high

plateaus or valleys by their summer occupants.8 It is probably that

Vijje Marg represents the Prajimathika which Kalhana mentions as the

position occupied by the Kasmir forces during their unsuccessful siege

of the fort.

As a characteristic point it may be mentioned that the garrison

depended for its water-supply on the storage of snow. This had become

exhausted at the late summer season when the siege took place, but,

luckily for the Darad defenders, was replaced by a fresh fall of snow.

The latter is explained by the elevation of the pass which I estimated

at about 11,500 feet. Snow-storms occur sometimes on the neighbouring

Trdgabal Pass so early as September.

From the Dud^khut Pass an easy track over the ridge marked

‘ Kiser ’ on the map leads down to Gurez, the chief place of the Valley.

The latter corresponds probably to the Daratpuri of the Rajatarangini.

The route over the Dud^khut, being very direct and comparatively

easy during the summer, was much frequented by Dard traders until

the recent construction of the ‘ Gilgit Transport Hoad.’ It was used by

the Sikhs for military convoys until a disaster caused by an avalanche

above Atavuth induced them to change it for the Tragabal route. It

also seems to have been mentioned to Baron Hug el.3 * * In Muhammadan

1 For detailed evidence regarding this location and a description of the site, see

Bajat. vii. 1171 note.

2 Skr. mathilcd is the diminutive of matha ‘ hut’, ‘ Sarai.’ The Ks. derivative

of the latter term, mar, is still used regularly for the rude shelter-huts which are

found on the higher passes particularly towards the north.

s See Kaschmir, ii, p. 169.

91

1899.] NORTHERN AND EASTERN MOUNTAIN RANGES.

times both routes were in charge of a ‘ Malik ’ who resided in the castle

of BandvJcoth, not far from the ancient Matrgrama shown on the

map.

In ancient times there probably existed in the same neighbourhood

a watch-station or Dranga. Ou-k'ong when speaking of the ‘ gate to

the north ’ through which the road led to Point or Baltistan, may have

meant either this Dranga or the fort of Dugdhagliata.

57. To the east of the Dud^khut Pass the summits of the range

, _ , , gradually get higher and higher until we reach

the great mountain-mass of the Haramukli

Peaks. Rising to close on 17,000 feet and surrounded by glaciers of

considerable size, these Peaks dominate the view towards the north

from a great part of the Kasmir Valley. Sacred legends have

clustered around them from early times. The lakes below their

glaciers belong still to the holiest of Kasmlrian Tirthas. The ancient

name of the Peaks is Haramukuta, ‘ S'iva’s diadem.’ This is explain¬

ed by a legend which is related at length in the Haracaritacintamani.1

Their height is supposed to be S'iva’s favourite residence.2 Hence

Kasmlrian tradition stoutly maintains that human feet cannot reach

the Peaks’ summit.3

The lake which lies at the foot of the north-eastern glacier, at a level

of over 13,000 feet, is looked upon as the true source of the Kasmir

Ganga or Sind River. It is hence known as Uttaraganga or popularly

Gangabal ,4 It is the final goal of the great ‘ Haramukutaganga ’ pil¬

grimage which takes place annually in the month of Bhadrapada and is

attended by thousands of pilgrims. The bones of those who have died

during the year, are on that occasion deposited in the sacred waters.

A short distance below this lake is another also fed by a glacier and

now known as Nundkol. Its old name Kalodaka or Nandisaras is

derived from a legend which makes the lake the joint habitation of

Kala, i.e., S'iva, and of his faithful attendant Nandin. From the

1 See Haracar. iv. 62 sqq .

2 The legends relating to S'iva’s residence on Mount Haramulcuta and his

connection with the several sacred sites of Nandiksetra, aro given at great length in

the Nilamata 1049 sqq.

3 Owing to this superstition I had great difficulty in inducing any of my

Kasmlrl Coolies (Muhammadans !) to accompany me on the ascent I made to the

Peaks in September, 1894. My Brahman friends could not give credence to my

having reached the summit. According to their opinion the very fact of my

having reached the Peak was a sufficient pi-oof of this not having been Haramukuta.

An argument as simple as incontrovertible to the orthodox mind.

4 Sec my note Rdjat. i. 57. Another name often used in the Nilamata and

other texts is Uttaramdnasa ; see Raj at. iii, 448 note.

92

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

latter the whole collection of sacred sites takes the name of Nandiksetra

by which Kalhana usually designates it.1

In the valley of the Kankanai stream (Skr. KanakavahinI) which

issues from these lakes, lies the sacred site of S'iva Bhutesvara (now

Buthbser). It is closely connected with the legends of Mount Hara-

mukuta and often mentioned in the Rajataranginl.2 A series of interest¬

ing temple ruins marks the importance of this Tirtha and that of the

ancient Jyesthesvara shrine which immediately adjoins it.3 Bhutesvara

is passed by the pilgrims on their way back from the sacred lakes, while

on their way up they reach the latter by another route, passing the

high ridge known as Bharatagiri and the smaller lake of Brahmasaras.

From the Gahga lake a track passable for ponies leads over the

Satsaran Pass to Tilel , a Dard district on the Kisanganga. It is probably

the route by which King Harsa’s rebel brother Vijayamalla escaped

from Laliara (Lar) to the Darad territory.4 *

58. Eastwards from the Haramukuta Peaks the range does not

overlook on the south the main Valley of

Kasmir, but that of the Sind River. The gene¬

ral level of the summits rises, and glaciers of fair size become frequent

on their northern slopes. Close to the head of the Sind Valley,

the range we have been so far following joins on to the great chain of

snowy mountains which stretches from Mount Nanga. Parvat in a south¬

easterly direction to the Nankun Peaks in S'uru.6 A few miles south

of this junction we arrive at a gap in the mountains which forms the

lowest watershed between the Indus and the Vitasta basins. It is the

Pass known generally by its Ladakhi name of Zoji-Ld. It leads at an

elevation of 11,300 feet from Baltal, on the headwaters of the Sind, to a

high-level valley draining into the Dras River and hence into the Indus.

The route leading over the Zoji-La undoubtedly has been already

Route over ZojI La anc*en^ fimes a most important thorough¬

fare. It connects Kasmir with Ladakh and

thence with Tibet and China. Here too the natural watershed has in

old as in modern times been also the ethnic boundary. Beyond the

Pass begins the land of the Bhauttas or Bhuttas , as the Tibetan inhabi-

North-eastern range.

1 See Rdjat. i. 36 note.

2 See regarding the history and remains of Bhutesvara , Rajat. i. 107 ; v. 55

notes. The Tirtha was rich enough to attract a special expedition of marauding

hill men in Kalhana’s time ; see viii. 2756.

3 See Rdjat. i. 113 note.

4 See Rdjat . vii. 911.

6 Compare regarding this great range which may fitly be called the main range

of the mountain system around Kasmir, Drew, Jummoo , pp, 194 sqq.

1899.] NORTHERN AND EASTERN MOUNTAIN RANGES.

93

tants of the Indus region are uniformly designated in our Kasmirian

texts (modern Ks. Butt).1

Ou-k'ong is the first who refers distinctly to this route when

speaking of the road which leads through the gate in the east to Tou-fan

or Tibet. Kalhana has scarcely occasion to refer to it, as the regions

beyond the Pass lay quite beyond the reach of the political power of

the later Kasmirian kings. He probably means, however, the Zoji-La,

when mentioning the route of the Bhutta- land ( Bhuttarastradhvan ) by

which the Darads offered to pass the pretender Bhoja into Kasmir,

while the more direct routes from their own territory were closed by

the winter.2 An easy pass connects Tilel at the head of the Kisanganga.

Valley with the Dras territory to the east. From there Bhoja could

then have entered Kasmir via the Zoji-La.

This Pass, the ancient name of which is not known to us, has more

than once witnessed successful invasions of Kasmir. Through it came

early in the 14th century the Turk (P) Dulca and the Bhautta Rincana

whose usurpation led to the downfall of Hindu rule in the Valley.3

About two centuries later Mirza Muhammad Haidar with his small

Mughal force successfully fought here his entrance into Kasmir

(a.d. 1532). The account he gives of this exploit in his Tarikh-i-

Rashidi , is not without topographical interest.4

59. A high peak situated about 10 miles east-southeast of the

Zoji-La, marks the point where the range forming the eastern

boundary of Kasmir branches off from the main chain. This range

runs in an almost southerly direction until it reaches the southern¬

most headwaters of the Vitasta. It then turns to the north-west and

at the Banahal Pass joins on to the Pir Pantsal Range. Through this

range there lead routes connecting Kasmir with the Madivadvan

Valley (see § 84) which drains into the Cinab, and with Kastavar,

the ancient Kastavata, on the Cinab itself. Both these Valleys are con¬

fined, difficult of access, and scantily populated. They have hence never

played an important part either in the foreign relations or trade of

Kasmir. On this account our notices regarding the old topography

of the dividing range are extremely meagre.

1 Compare Rajat. i. 312-316 note.

8 Compare Rajat. viii. 2887.

3 See Jonar. 142 sqq., and for the stratagem by which Rincana forced his way

into Lahara (Lar), 165 sqq. The Laharakotta mentioned in the last passage pro¬

bably represents the old watch-station of this route, but its position is un¬

certain.

4 See Tdrikh-i-Rashtdi , transl. by Messrs. Elias and Ross, pp. 423 sqq., and

below, § 131.

94

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

At its northern end and close to the great snowy peak already

Tirtha of mentioned, is the Tirtha of Amaresvara or

Amaranatha Amaranatha, known by its Kasmiri name as

Amburnath, Together with the sacred Gaiiga-

lake on Mount Haramukuta, it is now the most popular of Kasmirian

pilgrimage places. Its Yatra in the month of Sravana attracts many

thousands of pilgrims not only from Kasmir but from all parts of

India.1 Their goal is a cave situated at a considerable altitude and

formed by a huge fissure on the south side of a snowy peak, 17,300 feet

high (marked ‘Ambarnath’ on map). In this cave there is a large

block of transparent ice formed by the freezing of the water which

oozes from the rock. It is worshipped as a self-created ( svayambhii )

Liiiga, and is considered the embodiment of S'iva- Amaresvara.

Judging from the scanty references made to this Tirtha in the

Rajataraiigini and the Nilamata, it appears doubtful whether it could

have enjoyed in old times quite such great celebrity as now. But

Jonaraja already relates a visit to this sacred site paid by Sultan

Zainu-l-‘abidiu,2 and in the Mahatmya literature Amaresvara receives

its due share of attention. The pilgrims’ route described in great detail

by the Amaranathamahatmya ascends the valley of the eastern branch

of the Lidar or Ledari.

There the lake of the Naga Susravas, now known as Susramnag

or (with a popular etymology) S'esanag, is visited at the north foot of a

great glacier descending from the Kohenhar Peak. In this lake and

a small rock-bound inlet of it called Jamatrnaga (ZamHur1 Nag), the

local legend, related by Kalhana, i. 267 sgrq., and connected with the

ancient site of Narapura, has placed the habitation of the Naga Su¬

sravas and of his son-in-law.3 The route then crosses a high pass, known

as Vavajan (Skr. Vayuvarjana in the Mahatmya), into a high-level valley

drained by five streams which bear the joint designation of Panca-

taranginI. From there the pilgrims toil up a lofty spur to the north¬

east and descend into the narrow gloomy valley which lies at the foot

of the Amburnath Peak. It is watered by a stream (Amaravati)

which comes from the glacier of a still higher peak to the east.

Joining the Paiicatarangim it flows through an inaccessible gorge down

to the head of the Sind Valley near Baltal.

1 See for the old notices of the Tirtha, Rdjat. i. 267 note ; for a description of the

modern pilgrimage, Vigne, Travels , ii. pp. 10 sqq., and Bates, Gazetteer, pp. 121 sq.

2 Compare Jonar. (Bombay ed.) 1233 sqq.

3 Compare Rdjat, i, 267 note.

1899.]

NORTHERN AND EASTERN MOUNTAIN RANGES.

95

60. Connected with the eastern range is a mass of mountains

which it will be convenient to mention here

Watershed range ,, , , £ , £ .

, , ,, though it does not form part oi the mountain-

between Smdhu p f *8

and Vitasta barriers of Kasmir. It fills the great trian¬

gular space which lies between the Sind Yalley

and the range in the east we have just noticed, the level ground along

the right bank of the Vitasta forming as it were the base. This mass

of mountains separates from the eastern ridge between the Kohenhar

and Amburnath Peaks. Trending westwards it soon culminates in the

conspicuous pinnacle of Mount Gastbrar (map ‘Kolahoi’), close on

18,000 feet in height. From this conspicuous mountain numerous spurs

radiate with glaciers in their topmost hollows.

The highest of these ridges runs for about thirty miles along the

Sind Yalley, of which it forms the southern side. A high cross-spur, now-

known as Durun Nar, which descends to the north towards Sunamarg,

is probably identical with Mount Dhudavana, the scene of a siege

related in the Kajataranginl.1 The extremity of this ridge in tlie west

forms the amphitheatre of bold hills which encircle the Dal lake and

STinagar on the north. Here we have Mount Mahadeva which is

much frequented as a Tirtha.2

Facing it from the south is the rocky spur which lines the eastern

shores of the Dal. It bore in old days the name of S'rIdvaea,3 and is

the site of a series of ancient pilgrimage places, such as Suresvari,

Tripuresvara, Harsesvara, and Jyesthesvara, which will be discussed

below. The extreme offshoot of this spur is the ‘ Hill of Gopa ’

( Gopadri ), the present Takht-i Sulaiman, which is so conspicuous a

feature in the landscape of STinagar. Other spurs descending into the

vale further east form successively the semicircular side-valleys contain¬

ing the Parganas of Vihl and Vular.

We now return once more to the eastern range. South of the

Kohenhar Peak which is still over 17,000 feet high, its summit ridge

gets gradually lower. It is crossed by the Margan Pass into Madi-

vadvan. Of the latter valley I can find no old mention. Still further

south we come to the Marbal Pass, at an elevation of 11,500 feet, which

forms the usual route towards Knstavar.

This territory which is now partially inhabited by Kasmiris, is

mentioned as an independent hill-state by Kalhana.4 The valley into

1 See Rdjat. viii. 595 note and below, § 131.

2 It is mentioned in the NUamata , 1324, and frequently in the S'arvdvatdra.

3 See Rdjat. viii. 2422 note.

4 Compare regarding the old Kdsthavdta note vii, 588-590, where also the

references in the later Chronicles are given.

96

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

which the route descends immediately after crossing the Marbal Pass,

is known now as Khaisal. It is once mentioned as Khasali by Kalhana

and more frequently referred to in the last Chronicle by the name of

Khasalaya.1 From the latter source we learn that it was inhabited

by Khasas to whose occupation it may have owed also its name. So we

note here once more in the east the coincidence of the ethnic boundary

with the natural watershed.

Section V. — Upper Course of the Vitasta.

61. We have now completed the circuit of the great mountain-

barriers which enclose the Kasmir Valley, and can turn our attention

to its interior. This is naturally divided into two great parts. One

comprises the plain formed by the alluvium of the Vitasta and its main

tributaries ; the other consists of plateaus or Karewas elevated above

the river flats and largely caused by old lacustrine deposits. We shall

first notice the alluvial plain and the river-system which has created it.

The great river which is the recipient of the whole drainage of the

countrv, is now known to Kasmiris by the

Name of Vitasta. "» Tr ,7 mt. , , . , . .

name or Vyath. This modern designation is the

direct phonetic derivative of the ancient Sanskrit Vitasta which we meet

already among the river-names of the Kigveda.2 The intermediary

Prakrit form *Vidasta underlies the Hydaspes of the Greeks in which

we note, as so frequently in Greek renderings of foreign names, the

modifying action of popular etymology.3 In Ptolemy’s Bidaspes we

have another rendering which though later in date yet approaches

closer to the sound of the Indian original.4 The name Jelilam which is

1 Compare Rajat. vii. 399 note.

s The line of phonetic development may be roughly represented as Skr.

Vitasta >Pr. *Vidasta >Ap. *Vi[K]ath >Ks. Vyath.

The name Vitasta is still well-known to Kasmir Brahmans from the Mahatmyas

and similar texts, and is currently used by them. The form ‘ Vedasta ’ which Drew

and other writers indicate as the old name of the river “ still used by those who

follow Sanskrit literature,” is due to some error of hearing. It is curious to

meet a similar form *Vidastd in the transcription of the Chinese Annals of the 8th

century ; see my Notes on On-Wong, p. 31.

3 The ending in the form Hydaspes is undoubtedly due to the influence ' of the

numerous Persian names known to the Greeks which end in -acnnjs (Old Persian

aspa). For the rendering of initial Vi- by *Y compare Hystaspes : Vishtdspa.

4 Ptolemy’s Bi (for Vi) is the most exact phonetic reproduction possible in Greek

characters. It is evident from Ptolemy’s Panjab river names that he did not take

1899.]

UPPER COURSE OF THE YITASTA.

97

Legendary source of

Vitasta.

now borne by the Vitasta in its course through the Panjab, is wholly

unknown to the genuine usage of Kasmlr. It is apparently of Muham¬

madan origin and has been brought to Kasmir only by Europeans and

other foreigners.1

The river to which the name Vitasta or Vyath is properly applied,

is first formed by the meeting of the several streams which drain the

south-eastern portion of the Valley. This meeting takes place in the

plain close to the present town of Anatnag or Islamabad. But sacred

tradition has not failed to trace the holiest of Kasmlr rivers to a more

specific source.

An ancient legend, related at length in the Nilamata and reproduc¬

ed by the author of the Haracaritacintamani,2

represents the Vitasta as a manifestation of

S'iva’s consort Parvatl. After Kasmir had

been created, Siva at the request of Ka^yapa, prevailed upon the

goddess to show herself in the land in the shape of a river, in order to

purify its inhabitants from the sinful contact with the Pisacas. The

goddess thereupon assumed the form of a river in the underworld, and

asked her consort to make an opening by which she might come to the

surface. This he did by striking the ground near the habitation of the

Nllanaga with the point of his trident ( sulci ). Through the fissure

thus made which measured one vitasti or span, the river gushed forth,

receiving on account of this origin the name Vitasta. The spring-basin

where the goddess first appeared was known by the several designations

of Nilakunda , S'ulaghata (‘spear-thrust’) or simply Vitasta .3 It is clear

that the spring meant is the famous Nllanaga, near the village of

Vernag in the Shahabad Pargana. It is a magnificent fountain which

amply deserves the honour of being thus represented as the traditional

source of the great river.4

The legend makes Parvatl- Vitasta subsequently disappear again

from fear of defilement by the touch of sinful men. When brought to

light a second time by Kasyapa’s prayer the goddess issued from the

Naga of Pancahasta. In this locality we easily recognize the present

his nomenclature directly or indirectly from the historians of Alexander, but from

independent sources. Bidaspes, Zaradros, Bibasis, Sandabal, these all represent

unsophisticated attempts to reproduce in sound the genuine Indian forms. The

same cannot be said of the names given by Arrian, Pliny, etc.

1 Alberum already knows the name Jailam ; see above, § 14. STIvara when

relating an expedition of Sultan Haidar Shah into tiro Panjab, sanskritizes this

name into Jyalami ; see ii. 152.

8 Sec Nilamata, 23S sqq.-, Haracar. xii. 2-34.

3 See Nilamata, 1290; Haracar. xii. 17.

4 Compare for the Nllanaga and its round spring-basin (kunda), Raj at, i, 28 note

J. i. 13

98

ANCIENT GEOGRAPHY OF KAS'MlR.

[Extra No. 2,

village of Panzath, situated in the Div^sar Pargana and boasting of a

fine spring which is still visited by the pious of the neighbourhood.1

After another disappearance for a reason similar to the above, the

goddess came forth a third time at Narasiriihasrama. This place I am

unable to trace with certainty. Finally the goddess was induced to

abide permanently in the land when Kasyapa had secured for her the

company of other goddesses, who also embodied themselves in Kasmir

streams, like Laksmi in the Visoka, Gfariga in the Sindhu, etc.

Another version of the legend which, however, seems of less ancient

date, seeks the place of the Vitasta’s second appearance in the spring

of the modern Vitkfivutur , a small village situated about one mile to the

N. W. of Vernag.2 The place is known by the name of Vitastdtra to

Kalliana who mentions Stupas erected there by King Asoka.3 This

notice certainly seems to indicate some sacred character attaching to

the spot. Yet Kalliana’s direct mention of the Nilakunda ns the birth¬

place of the Vitasta leaves no doubt as to where the tradition prevalent

in his own time placed the source of the sacred river.4 *

62. The streams which unite close to Anatnag and there form the

true Vitasta river, are the Sdndran , the Bring ,

Arapath and Lidar . Of these the first and

southernmost drains the Shahabad (or Ver)

Pargana and receives the water of the sacred springs mentioned in

the preceding paragraph. Its old name I am unable to trace with any

certainty. The next affluent, the Bring , comes from the side-valley

which forms the Pargana of the same name. The ancient name of the

stream is unknown, the modern Vitastamahatmya which gives it as

Bhrhgi , being but a doubtful authority. The Bring too is fed by the

water of some well-known Nagas, among which the famous Trisamdliya

fountain and the springs of Ardhanarisvara (Naru) may be specially

mentioned.

The Arapath which comes from the north-east, is mentioned

repeatedly in the Nllamata by its ancient name of Harsajiathd .6 The

Headwaters of

Vitasta.

1 Compare Rdjat. v. 24 note.

8 This version is found in the Vitastamahatmya, ii. 37, sqq ., which calls the

place Vitastdvartikd ; see also Vigne, i. p. 335.

3 See Rdjat. i. 102 note.

4 Rajat. i. 28. I am unable to account for the mention made in the Mahdbh. iii.

Ixxxii. 90 of the Taksaka Naga in Kasmir as the Vitasta, i e., its source. No such

distinction is claimed for the well-known Taksaka spring near Zevan (Jayavana) ;

see Rdjat. i. 220. The author of the Tirthaydtrd in the Mahabh. shows no accurate

knowledge of Kasmir and seems to have made a mistake here.

6 See Nilamata , 232, 1299, etc.

1899.]

UPPER COURSE OF THE VITASTA.

99

valley it drains is known as tlie Kother Pargana and takes its name

from the sacred tank of Kapatesvcirci. At the western end of the spur

on the slope of which this Tirtha is situated, issue the magnificient

springs of Achabal (Aksavala). They form a small stream by themselves,

which flows into the Harsapatha. A short distance below the village

of Khanabal (map ‘ Kanbal ’) where the three streams hitherto men¬

tioned unite, their waters are joined from the north by those of the

Lidar.

This river, the ancient Ledari ,l receives a number of glacier-fed

streams which drain the high range towards the Upper Sind Valley. It

is hence in volume more considerable than any of the previously named

affluents. The Ledari spreads in several branches through the wide

valley forming the Parganas of Dachiinpbr and Khovurpor which take

their names, ‘ Right Bank ’ and ‘ Left Bank,’ respectively, from their posi¬

tion relative to this river. In old days a canal constructed on the hill¬

side to the east carried the water of the Ledari, and with it fertility, to

the barren plateau of Martanda or Matan.2

63. At Khanabal the Vitasta becomes navigable and continues

. so on its whole course through the valley.

Vitasta’s course m Tl . ,

alluvial plain I here too the great flat plain begins which

stretches on both sides of the river down to

Baramula in the north-west. In its course to the Volur lake, a direct

distance of about 54 miles, the river falls only some 220 feet.3 The

slope in the general level of the plain is equally gentle. The bed of

the river lies everywhere in the alluvial soil, the result of the deposition

of sediment at flood times when the river overflows its banks. Down

to STinagar the river keeps in a single bed and its islands are but

small, in fact mere temporary sandbanks. The course is in parts very

winding. But as far we can judge from the position of the old sites

along the river, no great changes are likely to have taken place in

historical times in this portion of the river’s course.

When the river is low as during the winter, the banks rise on an

average about 15 feet above the water. But in the spring when the

snow melts, the great volume of water brought down from the moun¬

tains rises to the top of the banks and often overflows it. Dangerous

floods may also follow long and heavy summer- rains, and sometimes

1 See Rajal. i. 87.

2 The construction of this canal by Zainu-l-‘abidin is described at length by

Jonar. (Bo. ed.) 1232-60. It is probable that there existed earlier irrigation works

on the same plateau. See below, § 111,

3 See Drew, Jummoo, p. 163.

100

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

cause immense damage to the crops over a great portion of the cultivat¬

ed area of the Valley.1

Such floods and the famines which are likely to follow, were a

danger well-known in old times already and are more than once men¬

tioned by Kalhana.2 Against them the villages and riverside towns

have always endeavoured to protect themselves by artificially raising

the banks. The allusions found in the Chronicle suffice to show that

the construction of embankments ( setu , now suth), with the accom¬

panying system of floodgates closing lateral drainage channels, has

existed since ancient times.3 4 One great regulation scheme which was

directly designed to diminish these risks, and of which we possess a

detailed historical account, will be discussed below. The equally

elaborate system by which water was secured for the irrigation of the

otherwise dry alluvial flats along the river, will also be specially

noticed.

The navigable waters of the Vitasta have from ancient times to

the present day formed the most important highway of Kasmir. The

value of the river and of the numerous canals, lakes, and streams which

are also accessible to boats, for the development of internal trade and

traffic can hardly be overestimated. Until a couple of years ago there

were nowhere in Kasmir, not even in the flattest parts of the Valley

roads fit for wheeled traffic. Carriages were practically things unknown

to the population bred in the Valley. As long as the communication

with the outer world was restricted to difficult bridle-paths or tracks

passable only for load-carrying Coolies, the construction of such roads

would have been, in fact, of very slight advantage. The importance

of river-traffic in Kasmir may be estimated from the fact that the

number of boatmen engaged in it (and their families) amounted accord¬

ing to the census of 1891 to nearly 31,000. 4 That boats were in old

days, just as up to the present time, the ordinary means of travel in

the Valley, is shown by the frequent references to river journeys in the

Chronicles.5 6

Equally eloquent testimony to the historical importance of river

navigation in Kasmir is borne by the position of the ancient sites. We

1 Compare for data as to modern floods, Lawrence, Valley, pp. 205 sqq.

2 See Rdjat, vii. 1219 ; viii. 2449, 2786 ; also vii. 1624 ; viii. 1417, 1422 ; Jonar.

403 sqq.

3 See Rdjat. i. 159; iii. 483; v. 91, 103, 120; viii. 2380, etc.; Jonar. 404, 887 ;

S'rlv. iii. 191 sq., etc.

4 Compare regarding the Han zi of Kasmir, Lawrence, Valley , p. 313; also

Rdjat, v. 101 note.

6 See Rdjat. v. 84 ; vii. 347, 714, 1628, etc.

1899.]

UPPER COURSE OF TIIE YITASTA.

101

shall see that all the towns which from time to time were the capitals

of the country, were built on the banks of the Vitasta, and that the

great majority of other important places of ancient date were similarly

situated. It is certain that then as now all produce of the country

was brought to the great centres by water. Villages even when

situated at a great distance, had, no doubt, just as at the present day,

their landing places (Ks. yarabcil ) on the river or the nearest navigable

waterway. Kalhana’s description of the semi-legendary city of

Narapurct shows how closely the busy “ coming and going of ships ” was

connected in the Kasmirian mind with the splendour of a large

town.1

The Gambhlra.

64. After these general remarks we may now proceed to follow

the Vitasta’ s course through the Valley noticing its tributaries in

due order as we reach the confluences. . Below Khanabal 2 3 the river

receives in succession the several branches of the Ledarl and then passes

the ancient town and Tirtlia of Vijayesvara, the present Vij^bror.

About a mile lower down, its course lies between high alluvial plateaus

or Karewas. One on the left bank, the Tsakadar Udar, will be noticed

below as one of the most ancient sites of the Valley (Cakradhara).

About three miles further down and not far from the village of

Marhom (the old Madavasramci) , the Vitasta

is joined by the Vesau and Rembyar?' Rivers

which meet a short distance above their common confluence with the

Vitasta. This river junction is known to the Mahatmyas by the name

of GambhIrasamgama (‘the deep confluence’) and is still visited as a

Tirtha.s The short united course of the Vesau and Rembyar^ bears

the old name of GambhIra and is referred to under this designation

repeatedly by Kalhana. The Gambhlra is too deep to be forded at any

time of the year, and being on the route from Vijayesvara to S'rinagar,

is of military importance. It was twice the scene of decisive actions.

King Sussala’s army on its retreat over the Gambhlra (a.d. 1122)

suffered a complete rout. Six years later Sujji, his son’s general, gained

an equally signal victory by forcing the passage in the face of a rebel

army.4

1 See Bdjat. i. 201 sq.

2 According to a gloss on NUamata 1307, Kkanabal, the port, so to say, of

Anatniig corresponds to the Khandapuccha Naga of that text. This Naga is else¬

where mentioned, but I have no distinct evidence for its identification.

3 See Bdjat. iv. 80 note. Junctions of rivers and streams ( samgamas ) are every¬

where in India favourite places for Tirthas.

4 See Rdjat. viii. 10G3 sqq., 1497 sqq.

102

ANCIENT GEOGRAPHY OF KASMIR. [Extra No. 2,

The Ve&iu, frequently mentioned by its ancient name of Visoka in

the Chronicles, the Nllamata and other texts,1

The Visoka. . . , ,, T . . ,,

is a considerable river. It receives all tne

streams coming from the northern slope of the Pir Panteal Range

between the Sidau and Banahal Passes. Its traditional source is placed

in the Kramasaras or Konsar Nag Lake below the Peak of Naubandhana.

The Nilamata, 271 sqq. relates a legend which identifies the Visoka with

Laksmi and accounts for its name (‘ free from pain ’). The fine water¬

fall which is formed by the stream of the Konsar Nag not far from

the village of Sidau, is now known as Alirabal. The Nilamata calls it

Akhor bila ‘ the mousehole,' which may possibly be the origin of the

modern name.2 As soon as the Visoka emerges from the mountains,

numerous irrigation canals are drawn from it which overspread the

whole of the old Parganas of Karala (AtGvin) and Devasarasa

(Div^sar).

One of these canals is the SunQman * Kul which is mentioned in the

Rajatarahginl by its ancient name of Suvarnamanikulya.3 If the story

of its construction by King Suvarna, reproduced from Padraamihira,

could be trusted, we should have to ascribe to this canal a high anti¬

quity. It leaves the Visoka near the village shown as ‘ Largoo ’ on the

map and rejoins it near the village of .Sd^vin (map ‘Arwin’). Another

old canal, called Nandi (not shown on the map), leaves the Visoka near

Kaimuli, the ancient Katimusa, and irrigates the land between the

lower course of this river and the Vitasta. Its name is connected

perhaps with that of the village Nandaka which is referred to in con¬

nection with Avantivarman’s drainage operations.4’ The Visoka is

navigable up to Kaimuli.

The Rembydr 9 which joins the Visoka a little above Gambhlra-

saihgama, we have met already before as the

river uniting the streams from the Pir

Pantsal and Rupri Passes. Kalb ana mentions it by its ancient name

RamanyatavI when relating the legend of the burned city of Narapura.5

The Rembyar?' after leaving the mountains below Hiir^por flows

divided in many channels within a wide and mostly dry bed of rubble

and boulders. This strip of stony waste along the river attains a width

of over two miles near the village of Tsuran (map ‘ Charran ’).

The local legend referred to attributes the creation of this waste to

The RamanyatavI.

1 See Rdjat. iv. 5 note.

2 Compare Nilamata, 27l sqq., and for Akhor bila, 283.

8 See Rdjat. i. 97.

4 See Rdjat. v. 85 note.

6 8ee Rdjat. i. 263-265, note.

1899.]

UPPER COURSE OF THE VITASTA.

103

the Nagi Ram any a. She had come down from the mountains carrying

masses of stone to assist her brother, the Susravas Naga, in the destruc¬

tion of Narapura. When she learnt that he had already completed his

task, she dropped the stones ‘more than a Yojana ’ from the site of the

doomed city. The distance indicated corresponds exactly to that of the

village of Litar where the Rembyar?' leaves behind its stony bed and

passes into alluvial soil. The village land for five Yojanas above that

place was buried by the mighty boulders which Ramanya left along

her trail. Similar tales regarding the origin of stone-wastes (“ Mur-

ren ”) are well-known to European alpine folk-lore.

65. Below Gambhirasamgama the Vitasta receives from the right

_ . the stream which drains the ancient district

The Vitasta near * rr i j- ai , Tr , T.,t

cs - of kLolada, the present Vular. It then passes

close to the foot of the Vastarvan spur, near

the old town of Avantipura. No important stream joins the river

from the right until we reach STinagar. The affluents on the left

like the Ramusa are also of small volume. Some do not reach the river

direct but end in low marshes, communicating with the latter only by

gates made in the river embankments. Of the ancient sites situated

along the river, the town of Padmapura, the present Pampar, is the

most considerable. As we approach S'rinagar we pass the site of the

ancient capital, Puranadhisthana, marked by the present village of

Pandrethan. It lies between the right river bank and the southern

foot of the ridge which encircles the Dal. For the streams we have

next to notice, a reference to the special map of Ancient STinagara is

necessary.

Just before we reach the area of the city proper, the Vitasta is

joined by a stream which drains .the lake to the east of the city. This

lake, known as Dal (Skr. Dala), is fed by plentiful springs and by

streams which reach it from the north. Its surplus waters flow out

towards the Vitasta by a canal which is now called Ts until Kul, but in

ancient times bore the name of Mahasarit.1 This canal passes through

an ancient embankment ( setu ) which protects the city as well as the

low shores of the Dal from floods of the river, and already figures in the

traditional account of the foundation of S'rinagar. The position of the

gate which closes the outflow of the Mahasarit is marked on the map by

the entry ‘ Durgagalika.’

A small channel from the river — whether artificial or natural cannot

1 For the identification of the Tsiinth Kul and the Mar canal in the city with

the Mahasarit, my note on Rajat. iii. 339-349 should be consulted. In addition to

the evidence there recorded, it should be noted that the Mahasarit is twice mention¬

ed by its old name also in the S'arvdvatara iii. 74 ; iv, 129 s<p

101

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

be ascertained now — joins the Mahasarit at this very point and turns

the ground between it and the river into an island. This is now known

by the name of MaySmim, derived from the ancient Mak^ikasvamin.

We shall have to refer to it again in our account of the topography of

STinagar. From Durgagalika downwards the Mahasarit or Tsunth Kul

was in old times the south-eastern boundary for that part of Srinagar

which lies on the right bank of the Vitasta. Being a natural line of

defence it is frequently referred to in the narrative of the various sieges

of the capital.1 *

The confluence of the Mahasarit and Vitasta which is just opposite

to the modern palace, the Shergarhi, has been a Tirtha from early times

and is mentioned by its correct name inMarikha’s description of Kasmir.8

STivara refers to it by a more modern name, Martsamgama , where

Mari is an evident adaptation from the Ks. form Mar .3 The latter

name, itself a derivative of Mahasarit, is applied at the present day to

another branch of the Dal outflow. This turning to the west passes

through the marsh known as BraiTnambal ( Bhattaranadvald ) and then

enters the city.

This canal is of considerable importance for the internal traffic of

the city as it opens a convenient waterway to the Dal and greatly

facilitates the transport of its manifold produce.4 After passing behind

the whole of the city quarters on the right river-bank the Mar issues

near the quarter of Narvot* (Skr. Nadavana ) into the marshes of the

AnclPar. Through the latter a connection is thus secured with the Sind

river delta. This extension of the Mar to the west seems, however, of

later date, as STivara attributes the construction of a navigable channel

towards the Sind to Zainu-l-‘abidin.5 * *

66. The lake which supplies the water of the Mahasarit, is in

some respects one of the most favoured spots of

the whole Valley. Its limpid water, the

imposing aspect of the mountain amphitheatre which encloses it on

three sides, and the charming gardens and orchards around it have

made the Dal justly famous.

The Dal lake.

1 See Bdjat. viii. 733, 753, 3131.

* See S'riTcanthac . iii. 24, Mahdsaridvitastayoh ... samgamah. Here too as in

former translations of the Rajataranginl, Mahasarit has been wrongly taken as an

common nonn and explained as ‘ great river.’

3 The term Mari is also elsewhere used in the later Chronicles and the Mahat-

inyas ; comp., e.g., S'riv. i. 442; iv. 298; Fourth Chron, 145, etc.

4 It is this narrow canal, more picturesque than sweet-smelling, which has led

to the frequent comparisons of S'rmagar with Venice. It has not received much

attention in recent years and for want of dredging seems in danger of silting np.

s See S'riv, i, 440 sg.

UPPER COURSE OF THE VITASTA.

105

1899.]

The Dal measures about four miles in length and two and a half

in width where it is broadest. Its depth nowhere exceeds 30 feet, and

in most parts it is far more shallow. At its southern end it is fringed

by lagoons, and a great portion of it is covered by the famous floating

gardens. Notwithstanding the superabundance of water-plants and

vegetable matter, the water everywhere retains an admirable clearness

and freshness. This is, no doubt, due to the ampleness of the springs

which rise within the lake. Though we find no direct mention of the

lake in the Rajataraiigini, and though it does not claim any parti¬

cular sanctity, there is no want around its shores of ancient and

holy sites.

The earliest reference to the lake itself occurs in the Chronicle of

Sri vara who describes at length how King Zainu-l-‘abidln diverted

himself on the lake and adorned its vicinity.1 2 S'rivara calls the lake

Dala, while the few Mahatmyas which condescend to mention it, use

the form Data} He also mentions the two small artificial islands

called Lanka , and now distinguished as BupHank and Sunaldnk (‘ the

Silver Lanka,’ ‘ Golden Lanka’). Different names are given to several

distinct portions of the lake. But of these only Hastavalika, the

present Ast^vol, can be traced in the Clnonicles.

The sacred sites of Gopadri, Jyesthesvara, Theda, S'uresvarl, etc.,

with their numerous Nagas line the eastern shores of the Dal. They

will be mentioned below in the description of the vicinity of the capital.

The well-known gardens of Shalimar, Nishat and Nasim are creations

of the Mughal Emperors who did much to enhance the natural beauties

of the lake.

Besides the springs of the lake itself the latter is fed also by a

stream which comes from the Mar Sar lake, high up in the mountains to

the east. The old name of this stream, marked ‘ Arrah ’ on the map, is

uncertain. The S'arvavatara seems to extend to it the name Mahasarit .3

In its lower course where it approaches the north shore of the Dal, it

now bears the name of Telbal Nal (stream). An earlier form is fur¬

nished by S'rivara who calls the stream at this point, by the name of

Tilaprastha ; the latter is also found in several Mahatmyas*

67 Erom the junction with the Mahasarit downwards the Vitasta

flows for over three miles between almost unbroken lines of houses

raised high above the water on stone embankments. The latter consist

now-a-days chiefly of large blocks of stone which belonged to ancient

1 6'riv. i. 418 sqq.

2 See, e.g., Vitastamah , xxi. 39.

3 See S'arvav. iii. 75 ; iv. 129.

* See S'rtv. i. 421 ; S'arvdv. iv. sqq., etc.

J. i. 14

106

ANCIENT GEOGRAPHY OF KASHMIR.

[Extra No. 2,

The Ksiptika.

temples and other structures of pre-Muhammadan date. Judging from

their size and careful carving we can well picture to our mind the

splendid appearance which the river-banks must have here presented in

bygone days.

The river within the city flows first in one long reach due north.

Near the fourth bridge in the heart of the city,

it makes a great bend and turns to the south¬

west. A canal which leaves the left bank of the river between the

Shergarhi palace and the quarter of Kathul (Kasthila), and rejoins the

river near the last bridge, allows boats to cut this great bend. It now

bears the name Kut^kul, derived from the ancient designation of

Ksiptikakulya.1 The Ksiptika is often mentioned in the later portions

of Kalhan.a’s Chronicle which relate the sieges of S'rinagar witnessed in

his own time. It forms to this day the natural line of defence for that

part of the city which occupies the left river-bank, and which could be

successfully attacked only by crossing the Ksiptika.2 No information

is available to us as to the origin of this canal. Judging from its posi¬

tion it is likely to have been a natural side-channel of the river which

was subsequently maintained or improved for the convenience of navi¬

gation.

A few hundred yards lower down the Vitasta is joined on its left

bank by a considerable river now known as

the Dudgangtl , ‘ the milk Graiiga/ or C'hatstlcul ,

‘ the white stream.’ Its ancient name is given as Dugdhasindhu in

Bilhana’s description of S'rinagar.3 The Mahattnyas know it by the

name of S'vetagangci , ‘the white Gaiiga,’ to which the alternative modern

designation, Chats^kul, exactly corresponds.4 * Its waters come from the

central part of the Pir Pantsal Range round Mount Tatakuti, its chief

sources being the mountain-streams marked as ‘ Sangsofed ’ (Sangsafed)

and ‘ Yechara ’ on the map. The confluence of the Vitasta and Dud-

gariga, opposite to the old quarter of Diddamatha, is still a Tirtha of

some repute and is probably alluded to already by Bilhana.6

The Dugdhaganga.

1 Ks. kul < Skr. Jculyd is the ordinary term for small streams or canals.

2 For detailed evidence of this identification, see Rdjat. viii. 732 note.

8 Vikramahkac. xviii. 7.

4 See Vitastamdh. xxiij S'vetagahgdmdh., etc. Skr. sveta becomes in Ks. by

regular phonetic conversion chuth, fern. cftotg®.

The Nilamata curiously enough does not mention the Dudganga unless it is

meant by Kfiranadi , 1281. The latter name, meaning ‘ the river of milk,’ is given

to the Dudganga by a passage of the modern Yitastamahatmya, xxii.

6 See Vikram. xviii. 22.

1899.]

LOWER COURSE OF THE VITA ST A.

107

Section VI. — Lower Course of the Vjtasta.

The Sindhu,

68. Immediately below S'rinagar we come to marshes which

stretch along both sides of the river for a considerable distance.

Those on the left bank, of which the Hukhasar and Panz^nor Nambal

are the nearest, are fed by mountain-streams of smaller volume. The

marshes to the north of the river are more extensive and belong to the

Delta of the Sind River, the greatest tributary of the Vitasta within the

Valley.

Our survey of the northern range of mountains has already taken

us to the true headwaters of the Sind near the

Zoji-La and the Amburnath Peak. Its tradi¬

tional source in the sacred Gaiiga-lake on Mount Haramukh has also

been noticed. This great river has a course of over sixty miles and

drains the largest and highest portion of the mountain-chain in the

north. Its ancient name, Sindhu, means simply ‘ the river’ and is thus

identical with the original designation of the Indus.1 2 The Raja-

tarangini mentions the river repeatedly, and it figures largely in the

Nllamata, Haracaritacintamani and the Mahatmyas.3 Everywhere it is

identified with the Granga, as already by Alberuni’s informants. The

valley of the Sind forms the district of Lar, the ancient Lahara, one of

the main subdivisions of Kasmir territory.

Where this valley debouches into the great Kasmir plain, near the

village of Dud^rhSm, the old Dugdhasrama, the river spreads out in

numerous branches. These form an extensive Delta, covered in its

greatest portion by shallow marshes and known as Anch'ar. Its eastern

side extends along the strip of high ground which connects STinagar

with the foot of the spur at the mouth of the Sind Valley. The western

1 It is customary in Kasmir to distinguish the two rivers by giving the designation

of * the Gi’eat Sind (Bud Sind),’ to the Indus. This is found as ‘ Brhatsindhiv,*

already in the Haracaritacintamani, xii. 45.

The identity of the two river names has led to a great deal of confusion in

geographical works down to the beginning of the present century. The Sind River

of Kasmir was elevated to the rank of one of the chief sources of the Indus, or else

represented as a branch of the great river taking its way through Kasmir (!). This

curious error is traceable, e.g., in the map of c L’Empire du Grand Mogol * reproduced

in Bernier’s Travels , ed. Constable, p. 238, from the Paris Edition of 1670, and in

the map of Ancient India attached to Tieffenthaler, Description de V Inde, 1786,

p. 60. Compare Huger, Kaschmir, i. p. 330. Even Wilson, writing in 1825, says

of the Kasmir Sind that “ it is not improbably a branch of the Indus.”

2 See Bajat, i. 57 note; also iv. 391; v. 97 sqq. ; viii. 1129; Jonar. 982; S’rtv. iv.

110, 227, etc.

108

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

side of tlie delta is marked by an alluvial plateau which continues the

right or western side of the lower Sind Valley down to the river’s con¬

fluence with the Vitasta. The base of the triangle is the Vitasta itself

which between STinagar and this junction flows in a bed separated by

artificial banks from the marshes on either side. The waters of the

Sind after spreading over this wide Delta leave it in a single channel at

its western extremity, opposite to the village of Shadipur.

The confluence of the Vitasta and Sindhu has from early times

enjoyed exceptional sanctity as a Tirtha. Ka§-

ConfluencG of Vitasta - ,

mir tradition, as recorded already m the

a n U Nilamata, identifies the Vitasta and Sindhu,

the largest a ad holiest rivers of the country, with the Yamuna and

Ganga , respectively. Their junction represents, therefore, the Kas-

mirian equivalent of the famous Prayaga at the confluence of the great

Indian rivers. The Vitastasindhusamgama is often referred to as an

important Tirtha in the Rajatarangini, the Nilamata and numerous

other texts. It is actually known by the name of Prayaga to the modem

tradition and the Mahatmyas.1 A small island built of solid masonry

rises in the river-bed at the point where the waters of the two rivers

mingle. It is the object of regular pilgrimages on particular Parvans

throughout the year. On it stands an old Cinar tree which to the

pious Kasmirian represents the far-famed Ficus Indica tree of the real

Prayaga.

Notwithstanding the accumulated holiness of this Tirtha there is

most explicit evidence to show that its present position dates back only

to about a thousand years. We owe the knowledge of this interesting

fact to the detailed account which Kalhana has given us of the

great regulation of the Vitasta carried out under King Avantivarman

(a.d. 855-883), As the change in the confluence of the Vitasta and

Sindhu forms one of the most striking results of this regulation,

Kalhana’s account of the latter may conveniently be noticed in the

present place. I shall restrict myself to an indication of the main facts

connected with these operations, referring for all detailed evidence to

Note I (v. 97-100) of my translation.

69. Kalhana tells us in his opening notice * that the produce of

_ Kasmir had in earlier times been greatly

Suyya’s regulation of , . , , . , , . _ J

the Vitasta restricted owing to disastrous floods, particu¬

larly from the Mahapadma or Volur lake, and

the general water-locked condition of the country. Drainage operations

1 For a detailed account of the references to the Vitasta-Sindhusamgama and

the ancient remains near it, see Note I ( Rdjat . v. 97-100), §§ 14, 15 ; also note iv. 391.

* See Rdjat, v. 68 sqq.

LOWER COURSE OF THE VITASTA.

109

1899.]

under King Lalitaditya had led to an increase of agricultural produce.

But these works were apparently neglected under his feeble successors,

and disastrous floods, followed by famines, became again frequent. In

Avantivarman’s time Suyya , a man of conspicuous talents but low

origin, offered to remedy these troubles. Receiving the king’s assent

for his scheme and the necessary means, he set about regulating the

course of the Vitasta with a view to abetter drainage of the whole Valley.

Omitting legendary details with which evidently popular tradition has

embellished Suyya’s story, the course adopted was briefly the following.

The operations commenced in Kramarajya at the locality called

Yaksadara where large “ rocks which had rolled down from the moun¬

tains lining both river banks,” obstructed the Vitasta.1 We

have already when describing the Vitasta Valley route, referred to

Yaksadara, the present Dyar^gul, as a spur projecting into the river-bed

some three miles below the commencement of the Baramula gorge. Its

rocky foot forms the first rapid of the river. By removing the obstruct¬

ing rocks the level of the river was lowered. Then a stone-dam was

constructed across the bed of the river, and the latter thus blocked up

completely for seven days. During this time “ the river-bed was cleared

at the bottom, and stone walls constructed to protect it against rocks

which might roll down.”8 The dam was then removed, and the river

flowed forth with increased rapidity through the cleared passage.

I must leave it to competent engineering opinion to decide to what

extent and at which point of the Baramula gorge the operations so far

described were practicable with the technical means of that age. What

follows in Kalliana’s account is so matter-of-fact and so accurate in

topographical points, that a presumption is raised as to the previous

statements also resting, partially at least, on historical facts.

Wherever inundation breaches were known to occur in times of

flood, new beds were constructed for the river. One of these changes

in the river-bed affected the confluence of the Vitasta and Sindhu, and

this is specially explained to us in v. 97-100. The topographical indi¬

cations here given by Kalhana are so detailed and exact that they

enabled me to trace with great probability what I believe to have been

the main course of the Vitasta before Suyya’s regulation.

70. Kalhana describes to us successively the position of the old and

Change of confluence the new confluence relative to certain temples

of Vitasta and situated at the village of Trigrami and other

Sindhu. points on the river-banks. Most of these struc¬

tures I have been able to identify, and a close examination of the

1 See v. 87 sqq.

2 Compare v. 92 sq.

110

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

general topographical features in tlieir neighbourhood has fully borne

out the correctness of Kalhana’s account. Without the help of a larger-

scale map it would, however, be impossible to explain here accurately

the topographical evidence collected.1 I must, therefore, once more

refer to the above-quoted detailed note in my forthcoming work,

where a special map, on the scale of one inch to the mile, has been

inserted for the illustration of this tract. In the present place I must

restrict myself to indicating the main results of my enquiries.

These have shown that while the new confluence which Kalhana

knew iu his own time, is identical with the present junction opposite

Shadipur,2 the old one lay about two miles to the south-east of it,

between the village of Trigam and the Par^spor plateau. The latter is

the site of the great ruins of Parihasapura , first identified by me and

.shown on the map (see below, § 121.) Trigam marks the position of the

ancient Trigrdmi , and a short distance south of it stands the temple ruin

which I identify with the shrine of Visnu Vainyasvdmin.

Ktilhana mentions this temple as the point near which “ the two

rivers, the Sindhu and Vitasta, formerly met flowing to the left and

right of Trigrami, respectively.” Standing on the raised ground before

the ruin and turning towards Shadipur, we have on our left a narrow

swamp about a quarter of a mile broad which runs north-east in the

direction of Trigam. In this swamp and a shallow Nala- continuing it

towards Shadipur, we can yet recognize the old bed of the Sindhu. On

the right we have the Badrihel Nala which divides the alluvial plateau of

Trigam and Pai^spor. This Nala is clearly marked as an old river-bed

by the formation of its banks and is still known as such to the villagers

of the neighbourhood.

The Badrihel Nala connects the great swamp to the east known as

Panz'nor Nambal with the extensive marshes stretching west and north¬

west of Par^spor towards the Volur. This channel still serves regular¬

ly as an outflow for the PanzHior Nambal whenever the latter is flooded

from the Vitasta at times of high-water. Were it not for the great

embankments which guard the bed of the Vitasta towards the low

Panzer Nambal, the latter would still form a regular course of the

1 In the copies of the map accompanying this memoir the faint outlines by which

the engraved Atlas of India sheet marks the low alluvial plateaus, the marshes, and

similar features of this tract, have become much effaced. A reference to the

original groundmap or the larger Survey map (2 miles to the inch) is hence recom¬

mended.

2 Shadipur is a modern contraction for Shahahuddln'pur , the name given to the

place by Sultan Shahabu-d-dm (A.D. 1354-73) who founded it, as Jonaraja, 409, tells

us, at the confluence of the Vitasta and Sindhu.

1899.]

LOWER COURSE OF THE V1TASTA.

Ill

river. Even so it is still liable to be invaded by tbe Vitasta at times of

flood. For tbe swamp as well as the fertile village lands reclaimed

around it, lie below the level of the river-bed.1

The old course of the rivers here briefly indicated explains the

curious position of the Nor (map ‘Noroo’). This canal which is of

importance for navigation leaves the Vitasta. on the left bank just

opposite to the present junction with the Sindhu and practically conti¬

nues the southwesterly course of the latter for some distance. Only

about | mile of low ground divides the Nor from the end of the swamp

which marks the bed of the Sindhu at the point of its old junction

opposite the Vainyasvamin ruin.

Similarly the position of Parihcisapura which King Lalitaditya

chose for his splendid capital, becomes now intelligible. The plateau

or Karewa of Par^spor which still preserves its name is now flanked

on the east by the Panzinor Nambal and on the west by the marshes of

HarHrath. Neither of them affords in their present condition the

convenient waterway we find invariably near all other Kasmir capitals.

Before Suyya’s regulation, however, the Vitasta flowed as we have seen,

immediatly to the north of the plateau and at the very foot of the great

temples erected here by King Lalitaditya.

71. The object and result of the change of the confluence can, I

? think, also be traced yet. By forcing the

Results of Suyya s Vitasta to pass north of Trigam instead of

south of it, the reclamation of the marshes south

of the Volar lake must have been greatly facilitated. The course thus

given to the river carries its waters by the nearest way into that part of

the Volur which by its depth and well-defined boundaries is naturally

designed as a great reservoir to receive the surplus water of dangerous

floods. The southern shores of the lake are still to this day the scene

of a constant struggle between the cultivator and floods. The reclama¬

tion of laud which has gone on for centuries in these low marshy

tracts,8 could never have been undertaken if the Vitasta had been

allowed to spread itself over them from the south, the direction marked

by its old course.

The change in the confluence of the Vitasta and Sindhu was a neces¬

sary condition for the subsequent course given to the united rivers. It

• See Lawrence, Valley, pp, 210 sq. Kalhana’s account shows that the huge

embankments guarding the Panzhior tract must be far older than the times of the

Mnghals to which they are popularly attributed.

8 Compare Drew, Jtnnmoo, p. 116, for a description of these tracts and the

amphibious ways of the inhabitants who get their living as much from the water as

the land around.

112

ANCIENT GEOGRAPHY OF K AS' MIR. [Extra No. 2,

was thus closely connected with the general scheme of regulation and

drainage. Kalliana indicates this by referring immediately after the

above passage to stone-embankments constructed along the Yitasta for

seven Yoianas (circ. 42 miles) and the damming-in of the Yolur lake.1

On the land reclaimed new populous villages were founded. From

the circular dykes which were built around these villages, they are said to

have received the popular designation of kundala, ‘ ring.’ We actu¬

ally still find two villages on the low ground near the Yolur showing

in their modern names the ending kundal, derived from Skr. kundala.

Ut,b^kundal (map wrongly ‘ Watr koondl ’) and Mar^kundH are situated

both close to the left bank of the Yitasta before it enters the marshes at

the south-eastern end of the Volur. Their names and position seem to

support the assumption that the present northerly course of the river

above Trigrami and Shad1 par is directly due to Suyya’s operations.

Kalhana adds that even in his own time, i.e., two and a half cen¬

turies later, there were “ seen, growing on the banks of the former

river-beds, old trees which bore the marks of the boat ropes fastened to

them.”2 Similarly the observant Chronicler noted the old pales secur-

ing the embankments “ which the rivers displayed when low in the

autumn.” 3 We must be grateful to him for the evident interest with

which he ascertained and recorded the details of Avantivarman’s opera¬

tions. For he has thus enabled us even at the present day to trace

some of the important changes then effected in the hydrography of the

whole Yalley.

72. Following the course of the Yitasta below its present conflu¬

ence with the Sindhu we soon pass the village

Course of Vitasta to of Sambal where the route from STiuagar to

wards the o ur a p|ie north of the Yolur lake and thence to the

Tragabal Pass, crosses the river. Here at some distance from the left

bank is the site of the ancient Jayapura, the capital founded by King

Jayapida in the second half of the eighth century.4 It is marked by

the village of And^rkoth situated on an island between the Sambal

marsh and a branch of the canal known as Nor. An ancient causeway

connects the island with the strip of land separating the marsh from

the present course of the Yitasta.

1 See v. 103 sqq.

8 v. 101.

8 It is still the common belief in Kasmir that “ no embankment on the riverside

is sound unless it has a foundation of piles ” ; Lawrence, Valley, p. 211. Consider¬

ing the peaty nature of the soil along the lower course of the river, this belief may

be justified by old experience.

4 See for the identification of this site, Raj at. v. 506 note, and below, § 122.

1899.]

LOWER COURSE OF THE VITASTA.

113

We should have some difficulty in understanding the position

chosen for a town which was intended to be a place of importance if we

did not know the great change effected in the course of the river by the

subsequent regulation of Avantivarman. In King Jayapida’s time one

of the main branches of the Vitasta probably followed the line of the

Nor in this neighbourhood. The island of And^rkoth which forms a

small alluvial plateau, raised perhaps artificially in parts, was then a

convenient site. This is no longer the case since the liver flows to the east

of And^rkoth and at a considerable distance. We can safely attribute to

this change the fact that Jayapura like the similarly situated Pari-

liasapura had fallen into insignificance already before Kalhana’s time.

Close to Sambal the river passes the foot of an isolated hill known

as AhHyung , rising about a thousand feet above the plain. Under its

shelter on the north is the small lake of Manasbal which is mentioned by

the name of Manasa[saras ] in the Nilamata and by Jonaraja.1 2 * It is about

two miles long, and occupying a rock-basin is deeper than the other

lakes of the Kasmir plains. It is connected with the river by a short

channel and partially fed by an irrigation canal carried into it from the

Sind River.8 Its ancient name is derived from the sacred lake on

Kailasa, famous in the Puranas and Epics and usually located in the

Mansarovar of the Tibetan highlands.

A short distance lower down the villages of Uts^kundal and Mar^-

kundal already referred to above, are passed on the left bank. There

are various indications which make it probable that in old times the

Volur lake reached much closer to these villages than it does at present.

Kalhana’s reference seems to indicate that these villages enclosed by

circular dykes were actually reclaimed from the lake, and Jonaraja

still places them on the very shore of the lake.8 In the same way

S'rivara speaking of the villages stretching from Samudrakota,4 * the

present Sudarkoth, to the vicinity of Dvarika, near And^rkoth, seems

to place them along the shore of the Volur.

A glance at the map shows that the land on the left bank of the

river below the ‘ Kundala’ villages projects like a peninsula into the lake.

1 As Jonaraja , 864 sq., makes the ancient name quite certain, the latter could

have safely been shown on the map. In some passages of the Nilamata and

Mahatmyas it might be doubted whether this lake or the Uttaramanasa on Mount

Haramukh is intended; see however NUamata, 1338, where the Manasa lake is

mentioned after the Yitastasindhusamgama.

2 The construction of this canal is ascribed by Jonaraja, 864 sq ., to Zainu-1-

‘abidin.

2 See Itajat. v. 120, and Jonar. 1230, (Bo, ed.j.

4 See Srlv. i. 400 sq,

J. i. 15

114

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

It can be safely assumed that the creation of this strip of land which

now accompanies the river-channel for some seven miles farther, is due

to the continual deposits of silt. This silting-up process is still going

on in this as in other portions of the Volur where streams enter it, and

is likely' to reduce the expanse of the lake still further in the future.1

A striking proof for the gradual change thus effected is afforded by

the position of the artificial island known now as Zainalank. It was

constructed by King Zainu-l-‘abidin from whom it took its proper

designation of Jainalanha. It was then, according to Jonaraja’s descrip¬

tion, in the middle of the Volur where the water was deep.2 It is

now situated in a shallow marsh close to the present embouchure of

the river.

73. The great lake, with the southern shores of which we have

Volur lake already become acquainted, is a very impor¬

tant feature in the hydrographic system of

Kasmir. It acts as a huge flood-reservoir for the greatest part of the

drainage of Kasmir and gives to the western portion of the Valley its

peculiar character. Its dimensions vary at different periods, owing to

the low shores to the south being liable to inundation. In normal

years the length of the lake may be reckoned at about 12 and its width

at 6 miles, with an area of about 78 square miles. In years of flood

the lake extends to about 13 miles in length and 8 miles in width.3

Its depth is nowhere more than about 15 feet and is continually lessen¬

ing in those parts where the streams debouch into it. Notwithstanding

this slight depth navigation on the lake often becomes dangerous when

violent storms sweep over it from the mountains in the north. The

boundaries of the lake are ill-defined in the south and partly in the east;

the marshes and peaty meadows merge almost imperceptibly into the

area of the lake. On the north the shores slope up towards an

amphitheatre of mountains from which some rocky spurs run down to

the water’s edge. The fertile tract at the foot of these mountains

forms the ancient Khuyasrama , the modern Pargana of Khuy^hom.

The ancient name of the lake is Mahapadmasaras, derived from the

, -_T_ Naga Mahapadma , who is located in the lake

Manapadma Naga. ° r ’

B as its tutelary deity. This designation is by

far the most common in the Chronicles, the Nilamata, and other old

1 Compare Drew, p. 166, and Lawrence, Valley , p. 20. The latter author ia

probably reproducing a popular tradition when mentioning that in King Zainu-1-

‘abidin’s time the waters of the Volur stretched south to Asam and Sambal.

2 See Jonar. (Bo. ed.) 1227 sqq. The name Jainalanka was mutilated in the

Calcutta^edition ; else it would have been shown on the map,

s See Lawrence, p. 20.

1899.]

LOWER COURSE OF THE YITASTA.

115

Legends of Maha-

padma Naga.

texts.1 It is also used, as we have seen, in tlie description of Kasmir

given by the T'ang annals.

The name Ullola from which the present Volur ( vulgo 1 Woolar ’)

seems to be derived, is found only in one passage of Jonaraja’s Chronicle

and in a single modern Mahatmya.8 Skr. ullola can be interpreted to

mean ‘ turbulent ’ or ‘ [the lake] with high-going waves.’ 3 Those who

have experienced the sensation of crossing the lake with a strong wind,

will readily allow the appropriateness of this designation. Yet it is

impossible to dismiss altogether the suspicion that the name which

seems wholly unknown to the older texts, may be only a clever adapta¬

tion of the Kasmiri name Yolur or its earlier representative. It is cer¬

tainly curious that in modern Mahatmyas we meet with Ullola as a nam

for the Vulgar Pargana, the genuine ancient designation of which is

Holada.4 * Jonaraja in his commentary on STikanthacarita, iii. 9, uses

Ullola as a paraphrase for Mahapadma.

74. From an early date various legends seem to have clustered

around this, the greatest of Kasmir lakes. The

Nilamata relates at length how the lake be¬

came the habitation of the Mahapadma Naga.6

Originally it was occupied by the wicked Naga Sadahgula who

used to carry off the women of the country. Nila, the lord of Kasmir

Nagas, banished Sadahgula to the laud of the Darvas. The site left

dry on his departure was occupied by a town called Gandrapura under

King Visvagasva. The Muni Durvasas not receiving hospitable recep¬

tion in this town, cursed it and foretold its destruction by water. When

subsequently the Naga Mahapadma sought a refuge in Kahnir and

asked Nila for the allotment of a suitable habitation, he was granted

permission to occupy Gandrapura. The Mahapadma Naga thereupon,

approached King Visvagasva in the disguise of an old Brahman and

asked to be allowed to settle in the town with his family. When his

prayer was agreed to he shewed himself in his true form and announced

to the King the approaching submersion of his city. At the Naga’s

direction the King with his people emigrated and founded two Yojanas

further west the new town of Visvagasvapura. The Naga then converted

the city into a lake, henceforth his and his family’s dwelling place. A

recollection of this legend still lives in popular tradition, and the ruins of

the doomed city are supposed to be sighted occasionally in the water.

1 For detailed references see Raj at. iv. 593 note.

2 See Jonar. (Bo. ed.) 1227-30; Dhyanesvaramah. 30, 33.

3 See Buhler, Report , p. 9.

4; See Vitastdmah. v. 48 ; Haridrdganesamdh.

6 See Mlamata, 976-1008, and Buhler, Rej)ort, p. 10,

116

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Another legend has found a lengthy record in Kalhana’s narrative

of King Jayapida’s reign, iv. 592 sqq. The Naga Mahapadma being

threatened with desiccation by a Dravidian sorcerer, appeared to the

King in his dream and asked for protection. As a reward he promised

to show a gold mine to the King. Jayapida agreed to the Naga’s

prayer. Curiosity, however, induced him to let the Dravidian first try

his magic on the lake. When the waters had been dried up so far

that the Naga and his dependents were seen as human-faced snakes

wriggling in the mud, the king interfered and caused the lake to be

restored. The Naga, however, resented the insult and showed to the

king only a rich copper ore instead of the gold mine.

With reference to a Puranic legend the Mahapadma is sometimes

identified with the Naga Kaliya who was vanquished by Krsna. As

the foot of the god when touching the Naga’s head made lotuses

(padma) appear on it, Mahapadma is treated by Kasmirian poets as

another form of Kaliya,1 2

75- Of the streams which fall iuto the Volur lake besides the

Lower affluents Vitasta, the stream of the Band^por Nala is

of Vitasta. the mos^ considerable. It drains the range

between Mount Haramukh and the Tragabal

Pass and forms a small Delta of its own to the north of the lake. Its

ancient name is 21 ladJiumati?’ It is repeatedly mentioned in the

Rajatarangini in connection with the route leading to the Dard terri¬

tory, but must be distinguished from another, smaller Madliumati which

flows into the Kisanganga near the S'aradatirtlia.

The outflow of the lake’s waters is at its southwest corner

about two miles above the town of S5pur. The latter is the ancient

Suyyapura , founded by Suyya and commemorating his name.3 If we

may judge from the position of the town and the words used by Kalhana

in another passage,4 it appears probable that the operations of Avan-

tivannan’s great engineer extended also to the river’s bed on this side

of the lake.

About four miles below S5piir the Vitasta which now flows in a

winding but well-defined bed, receives its last considerable tributary

within Kasmir. It is the Pohur which before its junction has collected

the various streams draining the extreme northwest of the Valley.

1 Compare SHhanthac. iii. 9 ; Jonar. 933, and my note Raj at. v. 114.

2 See Rdjat. vii. 1179 and note 1171 j also viii. 2883 j Nilamata 1259 sqq.,

1398, etc.

8 See Rdjat. v. 118 note.

4 V. 104 : “ Trained by him, the Vitasta starts rapidly on her way from the

basin of the Mahapadma lake, like an arrow from the bow.”

SOIL AND CLIMATE OF THE VALLEY.

117

1899.]

This portion of the country figures but little in Kalhana’s narrative ;

hence we find in the Rajatarahginl no reference to the Pohur or any

of its affluents. The old name of the river is uncertain. Jonaraja in a

passage which is found only in the Bombay edition, calls this river

Pahara ; the Mahatmyas vary between Prahara and Prahara.1 2 Of the

side-streams the Mavur (map ‘ Maur’) flowing through the Machipor

Pargana is named in the Nilamata as Mdhurl.% The name of the Hamal

stream is identical with that of the Pargana through which its course

lies, the ancient S'amald .3

About 18 miles from the point where the Yitasta leaves the Volur,

it reaches the entrance of the gorge of Baramula. Through this defile

we have already before followed the course of the river. At Baramula

navigation ceases. After passing with a violent current the ravine

immediately below the town, the river, so placid within the Yalley,

turns into a large torrent rushing down in falls and rapids.

Section YII. — Soil and Climate op the Yalley.

Alluvial Plateaus

(Udar).

76. 0 ur survey of Kasmir rivers has taken us along that great

flat of river alluvium which forms the lowest

and most fertile part of the Yalley. We

must now turn to the higher ground of the

Yale which consists of the peculiar plateaus already alluded to.

The genuine Kasmiri term for these plateaus is udar, found in its

Sanskrit form as uddara in the Chronicles. Auother modern designa¬

tion of Persian origin now often used, is karewa. The word uddara is

twice found as an ending of local names in the Rajatararigini 4 *, while

the latter Chronicles use it frequently in designations of well-known

plateaus.6 An earlier Sanskrit term no longer surviving in use, is

suda, originally meaning ‘ barren waste ground.’ Kalhana employs it

when speaking of the well-known Damadar Udar.6

The Udars of the Kasmir Yalleys are usually considered by geolo¬

gists to be due to lacustrine deposits. They appear either isolated by

1 See Jonar. (Bo. ed.) 1150, 1152 ; Vitastdmdh. xxvii. 2; Svayamhhdmdh.

2 Nilamata, 1322 sqq.

3 See Bdjat. vii. 159 note.

4 Locanodddra and Dliyanodddra, Rdjat. viii. 1427 note.

6 See Gusikodddra, the Udar of Gus near Ramuh, S'riv. iv. 465, 592, 596; Ddmo-

darodddra, the Datnadar Udar, S'riv. iv. 618 ; Laulapurodddra, Fourth Chron. 175 etc

6 See Rdjat. i, 156 note.

118

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

lower ground around them or connected by very gentle slopes with

spurs descending from the mountains. Often the tops of these plateaus

seem almost perfectly flat, forming table lands of varying dimensions.

They rise generally from 100 to 300 feet above the level of the ravines

and valleys which intersect them, and through which the streams from

the mountains and their own drainage find their way to the Vitasta.

Most of the Ildars are found on the south-western side of the Valley,

stretching from S'upiyan to Baramula. But they also occur across the

river on the north-eastern side of the Valley, and at both extremities

of the river-flat in the south-east and north-west.

Owing to the inferiority of the soil and the difficulty of irrigation,

the Udars show a marked difference in point of fertility from other

parts of the Valley. Those which slope down from the foot of the

mountains have been brought under cultivation with the help of water¬

courses conducted over them from the higher ground behind. Most of

these irrigation-channels are, no doubt, of ancient date, and some are

specially mentioned in the Chronicles. To other Udars, particularly

those which are entirely isolated, water could not be brought. These

are either barren wastes covered with low jungle or if cultivated, yield

only precarious crops owing to the uncertainty of the rainfall.

Some of the Udars, owing to their position near the Vitasta or for

other reasons, are sites of importance in the ancient topography of

Kasmir. Such are the plateaus of Martanda, Cakradhara, Padmapura,

Parihasapura. Another, the ‘ Udar of Damodara,’ plays an interesting

part in the legendary lore of the country. All these will be duly

noticed in the next chapter.

77. Climatic conditions are so closely connected with a country’s

, topography that the few old notices and

Kasmir climate. n , - , ,

references which we have regarding those or

Kasmir, may fitly find mention here.

The only distinct account of the Kasmir climate is given by

Alberuni.1 He clearly indicates the reason why Kasmir is exempt from

the heavy Monsoon rains of India proper. When the heavy clouds, he

explains, reach the mountains which enclose Kasmir on the south, “the

mountain-sides strike against them, and the clouds are pressed like

olives or grapes.” In consequence “ the rain pours down and the rains

never pass beyond the mountains. Therefore Kasmir has no varsaJccila ,

but continual snowfall during two and a half months, beginning with

Magha, and shortly after the middle of Caitra continual rain sets in

for a few days, melting the snow and cleansing the earth. This rule

has seldom an exception ; however, a certain amount of extraordinary

meteorological occurrences is peculiar to every province in India.”

1 See India, i. p. 211.

1899.] SOIL AND CLIMATE OF THEWALLEY. 119

That this description is on the whole as accurate as Alberuni’s other

data regarding Kasmir, will be easily seen by a reference to the detailed

statements of Mr. Lawrence and Mr. Eliot.1 2 What chiefly characterizes

the climate of Kasmir as against that of the Indian plains, is the

absence of a rainy season and the equally marked absence of excessive

heat. The moderate temperature of the Kasmir summer is ensured by

the high elevation of the Valley, and has at all times been duly appre¬

ciated by its inhabitants as well as its visitors.

Kalhana already proudly claims this exemption from the torments

of a fierce sun as one of the favours accorded to his country by the gods.8

His enthusiastic description of a Kasmir summer passed “ in the regions

above the forests ” shows that he was no stranger to the charms of that

season in the alpine parts of the country.3 More than once he refers to

the sufferings which the heat of an Indian summer outside the Valley

inflicts on Kasmirian exiles. Even in the hill regions immediately to

the south of Pir Pantsal the hot season with its accompanying fevers

has often proved disastrous to the Kasmirian troops employed there.4 *

On the other hand we find also the rigours of a Kasmir winter

duly illustrated by the Chronicle's narrative. We may refer to the

description of the heavy and continued snowfall which followed Sussala’s

murder in Phalguna of 1128 a.d., the freezing of the Vitasta in the

winter of 1087-8 a.d., etc.6 The graphic account of Bhoja’s flight to the

Upper Kisanganga Valley shows us in full detail the difficulties which

attend a winter-march over the snow-covered mountains to the north of

the Valley.6 Eor do we fail to be reminded otherwise of the great"

differences in climate which are implied by the varying altitudes of

Kasmir localities.7

Exceptionally early snowfall in the autumn such as saved the

garrison of the frontier fort on the Dugdhaghata Pass, has always been

known and dreaded even low down in the Valley. The danger it

represents for the rice crops is illustrated by Kalhana’s account of the

famines resulting from such premature snowfalls.8

1 See Lawrence, p. 24 sqq.

2 See i. 41.

8 ii. 138.

4 Compare vii. 970 ; viii. 1634, 1830, 1836,1865; regarding the fever-season of

Rajapuri and neighbouring districts, my note viii. 1873.

8 Bdjat. viii. 1376 sqq. ; 1434 sqq. ; vii. 592.

6 See viii. 2710 sqq. It must be remembered that as much as forty to sixty feet

of snow falls in a severe winter on the higher ranges around Kasmir ; see also viii.

411.

I Compare vii. 916; viii. 2511; ii. 138.

8 See ii. 18 sqq. ; viii. 2449.

120

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

Cultivation.

In this as in other respects there is nothing to suggest any material

change of the climatic conditions during historical times. Kalhana, it

is true, in describing the reign of Abhimanyu I., speaks of deep snow

as “ falling each year to cause distress to the Bauddhas ” and obliging

the king to pass six months of the cold season in Darvabhisara. But

the whole story there related is nothing but a mere rechauffe of the

ancient legend told in the Nilamata of the annual migrations caused

by the presence of the Pisacas. It therefore can claim no historical

value whatever.1

78. Cultivation such as appears to have been carried on in Kasmir

since the earliest historical period, must neces¬

sarily leave its traces in the topography of a

country and may hence claim a passing notice.

Rice has as far as we can go back, always been the largest and

most important produce of the Valley. Its character as the main cereal

is sufficiently emphasized by the fact that it is usually referred to in

the Chronicles by the simple term of dhanya 1 grain.’2 The conditions

of its cultivation presuppose an extensive system of irrigation, and for

this the Kasmir Valley with its abundance of streams and springs is

admirably adapted by nature. The elaborate arrangements which

exist at present for taking water from the streams large and small and

distributing it over all the ground capable of irrigation, will be found

fully detailed in Mr. Lawrence’s valuable and exhaustive account of

Kasmir agriculture.3 There is every reason to believe that they have

come down with little, if any, change from a very early period.

Many of the larger irrigation channels which intersect the fertile

Irrigation alluvial flats, or skirt the terraced slopes of

the Udars and mountain-sides, are shown on

the map ; see, e.g., the tracts on the lower course of the Lidar, Vesau,

Sind, and other rivers. In old times when the population was larger

than now, much land which is at present allowed to lie waste on the

liill-sides, on the Udars and in the low-lying tracts by the marshes,

must have been under cultivation.4 I have often come across traces of

old irrigation-cuts long ago abandoned, which brought down the water

of the melting snows from alpine plateaus high above the forest zone.

Their distance from any lands capable of rice-cultivation is so great

1 See i. 180, and note i. 184.

2 “ The Kashmiris, so far, have considered no crop worthy of attention save

rice;” Lawrence, Valley , p. 319.

3 See Valley , pp. 323 sq.

4 Compare Valley, pp. 239 and 356, as to the extensive areas which were once

cultivated and are likely to be so again in future.

1899.]

SOIL AND CLIMATE OF THE VALLEY.

121

and the trouble of their construction must have been so considerable

that only a far greater demand for irrigation than the present one can

account for their existence.

In the earliest traditions recorded by Kalhana the construction of

irrigation canals plays already a significant part. The Suvarnamcini-

hulya which is ascribed to King Suvarna and which still brings water

to a great part of the AdSwin district, has already been noticed.1 The

reference to the aqueduct by which King Damodara is supj)Osed to have

attempted to bring water to the great Udar named after him, though

legendary in the main, is also characteristic.2 Lalitaditya is credited

with having supplied villages near Cakradhara (Tsakadar) with the

means of irrigation by the construction of a series of water-wheels

(araghatta) which raised the water of the Vitasta.3

To Suyya, however, Avantivarman’s engineer, is ascribed the

merit of having on an extensive scale secured river-water for village-

lands. From .Kalhana’s detailed description it is evident that Suyya’s

regulation of the Vitasta was accompanied by systematic arrangements

for the construction of irrigation channels. For these the water of

various hill-streams was utilized as well as that of the main-river. The

size and distribution of the water-course for each village was fixed on a

permanent basis. He is thus said “ to have embellished all regions

with an abundance of irrigated fields which were distinguished for ex¬

cellent produce.” The increase in produce consequent on these measures

and the reclamation of new lands from the river and marshes is said to

have lowered the average price of a Khari of rice from two hundred to

thirty-six Dinnaras.4

The importance of irrigation from a revenue point of view must

have always been recognized by the rulers of the country. Hence even

in later times we find every respite from internal troubles marked by

repairs of ancient canals or the construction of new ones. The long

and peaceful reign of Zainu-l-‘abidin which in many respects revived

the traditions of the earlier Hindu rule, seems in particular to have

been productive of important irrigation works. Jonaraja’s and S'rivara’s

Chronicles give a considerable list of canals constructed under this

king.6 Among these the canal which distributed the water of the

Pohur River over the Zain^gir Pargana, and the one by which the

1 See above, § 64.

2 See Rdjat. i. 156 sq. note.

8 See Rdjat. iv. 191 note.

4 See Rdjat. v. 109-112 and note.

6 See Jonar. (Bo. ed.) 1141-55, 1257 sqq. ; S'rtv. i. 414 sqq. For repairs of old

canals, see Rdjat. viii. 2380.

J. i. 16

122

ANCIENT GEOGRAPHY OF KAS'MlK. [Extra No. 2,

Saffron-cultivation.

water of tlie Lidar was conducted to t-lie arid plateau of Martanda,

deserve special mention. In the latter locality some work of this kind

must have existed already at a far earlier period. Or else we could

scarcely understand how it could have been chosen as the site for

Lalitaditya’s magnificent temple and tli e flourishing township which

once surrounded it.1

Of the other products of the Valley only two may be mentioned

here, since they have from old times received

special attention in all descriptions of the

country. Already Kalhana in his introduction designates snffron and

grapes among “the things that even in heaven are difficult to find but

are common there.”2 Snffron ( hunkuma ) has to the present day re¬

mained a famous product of Kasmir. Its cultivation has apparently

from an early time specially flourished about Padmapura, the present

Pampar, where the Udar lands are still chiefly utilized for it. The

Fourth Chronicle describes at length the plant and its treatment.

Abii-l-Fazl mentions it also in the same locality and devotes to it a long

notice.3

The grapes of Kasmir which Kalhana mentions repeatedly,4 * have

not retained their area of cultivation with

equal persistence. They must have enjoyed

reputation outside Kasmir, because the name Kasmira is given by

Sanskrit Kosas as the designation of a special variety of grapes.6 They

were once plentiful at Martanda where both Kalhana and the Fourth

Chronicle mention them, and at many other localities.6

In Akbar’s time grapes were abundant in Kasmir aud very cheap ;

but Abu-1-Fazl notes that the finer qualities were rare.7 Since then,

viticulture among the people generally has greatly declined. Though

vines of remarkable size and age can still be found in many places, they

are mostly wild. The produce of grapes is now restricted to a few old

gardens at the mouth of the Sind Valley and to the new vineyards

established on the Dal shores by the late Maharaja for the cultivation

of French viues.8

Grapes.

1 See Rajat. iv. 192.

2 i. 42.

s See Fourth Chron. 926 sqq. ; Aln-i-ATcb ., i. pp. 357 sq.

4 Rajat. i. 42 ; iv. 192 ; vii. 498.

6 See Bdhtlingfc-Roth, s. v.

6 Fourth Chron. 851, 928.

7 Ain-i-Alcb., i. p. 349.

8 For a detailed account of Kasmir vineyards, see Lawrence, Valley , pp. 351 sq.

1809.]

SOIL AND CLIMATE OF THE VALLEY.

123

79. It will be useful to refer liere briefly to the data we possess

Old ethnography regarding the old ethnography of Kasmir and

of Kasmir. the adjacent hill regions.

As far as Kasmir itself is concerned our information does not allow

us to connect any particular localities with ethnic divisions. Judging

h-om Kalhana’s Chronicle and what other sources of information-are

available to us, the population of Kasmir has shown already in old times

the same homogeneity that it does at present. The physical aud ethnic

characteristics which so sharply mark off the Kasmlrl from all surround¬

ing races, have always struck observant visitors to the Valley and have

hence often been described.1 Hiuen Tsiang’s brief sketch reproduced

above is the earliest in date and ye t applies closely to the modern

inhabitants.

That the Kasmlris form a branch of the race which brought the

languages of the Indo- Aryan type into India, is a fact established by the

evidence of their language and physical appearance. But when their

settlement in the country took place, and from which direction they

immigrated, are questions beyond the present range of historical

research. The purity of race which has often been noted as distinguish¬

ing the great mass of the Kasmir population, may be admitted with a

qualification. It is probably due not only to the country’s natural

isolation but also to a curious faculty for absorbing foreign elements.

Colonies of Mughals, Pathans, Panjabis, and Paliarls, settled within

comparatively recent times in the Valley, are being amalgamated with

remarkable rapidity through intermarriage and other means.

The complete absorption of these settlements which is going on

under our own eyes as it were, furnishes a

likely analogy for the ethnic history of earlier

times. We have reason to assume that Kasmir

has also in Hindu times been often under

foreign rule. It is difficult to believe that the reign of foreign dynasties

has not been accompanied also by settlements of immigrants of the same

nationality. But it is not likely that these foreign colonies were ever

extensive. In any case we find no trace of their having retained a

distinct and independent existence.

Various tribal sections of the population are mentioned in Kalhana’s

narrative, but we have no means of deciding to what extent they were

based on race or caste distinctions. The names of the Lavanycis and

Tantrins survive in ‘ Krams,’ or tribal names, still borue by sections of

Absorption of

foreign ethnic

elements.

1 For a general account of the Kasmlrl population Drew’s remarks, Jummoo ,

pp. 174 sqq., may still be recommended. Fuller details regarding the various

classes, castes, etc., will be fouud in Mr. Lawrence’s work, pp. 302 sqq.

124

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

the Muhammadan rural population ( Lun 1 and Tdntr *)d But whatever

distinctions of race or caste may have originally been indicated by these

‘ Krams,’ they have long ago disappeared.

It is equally certain from an examination of the Chronicle that

these sections were never confined to particular territorial divisions, but

spread all over the Valley. The humblest of these sections is probably

the one which has least changed its character during the course of cen¬

turies. The modern Dumbs, the descendants of the old Dombas ,2 are

still the low-caste watchmen and village-menials as which they figure

in Kalhana’s narrative. They, like the still more despised Vatals or

scavengers, cannot intermarry with other Kasim ids. They have thus

retained in their appearance a distinctive type of their own which

points to relationship with the gipsy-tribes of India and Europe.

It is difficult to come to any definite conclusion as regards the

Ki-li-to whom Hiuen Tsiang mentions as a low-born race settled in

Kasmir from early times and opposed to the Bauddhas.3 Their name,

usually transcribed Kritiya , cannot be traced in indigenous records.

There is nothing to support their identification with the Kiras, as

suggested by General Cunningham.4 The latter seems to have been a

tribe settled somewhere in the vicinity of Kasmir.5

80. The ethnography of the territories immediately adjoining

Baees on Kasmir Kasmir can be traced quite clearly from the

borders. notices of the Rajataraiiginl.

In the south and west the adjacent hill-regions were occupied by

Khasas. Their settlements extended, as shown by numerous passages

of the Chronicle, in a wide semi-circle from Kcistavdr in the south-east

to the Vitasta Valley in the west.6 The hill-states of Rajapuri and

Lohara were held by Khasa families ; the dynasty of the latter territory

succeeded to the rule of Kasmir in the 11th century. I have shown

elsewhere that the Khasas are identical with the present Khakha tribe

to which most of the petty chiefs in the Vitasta Valley below Kasmir

and in the neighbouring hills belong. We have already seen that the

1 Compare notes v. 248; vii. 1171.

2 See Raj at. note iv. 475; also v. 353 sqq., vi. 84, 182; vii. 964, 1133, viii. 94.

These passages show that the Dombas also earned their bread as hunters, fishermen,

buffoons, quacks, etc,, and their daughters as singers and dancers. Their occupa¬

tions thus closely resembled those of the gipsies whose name, Rom , is undoubtedly

derived from Skr. domba ; see P. W. s. v.

8 See Si-yu-ki, transl. Beal, i. pp. 150, 156 sqq.

4 See Anc. Geogr., p. 93.

6 Compare my note viii. 2767.

6 See Iidjat, i. 317 note.

1899.]

SOIL AND CLIMATE OF THE VALLEY.

125

Kliaklias have until very recent times worthily maintained the reputa¬

tion which their forefathers enjoyed as marauders and turbulent hill-

men.

North of the Vitasta Valley and as far as the Kisanganga we now

find the Bombas as the neighbours of the Khakhas to whom they are

closely related. It is probable that the Karnav district was held by

them already in old times. Kalhana seems to comprise them, viii. 3088,

under the designation of Khasa.

The upper Kisanganga Valley above S'ardi wa§ in old days already

as at present inhabited by Dards (Skr. Darad , Daradct ) who are often

referred to by Kalhana as the neighbours of Kasmir on the north.1

Their seats extended then too probably much further to ilie north-west,

where they are now found in Citral, Yasin, Grilgit and the interven¬

ing regions towards Kasmir. Megasthenes already knew them in the

Upper Indus regions. Kalhana relating events of his own time speaks

of Mlecchas further to the north. These might have been Muham-

madanized Dards on the Indus, and beyond.2

The regions immediately to the north-east and east of Kasmir were

held by the Bhauttas. We have already seen that these represent the

people of Tibetan descent, the modern But?, of Dras, Ladakh and the

neighbouring mountain districts.3

1 See Raj at. i. 317 note.

2 See note viii. 2762-64.

8 See above, § 58.

126

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

CHAPTER IY.

POLITICAL TOPOGRAPHY.

Section I. — Frontiers oe ancient Kasmir.

81. Our account of tlie political topography of ancient Knsmir

may conveniently open with a survey of its frontiers. These agree so

closely with the natural boundaries of the Valley that we have already

had occasion to trace them when dealing with the mountain ranges

enclosing the latter. It will however be useful to supplement our

information regarding these frontiers by a brief notice of the territories

which lay beyond them and formed the neighbours of the Kasmir

kingdom in Hindu times.

Beginning in the south-east we have first the Valley of Kas-

thavata, the present Kastavar (‘ Kishtwar ’ of

the maps) on the upper Cinab. It is mention¬

ed by Kalliana as a separate hill state in the

time of Kalasa.1 Its Rajas who were Hindus till Aurangzeb’s time,

practically retained their independence until the conquest of their terri¬

tory by Maharaja Gulab Singh.

The hill-district of Bhadravah lower down on the Cinab is once

named in the Rajatarangini as Bhadrdvalcasa .3 Its Rajas were tributary

to Camba in recent centuries. This was probably the case also in

earlier times as we do not find a ruler of Bliadravakasa referred to in

Kalhana’s lists of hill Rajas.

Territories S. E. of

Kasmir.

1 See Rdjat, vii. 590 note.

* See Rdjat. viii. 501 note.

1899.]

FRONTIERS OF ANCIENT KASHMIR.

127

The Rajas of Camba, the ancient Campa, on the other hand figure

often in the Kasmir Chronicle.1 Their territory has since early times

comprised the valleys of the sources of the Ravi between Kangra, the

ancient Trigarta, and Kasthavata. The ancient Rajput family which

rules this hill state to the present day, often intermarried with the

Lohara dynasty which reigned in Kasmir.

To the west of Campa and south of Bhadravakasa lay the chiefship

of Vallapura, the modern Ballavar.2 3 Its rulers are repeatedly referred

to in Kalhana’s narrative and retained their independence as petty hill-

cliiefs till the rise of the Jammu family early iti this century. 4 * Balla-

war ’ was known also to Alberuni.

Of the political organization of the hill- territories between Valla¬

pura in the south-east and Rajapuri in the north-west we have no

distinct information. The Hindu inhabitants of this tract including

Ballavar call themselves now Dogrds and their country JDugar, This

name is traditionally derived from Skr. *Dvigarta,z but this term is

nowhere found in our historical texts and has probably been created for

the sake of an etymology in analogy of the ancient Trigarta. The

original of the name seems to be Durgara A

It is \ery probable that the region of the lower and middle hills

between the limits indicated was already in old times divided into a

number of small cliiefships. Of these some eleven seem to have existed

up to the time of the extension of the Sikh power into the Panjab

Kohistan.6 They were all absorbed in the growing state of Jammu

which was originally one of them.

Among these small hill-chiefs of limited territory but ancient

descent, we have probably to class the Thakkura Dengapala on the Cinab

who gave his daughter to the pretender Bhiksdcara in marriage.6 Also

the Raja of Kanda must probably be located in this hill tract.7 Other

Thakkuras in this region are mentioned as levying blackmail on Prince

Mallarjuna when on his march to Kasmir from the plains.8 Immediately

at the foot of the Banahal Pass in the territory of Visalata we find the

castle of a ‘ Khasa lord,’ who gave shelter to Bhiksacara and at the time

l Compare Rajdt. vii. 218 note, and Cunningham, Anc. Geogr., p. 141.

8 See Rdjat. vii. 220 note, and Cun'ningham, Anc. Geogr. p. 135.

3 See Drew, Jummoo, pp. 43 sq.

4 Compare the Camba copperplate, edited by Prof. Kielhorn, Ind. Ant., 1888,

p. 9.

3 See Cunningham, Anc. Geogr. pp. 133 sqq., where a useful synopsis of the

hill-states in the central portion of the Panjab Kdhistan is given.

6 See Rdjat. viii. 554 sqq.

1 See note vii. 590.

8 viii- 1989 sqq.

128

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

Frontier territories

to the south-west

and west.

was evidently independent.1 Temporarily tlie Rliasas of the hills imme¬

diately south of the Pir Pautsal Range may have acknowledged the

suzerainty of strong Kasmir rulers. But during the greatest part of the

period which is known to us from historical sources, they appear to have

held their own and rather to have levied subsidies, i.e., blackmail from

the Kasmir rulers.2

82. Some of the petty hill states here referred to must have been

included in the region which by its ancient

name was known as Darvabhisara. I have

elsewhere shown that this name, as a geo¬

graphical term, was applied to the whole tract

of the lower and middle hills between the Candrabhaga and Vitasta.3

The combined names of the Darvas and Abhisaras are found already in

the ethnographical lists of the Ivlaliabharata and Brhatsamhita. A

chief of this region figures by the ethnic appellation of Abisares in the

accounts of Alexander’s Indian campaign.

The most important of the hill-states in this territory was certainly

the ancient RajapurI represented by the modern district of Rajaurl.4 *

It comprised the valleys drained by the Toll! of Rajaurl and its tribu¬

taries. Owing to its position on the most direct route to the Pan jab,

RajapurI was necessarily often brought into political relations with

Kasmir. When Hiuen Tsiang passed through it, the ‘ kingdom of

RajapurI ’ was subject to Kasmir. From the 10th century onwards we

find the chiefs of RajapurI as practically independent rulers, though the

Chronicle tells us of numerous expeditions undertaken into their terri¬

tory by the later Kasmir kings. The upper valley of the Toll! of Prunts

leading to the Pir Pantsal Pass, was included in RajapurI terri¬

tory.6 Here lay probably the famous strong-hold of Rajayiri known also

to Alberuni.6

RajapurI took its name from its capital which is repeatedly

mentioned by Kalhana and undoubtedly occupied the position of the

present town of Rajaurl.7 The ruling family belonged to the Khasa

tribe. Its descendants were the Muhammadanized Rajput chiefs who

retained this territory down to the present century.

On the north-west RajapurI was adjoined by the territory of

1 viii. 1665 sqq.

2 See Rdjat. viii. 2283 note.

8 See note i. 180.

4 For a detailed account, see Rdjat. vi. 286 note,

6 See Rdjat. viii. 959 note.

8 See vii. 1270 note.

7 See vii. 973 sqq.

1899.]

FRONTIERS OF ANCIENT KASHMIR.

129

Lohara.1 2 The chief valley belonging to this hill-state was the present

Loh?rin which we have already visited when examining the Tos^maidan

route. Lohara became important for Kasmir from the end of the 11th

century when a branch of its ruling family acquired the Kasmir throne.

Subsequently this branch succeeded also to Lohara which thus became

united to Kasmir under the same ruler. As the ancestral home and

stronghold of the dynasty, the castle of Lohara has played a great part

during the last reigns related by Kalhana. The chiefs of Lohara are

distinctly named as belonging to the Kliasa tribe.

Lohara seems to have included in those times also the town and

district of Parnotsa corresponding to the present Punch or Prunts

(the Kasmlri form), in the lower valley of the Tohi (Tausi).8 In Hiuen

Tsiang’s time Parnotsa gave its name to the whole hill-state which was

then tributary to Kasmir. The Muhammadan Rajas of Prunts, closely

related to the Khakhas of the Vitasta Valley, remained more or less

independent till the conquest of Maharaja Grulab Singh. Their terri¬

tory forms now a separate small principality under a branch of the

Jammu family. Parnotsa being on the great route to the western

Panjab is often mentioned in the Kasmir Chronicles. The large per¬

centage of the Kasmlri element in the population of Prunts attests the

closeness and ancient date of its relations to Kasmir.

The hills to the south-west of Prunts were held till early in this

century by petty chiefs, known as the Rajas of Kotli. It is possible

that the small hill-state of Kalinjara repeatedly referred to by Kalhana

and known also to Ferishta, lay in this direction.3 *

Proceeding to the north-west of Parnotsa we come to the valley of

the Vitasta. This, as has already been shown above, was held in old

times as an outlying frontier-district of Kasmir as far down as Bolya-

saka, the present Buliasa. Beyond this point it was occupied by

Khasas. In Muhammadan times the valley was divided between several

petty chiefs of the Kliaklia and Bomba clans who seem to have acknow¬

ledged as their nominal head the Khakha Raja of Muzaffarabad. The

portion of the valley between Muzaffarabad and Buliasa bore the

old name of Dvaravati from which the modern designation of this tract,

Dvarbidl , is derived (see above, § 53).

1 Compare for the history of Lohara and its various localities, Note E, Rdjat. iv.

177, reproduced in Ind. Ant., 1897, pp. 225 sqq.

2 See for details note iv. 18. Hiuen Tsiang’s reference shows that the town of

Parnotsa must be older than the time of Lalitaditya to whom Kalhana ascribes its

foundation.

3 See note Rdjat. vii, 1256.

J. i. 17

130

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

Urasa-Hazara.

83. Further to the west and beyond the course of the Vitasta

after its great bend, lay the ancient kingdom

of Urasa.1 2 Its greatest part is comprised in

the British district of Hazara, between the Vitasta and Indus. It is

the Ovapaa or "Apcra of Ptolemy; its ruler figures as Arsakes in the

accounts of Alexander’s campaigns. Hiuen Tsiang mentions the terri¬

tory by the name of Wu-la-shi and found it tributary to Kasmir.

Though this dependence seems soon to have ceased we find Urasa often

referred to in the Rajataraiigini. The account of S'arhkaravarman’s

ill-fated expedition in this direction furnishes us with a cine as to the

position of the old capital of Urasa. It probably lay between the

present Mansahra and Abbottabad.3 Kalhana’s notice of an expedition

undertaken in his own time mentions in Urasa the town of Atyugra-

pura.3 I have shown in my note on the passage that this locality is

probably represented by the modern Agror , situated on the border of

Hazara towards the ‘Black Mountains.’ We have an intermediary

form of the name in Ptolemy’s T#dyorpo<>, given as the designation of

a town in Uarsa or Arsa north of Taxila.

In Muhammadan times Urasa was included in the region known as

Pakhli. This is defined by Abu-l-Fazl as comprising the whole of the

hill territory between Kasmir in the east and the Indus on the west.4 * To

Pakhli belonged also the lower valley of the Kisanganga and the

valleys of the streams which flow into the latter from the Kajnag Range

and the mountains to the north-west of Kasmir.

This tract which is now known as Karnau , bore the old name of

Kisanganga Valley. Karnaha.. It seems to have been held by

small chiefs nominally tributary to Kasmir

even in later Hindu times.6 It is but rarely mentioned in the Chronicle.

The inhabitants were Khasas,6 who are represented by the modern

Bomba clans still holding Karnau. Their Rajas were practically

independent till the Sikh conquest and often harried the north-western

parts of Kasmir.7 The last irruption of the Karnau Bombas and their

allies, the Khakha chiefs of the Vitasta Valley, occurred as late as

1846.

1 For a detailed synopsis of the old notices, see Raj at. v. 217 note.

2 See Rdjat. v. 217 note and Cunningham, Anc. Geogr., p. 104.

3 Compare note viii. 3402.

4 See Ain-i-Akb., ii. pp. 390 sq.

6 Compare Rdjat. viii. 2485 note.

8 See viii. 2756, 3006, 3088.

7 Compare for the modern Karnau, Bates, Gazetteer, p. 228.

1899.]

FRONTIERS OF ANCIENT KAS'MIR.

131

Darad. territory.

The valley of the Kisangaiiga above its junction with the Karnau

river and as far as Sitrdi, forms a separate tract known as Drava. This

is possibly tlie Duranda mentioned in a passage of Kalhana’s Chronicle.1

The northernmost portion of the tract seems to have been a dependency

of Kasmir even during the later Hindu reigns. At S'ardi we find the

shrine of S'arada, one of the most sacred Tirthas of old Kasmir. To

this as well as an old feudal stronghold in its neighbourhood we shall

have occasion to refer thereafter (§ 127).

Through S'ardi leads a route to Cilas on the Indus. But this

territory as well as the other portions of the Upper Indus Valley lay

apparently quite beyond the sphere of Kasmir political influence. Hence

we meet nowhere in the Chronicles with their ancient names.

84. Immediately above S'ardi the valley of the Kisangaiiga turns,

as we have seen, into a narrow uninhabited

gorge. At the other end of this gorge we

reach the territory of the Darads. Their settlements on the Upper

Kisangahga and its tributaries seem to have formed a separate little

kingdom, called by a general name Daraddesa in the Chronicle.2

Its inhabitants who bore Hindu names, more than once attempted inva¬

sions of Kasmir. DaratpurI, ‘ the town of the Darads,’ which was the

capital of their chiefs, may have occupied the position of the modern

Gurez (map ‘ Goorais’).3 The latter is the chief place of the valley

where the Nawabs governing it till the Sikh conquest resided. The

‘Mleccha’ chiefs who on two occasions figure as the Darad Rajas’ allies

from the north, were perhaps rulers of other Darad tribes further

towards the Indus who had early been converted to Islam.4

Crossing from the head- waters of the Kisangaiiga to those of the

Dras River we arrive in high-level valleys

inhabited by people of Tibetan race and

language, the Bhauttas of the Chronicles. The Rajatarariginl tells us

nothing of the political organization or topography of the Bhautta

territories. It is, however, possible that we have a reference to Leh ,

the capital of Ladakh, in the “ foreign country called Loh,” which

Kalhana names in iii. 10.

Nor do the later Chronicles supply us with any details in this

direction, though the several invasions which Kasmir suffered from

this side give Jonaraja and S'rlvara occasion to refer more frequently to

the Bhauttas and their rulers. It may, however be noted that S'rlvara

Bhauttas.

1 See viii. 2709 note.

2 Compare Rdjat. vii. 911 ; for other references to the Darads, i. 312 note.

8 See vii. 911 note.

4 See viii. 2762 note.

132

ANCIENT GEOGRAPHY OF KASAI IR. [Extra No. 2,

already knows the terms ‘ Little and Great Bhutfa-land.’1 They refer to

Baltistan (Skardo) and Ladakh which have continued to be known to

the present day as ‘ Little and Great Tibet,’ or among Kasmiris as Lukh

Butun and Bud Butun .2 * These terms are in fact of a far older date,

as they are found already in the Chinese Annals as Little and Great

Poliu .s

The eastern frontier of Kasmir is, as we have seen, formed by a

mountain range which runs from the Z5ji-La almost due south towards

Kastavar. Along this range on the east lies a long narrow valley

marked as Maru-Wardwan on the map (in Kasmirl Madivadvan) . It

is drained by a large river which joins the Cinab near the town of

Kastavar. Owing to its high elevation and the rigours of its climate it

is inhabited only by a scanty population. According to Mr. Drew’s race

map and other authorities, this consists now chiefly of Kasmiris.

Whether this was already the case in old times, is uncertain. The

Valley is nowhere mentioned in our old Kasrairian texts.4 * It is hence

doubtful whether it belonged to Kasmir territory in Hindu times. Yet

Abu-1-Fazl counts it among the Parganas of Kasmir.6 Beyond it to

the east stretches an uninhabited belt of high mountains and glaciers,

dividing Madivadvan from the Tibetan tracts of Suru and Zanskar. To

the south we reach once more the territory of Kastliavata from which

our present survey has started.

1 See S'nv. iii. 445 ( Suksmabrhadbhuttadesau ).

2 Butun (connected with the ethnic term But <? < Bhautta; see above, § 58), is

the Kasmirl term for Tibet in general.

s Compare A. Remusat, Nouveaux melanges asiatiques, i. p. 194; and Sir H.

Yule, Cathay, p. lxx.

4 The Trisamdhyamahatmya which refers to the Valley as Madavatlra , cannot

claim any particular antiquity.

6 See Aln-i-Akb., ii. p. 369.

1899.]

ANCIENT POLITICAL DIVISIONS.

133

Section II. — Ancient Political Divisions.

85. The Valley of Kasmir to which we may now return has

from early times been divided into two great

Kramarajya, Mada- parts, known by their modern names as

varajya. Kamrdz and Maraz. These terms are derived

from Skr. Kramarajya and Madavarajya, which are found very

frequently in the Rajatarahginl as well as the later Chronicles.1 The

original form of the modern Kamraz was known to the tradition of the

Shin agar Pandits generally. With the old name Madavarajya, however,

I found only those few acquainted who, like the late Pandit Damodara

and Pandit Govind Kaul, had specially studied Kalhana’s Chronicle.

According to the generally prevailing notion Maraz comprises the

districts on both sides of the Vitasta above STinagar, and Kamraz those

below. The present tradition places the boundary of the two great divi¬

sions more accurately at the Shergarhi palace. That the boundary was

already in old times indicated by a line drawn through the capital is

easily proved by an examination of all passages in the Rajatarahginl

and other Chronicles naming Madavarajya and Kramarajya. They

invariably show localities situated above STinagar in the former and

those below in the latter division.

We arrive at the same result on a reference to the Ain-i Akbari.

Abu-1-Fazl distinctly informs us that “ the whole kingdom was divided

under its aucient rulers into two divisions, Mardj on the east and Eamrdj

on the west.”2 He then proceeds to tabulate the thirty-eight Parganas

into which Kasmir was divided under Akbar’s administration, separately

under the two main-heads of Maraj and Kamraj. The city of STinagar

is counted with the former, and so are also all Parganas above the capital,

while those below are shown in Kamraj.

The term of Kamraz has in modern times occasionally been used

also in a more restricted sense, for the designation of the Parganas to

the west and north-west of the Volur lake. This usage probably arose

from the fact that at various periods several of the small Parganas in

this portion of the Valley were for administrative purposes grouped

together in one Pargana, to which the name Kamraz was given.3 This

1 See my note on j Raj at. ii. 15.

2 Compare Ain-i- Alcb., ii. p. 368.

& Thus Abu-l-Fazl’s table seems to show that in Akbar’s time the old Parganas

of Uttar, Lolau, Hamal and Machipur were embodied in the large Pargana of

‘ Kamraj see Ain-i Akb., ii. p. 371. In Moorcroft’s and Baron Hiigel’s list the Par¬

gana Kamraj includes Uttar, Ilamal and Machipur. Owing to the frequent changes

134

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

circumstance explains the different accounts referred to by Prof. Biihler

in his note on the term Kramarajya.1

Though the terms Madavarajya and Kramarajya are so often

employed in the Chronicles, we have no distinct evidence of the two

divisions having in Hindu times formed separate administrative units

or provinces. It is possible that this was the case at one or the other

period. But Abu-l-Fazl’s account as well as the usage traceable from

his time to the present day show that the terms in their popular geo¬

graphical significance could maintain themselves quite independently

of actual administrative divisions.2

86. The whole of the Valley has from an early date been sub¬

divided for administrative purposes into a

considerable number of small districts known

in recent times as ‘ Parganas.’ Their ancient

designation was visaya .3 The number, names and limits of these sub¬

divisions have been subject to considerable variations during the period

over which our documents extend.

The great majority of the Parganas known in recent times can be

safely assumed to have existed already during the Hindu rule. This is

proved by the fact that the names of numerous Parganas are found in

their ancient forms already in the Rajatarangini and the other Chroni¬

cles. But these texts do not furnish us anywhere with a complete list

of the Parganas. It is hence impossible for us to restore in full detail

the map of the administrative sub-divisions for any particular epoch

Administrative

Districts.

of the Pargana divisions (see below) the extent of the ‘ Pargana Kamraj ’ has also

varied from time to time.

1 See Report, p. 11.

2 The only trace I can find of a general division of Kasmlr other than that into

Madavarajya and Kramarajya, is contained in an unfortunately corrupt and fragmen¬

tary passage of the Lokaprakasa, iv. It seems to divide the twenty-seven Visayas

or Parganas of Kasmlr (see below) into three tracts, viz. (i) Kramarajya from

Khdyasramika onwards (Khuy4hom, the old Khuyasrama is meant) ; (ii) Madhyama •

rdjya from the Canula [river ?] to Lahara or Lar ; (iii) Madavarajya from

S'rivantaha ( ?).

The text is in a deplorable condition and the explanation of Canula and S'rivan-

taka quite uncertain. The former may be the river of doubtful name and identity

referred to in Rajat. note v. 109. It appears as if at the time to which the Loka-

prakasa’s notice goes back, an intermediate slice of territory had been formed

between Kramarajya and Madavarajya and dubbed Madhyamarajya ‘ the middle

province.’ Five thousand villages out of the 66,063 with which the text credits

Kasmlr, are attributed to this intermediate division.

& Compare for the term visaya Rajat. v. 51 ; viii. 1260, 1413, 2697.

The expression Pargana may have been introduced by the Mughal administra¬

tion. Its Skr. original * puragana is not found in the Chronicles.

ANCIENT POLITICAL DIVISIONS.

135

1899.]

during Hindu times. The Lokaprakasa, it is true, tells us of the

division of Kasmir into twenty-seven Visayas and enumerates some

nineteen of the latter. But several of the names are so corrupt as to be

beyond recognition, while others bear a distinctly modern look. In any

case it is impossible to fix the date to which this notice may belong or

to judge of its authenticity.1

Abu-l-Fazl’s account is the first which presents us with a systematic

statement of Kasmir Parganas. It is of special interest because it

shows us how their list could be increased, or re-adjusted within certain

limits according to fiscal requirements or administrative fancies. The

return of Asaf Klian reproduced by Abu-1-Fazl shows thirty-eight

Parganas, while the earlier one of Qazi ‘All contained forty-one. The

difference is accounted for by the amalgamation of some and the splitting-

up of other Parganas. The Parganas varied greatly in size, as shown

by the striking contrasts in the revenue-assessments. Thus, e.g , Patan

was assessed at circ. 5300 Kharwars, while the revenue from ‘ Kamraj ’

amounted to 446,500 Kharwars.

The number of Parganas had changed but little during Mughal and

Pathan times. For the Sikhs on their conquest of the V alley seem to

have found thirty-six as the accepted traditional number. But there

had been various changes in the names and extent of these Parganas.

These changes became still more frequent under the Sikh administra¬

tion, as is seen by a comparison of the lists given by Moorecroft (1823),

Baron Hiigel (1835) and Yigne (circ. 1840). They all show a total

of thirty-six Parganas but vary among themselves in the names of

individual Parganas.

These frequeut changes and redistributions of the Parganas conti¬

nued during JDogra Rule. The most accurate list I am able to refer to

for this most recent period, is that given by Major Bates. It shows a

total of forty-three Parganas for the year 1865.2 Subsequent reforms

introduced Tahsils after the fashion of British provinces with a view to

reducing the number of sub-divisions. The latest list shows eleven

Tahsils.3 In their constitution little regard was paid to the historical

divisions of the country. Fortunately, however, Kasmirls are as con-

1 Of the Lokaprakasa’s Visayas Khoydsrami, S'amala , Lahari, Auladlya , Nlldsa ,

Khaduviya correspond clearly to the Khfiyasrama, S'amala, Lahara, Holada, Nllasva,

Khaduvl of the Rajataranginl. Ekena, Devasiivi may possibly be corruptions for

Evenaka and Devasarasa. Krodhana, Dvavimsati, Bhrnga , Phdgvd probably repre¬

sent the modern Parganas of Kruhin, Dunts, Bring, Phakh. Cdlana , Vitasthd,

Satrava, Svanavdri , Nzld, Hart , Jalahadiya, are quite uncertain.

2 See Gazetteer , p. 2 sqq.

3 Compare the sketch-map attached to Mr. Lawrence’s Valley.

136

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

servative in tlieir topographical nomenclature as in many other matters.

The old Pargana names are hence still in ordinary use and likely

to remain so for some time to come.1 2

The absence of a complete list of Parganas for an earlier period

and the changes in their constitution during more recent times make a

systematic exposition of the ancient territorial divisions impracticable.

In a separate note I have given a comparative table of the Pargana lists

we possess since Akbar’s time. There too I have indicated the ancient

equivalents of the Pargana names, as far as they can be traced in the

Sanskrit Chronicles.3 We shall have occasion to refer to these names

and their history in the course of our detailed survey of ancient locali¬

ties in the Valley.

87. The large number of administrative sub-divisions which as

we have seen goes back to an early date, may

Density of popula- u £ , ,

.. . , - be taken as an indication or the dense popu-

tion in old Kasmir. . 11

lation then occupying the Valley. We have

no means of forming any accurate estimate as to the number of the

population which the country contained in Hindu times. But there is

every reason to believe that even at a later period it was far larger than

at the present day. The existence of a very great number of deserted

village- sites, in all parts of the country, the remains already alluded to

of a far more extended system of irrigation, the number of great temple

ruins, and the uniform tradition of the people, — all point to the same

conclusion.

The present century has witnessed in Kasmir a series of appalling

famines and epidemics, which wrought terrible havoc in the mass of the

rural population particularly. The last famine, 1878-79, alone is

supposed to have removed three-fifth of the population from the Valley.3

The political vicissitudes of the first half of the century had a baneful

influence on the economical condition of Kasmir and brought about

an extensive emigration both among the industrial and agriculturist

classes. Notwithstanding all these trials the population which in 1835

was estimated at about 200,000 souls, had risen to 814,000 according to

the census of 1891.

These figures indicate great powers of recuperation. Yet it is held

by competent judges that the present agricultural population is by no

means sufficient even for the land actually under cultivation. It would,

hence manifestly be hazardous to make any guess as to the numbers

1 The Survey of India maps indicates the approximate extent of the Parganas

recognized in the fifties.

2 See Supplementary Note BB.

s Compare for this and other statistical details Mr. Lawrence’s Valley, p. 223 sqq,

1899.]

THE OLD AND NEW CAPITALS.

137

which the country might have supported in the most prosperous times

of Hindu rule.

The fact of Kasmir having possessed a far greater population in

ancient times helps to explain the curious traditional verse which puts

the number of villages of Kasmir at 66,063. The verse is found twice

in the Lokaprakasa and still lives in the oral tradition of the Brahmans

throughout the Valley. It has been reproduced from the latter in

Pandit Sahibram’s Tirthasamgraha.1 2 That it can claim some antiquity

is evident from the allusion made to the number in Jonaraja’s Chronicle.*

Though that figure must have at all times implied a considerable ex¬

aggeration, it is nevertheless characteristic of the popular notion on the

subject. Even Sharifu-d-din whose information, collected about a.d. 1400,

is generally accurate and matter-of-fact, records : “ It is popularly

believed that in the whole of the province — plains and mountains

together — are comprised 100,000 villages. The land is thickly inhabit¬

ed.” 3 It is curious that Mirza Haidar who had ruled Kasmir himself

copies this statement without modification or dissent.

Section III. — The old and new Capitals.

88. The ancient divisions of Kramarajya and Madavarajya are

separated by a line drawn through S'rinagar. This fact as well as the

great historical interest attaching to S'rinagar as the capital of the

country make it the convenient starting-point for our survey. The

history of Kasmir has always been reflected as it were in that of its

capital. The site of the latter has not changed for more than

thirteen centuries. It is thus easy to account for the ample historical

data which enable us to restore in great part the ancient topography of

S'rinagar and to trace back the city’s history to the time of its foundation.

Hiuen Tsiang who visited the Kasmir capital about a.d. 631, and

whose record is the earliest we possess, found

it already in the position of the present S'ri¬

nagar. He describes it as situated along the

Srinagara in the

Hiuen Tsiang’s time.

1 Sastir gramasahasrani fastir grdmaiatani ca\ sastir gramas trayo grdmd hyetat

Kaimiramandalam II ; comp. Lokaprakasa, Ind. Studien, xviii. p. 375.

2 See Jonar. (Bo. ed), 153.

8 See Tdrlkh-i-Rash'/di , p. 430. Ritter who reproduces the passage of the

Zafarnama from De la Croix’s translation, shows the number of villages as 10,000;

see Asien, ii. p. 1123. It may be noted in passing that according to the Census of

1891 the number of villages in Kasmir was then reokoned at 2870.

J. i. 18

138

ANCIENT GEOGRAPHY OF KAS'MlR.

[Extra No. 2,

east bank of a great river, i.e. the Vitas ta, 12 or 13 li long from north

to south and 4 or 5 li broad from east to west, About 10 li to the

south-east of this, “ the new city,” the pilgrim notices a Buddhist convent

which lay between a high mountain on the north and the site of ‘ the

old city ’ on the south.

It is the merit of General Cunningham to have first recognized that

the situation here indicated for the new capital of Hiuen Tsiang’s time

corresponds exactly to that of the modern S'rinagar.1 A glance at the

map shows that the position and dimensions ascribed by Hiuen Tsiang

to the new city apply closely to that part of S'rinagar which occupies

the right or eastern riverbank, and which, as we shall see, forms the

older portion of the city. The two and a half miles represented by the

12 or 13 li of the Chinse measurement, agree accurately with the length

of the city within its ancient limits along the eastern bank of the Vitasta.

The estimate of its breadth at somewhat less than one mile (4 or 5 li)

is equally correct.

89. The position of ‘ the old city’ is marked by the present village

of Pandr ethan which derives its name from

the appellation Puranadhisthana, meaning * the

Old Capital.’ It lies to the south-east of S'rinagar just as Hiuen Tsiang

says, at the south foot of a mountain spur which rises with bold slopes

to a height of about 3000 feet above the village. Measured from the

nearest point of old S'rinagar, the distance to the presumptive site of

the monastery between Pandrethan and the steep hill-side is exactly two

miles or 10 li.

The history of ‘ the Old Capital ’ is so closely connected with that of

S'rinagara that it will be useful to acquaint ourselves first with the

data bearing upon it. The name of Puranadhisthana meets us first in

Kalhana’s account of the reign of King Pravarasena I. (or S'resthasena)

who is said to have erected there a shrine known as that of S'iva

Pravarehara .2 At the beginning of the tenth century the minister

Puranadhisthana.

Meruvardhana built at Puranadhisthana a Visnu temple called after his

own name. This has been rightly identified by General Cunningham

with the well-preserved little temple which still stands in the village

of Pandrethan and has often been described by European travellers.3

1 Gen. Cunningham’s identification was first indicated in his paper on the

architecture of Kastnlr temples, J. A. S. B., 1848, p. 283. For a fuller account, see

Anc. Geogr., pp. 93 sqq.

2 See Rajat. iii. 99 note, where detailed references have been given regarding the

site

3 ®ee v. 267 note, also for descriptions of the temple.

THE OLD AND NEW CAPITALS.

1899.]

139

Even in Kalliana’s own time pious foundations are recorded at this

ancient site.

The identity of Pandrethan with the site named in the Chronicle

as ‘ the Old Capital ’ is proved by ample evidence. It is indicated in

the old gloss on Rajat. v. 267 and is still known to Pandit tradition.

STivara in describing the flight of some troops which had been defeated

in Srinagar and were retiring along the Vitasta to the east, speaks of

the road from the Samudramatha (Sudarmar on the right bank of the

river near the second bridge) to Purvddhisthdva as covered with the

corpses of the slain. 1 It is clear that by the latter designation which

also means ‘the Old Capital,’ he refers to our present Pandrethan.

This name itself is the direct phonetic derivative of Puranadhisthchia .2

90. General Cunningham has assumed that ‘the Old Capital’

, _ _ marked by the site of Pandrethan was in

Aso a s Srinagar!. reajHy the ancient S'rInagarT which Kalhana

mentions as the capital founded by the great Asoka.3 * * * * 8 His assumption

was based on another passage of the Chronicle which mentions the

foundation of the shrine of Jyestharudra at STlnagari by Jalauka, the

son of Asoka. General Cunningham thought he could recognize this

shrine in the extant temple on the top of the Takht-i Sulaiman hill,

below which at a distance of about one and a half miles Pandrethan

is situated.

I have shown in my note on the passage that no reliance can be

placed on the alleged tradition which General Cunningham had adduced

as the sole proof of his location of the shrine. Yet at the same time

the evidence recorded by me proves that Jyestharudra must have been

worshipped either on the hill itself or in its close vicinity. Accordingly

Asoka’s STlnagari may safely be looked for in the same neighbourhood.

Our present data do not allow us to decide with absolute certainty

whether its site was at Pandrethan or elsewhere. But there are at

least sufficient indications to make General Cunningham’s view appear

very tempting and probable.

1 See S'nv. iv, 290.

2 The Ks. derivative of Skr. Purdna is pr<z«* ‘old’; this forms, with assimilation

of the initial double consonant, the first part, Pan -, of the modern name. The elision

of the second d in the assumed intermediary form # P[u]rdn[d]dethan is accounted

for by the influence of the stress accent which lies on the second syllable of the

modern name. The development of the combination nd into ndr is paralleled by

similar cases in other Indo- Aryan Vernaculars; comp. Dr. Grierson, Phonology of

Indo-Aryan Vernaculars, Z.D.M.Q., 1. p. 37, § 115. The nazalisation of e may be of

recent date, as the old gloss of A2 on Rajat. v. 267 shows the name as Pdmydrthan ,

i. e. Pan'drethan.

8 See Note C, i. 124.

140

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

There is in the first place the significant name Puranadhisthana, ‘ the

Old Capital,’ which shows that the site of Pandretlian must have once been

occupied by an important city. Next it is to be noted that Kalhana’s

narrative knows nothing of any other capital which might have been

founded in this vicinity previous to the new capital built by Pravarasenall.

on the site of the present S'rinagar. Lastly we have an indication in

the very name S’rlnagara which Pravarasena’s city has come to bear in

general usage instead of its proper and official designation Pravarapura.

If Asoka’s S'rinagari actually lay at or near the present Pandrethan

the transfer of its name to the new capital is most readily accounted

for. General Cunningham already has rightly pointed out the numerous

analogies for such a transfer furnished by the history of other Indian

capitals.1 Pravarasena’s city was practically contiguous to the older

S'rinagari and existed for centuries side by side with it. We can

hence easily understand that popular usage retained for the new capital

the old familiar designation.2 Exactly in the same way the several new

cities founded by successive kings in the vicinity of Delhi all continued

to be known simply by the name “ Delhi,” though each of them was

originally intended to bear the distinctive name of its founder.

Though Puranadhisthana had sunk to small importance already in

Hindu times, extensive remains of ancient buildings can still be traced

on the terraced slopes rising immediately to the north and north-east of

Pandrethan. Foundations of old walls, carved slabs, and architectural

fragments cover the foot of the hill-side for about one and a half miles.

Broken Lihgas of colossal dimensions are scattered among them. All

the remains above ground, however, are far too much decayed to permit

of a distinction of individual structures.3

The advantages of Pandrethan as the site for a great city cannot be

compared with those offered by the position of S'rinagar. Yet the

close vicinity of the Vitasta, coupled with the security from floods which

the near hill-slopes afford, must have been appreciated in an earlier

1 See Anc. Geogr ., pp. 97 sq.

2 The feminine form S'rinagari is used also for the new capital; comp. Rajat.

i. 104 note. There is thus no difference in the name as applied to both Asoka’s and

Pravarasena’s cities. S'rmagara or S'rinagari means the “City of S'ri ”, i.e. of

Laksmi, the Goddess of Fortune. For a whimsical etymology of European growth,

which has turned S'rinagar into the “ City of the Sun ”, see above § 4, note.

8 Compare for an account of these ruins, Cunningham, J. A. 8. B., 1848, pp. 283

sq ., Ajic. Geogr. 95 sq. [The remarks made in the latter place as to the supposed

cause of the desertion of Puranadhisthana rest on a misinterpretation of certain

Rajatarangini passages. The reconstruction of an alleged ‘ Pravaresvara symbol *

at Pandrethan, J. A, S. B 1848, pp. 324 sq.} is also unsupported by evidence.]

1899.]

THE OLD AND NEW CAPITALS.

141

Pravarasena’s

capital.

period when probably the riveraine flats of the valley were less drained.

The small semi-circular glens which are formed between projecting spurs

both north and east of the present village, with their gentle slopes offer

convenient building sites. The fertile shores of the Dal are also within

easy reach of Pandrethan through the gap in the hill-range which

separates the Takht-i Sulaiman hill from the greater heights to the east.

It is probably in this direction that we have to look for the Sarigharama

mentioned by Hiuen Tsiang in connection with ‘ the old city.’

91. Kalhana’s Chronicle furnishes us with a full account of the

origin of the new city which was the capital

of the Kasmir in his time and destined to

remain so to the present day.1 Kalhana attri¬

butes the foundation of this capital to King Pravarasena II. The topo¬

graphical details of his description make it clear beyond all doubt that

its site was that of the present S'rinagar.

The identity of the latter with Pravarasena’s town was duly recog¬

nized by General Cunningham who referred to the close agreement

between the general features of Kalhana’s description and the situation

of the present capital. He also pointed out that Kalhana distinctly

mentions as one of the pious buildings founded in Pravarasena’s city

that very Jayendrcivihara in which Hiuen Tsiang resided during his long

stay in the Kasmir capital.2 Subsequently Professor Biihler noticed the

survival of several old local names for parts of the modern city which

also prove its identity with Pravarasena’s capital.3 The most convinc¬

ing evidence, however, is contained in the long list of ancient buildings

and localities which Kalhana mentions in Pravarasena’s towTn. In the

course of our survey we shall be able to identify many of them within

the modern S'rinagar and its environs.

The attribution of this new capital to King Pravarasena rests on

equally strong proof. Through a chain of references extending over

nearly twelve centuries we can trace the use of the name Pravarapura,

shortened ( bhimavat ) for Pravarasencipura , as the official and correct

designation of the city occupying the site of the present S'rinagar. We

have found this appellation already in the record of the T'ang Annals

going back to the commencement of the eighth century. It is also found

in the works of Ksemendra, Bilhana, and numerous other Ka^mirian

authors. It has continued to be used to the present day in colophons

of Sanskrit Manuscripts, in horoscopes and similar documents.4

! See Rdj at, iii. 336-363.

2 See Anc. Geogr., p. 97 ; also Rajat. iii. 355 note.

3 Compare Report, p. 16.

4 For detailed references see my note Rdjat, iii. 339-349. S'ri-Pravarapure for

142

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Tlie date of King Pravarasena IT. whose name tlie above desig¬

nation of the new capital was intended, to preserve, cannot be fixed

with accuracy. Various historical and numismatic indications, however,

make it probable that he ruled at some period of the 6th century.

Thus we can easily understand that at the time of Hiuen Tsiang’s

visit (a.d. 631) S'rinagara or Pravarapura was still the ‘ new city.’

92. The traditional account of the foundation of Pravarapura as

t , „„ , ,. recorded by Kalhana is of considerable interest.

Legend of foundation J

of Pravarapura. Though largely interwoven with legendary

matter it preserves for us a series of exact

topographical data. Kalhana’ s story is contained in verses 336-349 of

the Third Book, and runs briefly as follows.1

When King Pravarasena II. had returned from his victorious ex¬

peditions abroad, he desired to found a new capital which was to beau

his name. He was then residing in the city of his grandfather Prava¬

rasena I., i.e ., in Puranadhisthana.2 From there the king went forth

at night in order, as the text says, “to ascertain in a supernatural way

the proper site and the auspicious time for the foundation of the new

city.” On his way he reached a stream which skirted a burning ground,

and was illuminated by the glow of funeral pyres. Then on the other

bank of the stream there appeared to him a demon of terrible

form. Promising him fulfilment of his desire, the demon invited the

king to cross over to his own side by the embankment he was preparing

for him. Thereupon “ the Raksasa stretched out his own knee from

the other bank, and thus caused the water of the Mahasarit to be parted

by an embankment (Setu).” The courageous Pravarasena drew out

his dagger ( ksurikd ), cut with it steps into the flesh of the Raksasa,

and thus crossed over to the place which has since been known as

Ksurikabala. The demon then indicated to him the auspicious time

and disappeared, after telling him to build his town where he would

see the measuring line laid down in the morning. This line ( sutra ) of

the Vetala the king eventually discovered “ at the village of Sdritaka

at which the goddess S'drikd and the demon Atta resided.” There

he built his city in which the first shrine erected was the famous one

of S'iva Pravaresvara .

Srtpravarasenapure is often written in the abbreviated form S'ripre in the formulas

of the Lokaprakasa, almanacs, etc. Kalhana often uses the simple Pur a for Par-

varapura and Nagara for S'rlnagara.

1 For all detailed references in connection with this story, note iii. 339-349

should be consulted.

2 That Puranadhisthana is meant is proved by iii. 99. There Kalhana, speaking

of a foundation of Pravarasena I. in his capital, by a kind of anachronism uses the

designation of Puranadhisthana.

THE OLD AND NEW CAPITALS.

143

1899.]

Keeping in view the details of the ancient topography of S'rinagar,

we can still follow np step by step the localities by which. the legend

here related leads King Pravarasena to the site of his new city. We

have already seen that the Mahasarit is the stream now known as

Tsunth Knl which flows from the Dal into the Vitasta. Near its con-

fluence with the Vitasta which we have also found already mentioned

as a Tirtha, there existed, until the times of Maharaja Ranbir Singh,

a much frequented Hindu burning Ghat. It was undoubtedly of ancient

date. Kalhana relates how the body of King Uccala, murdered

in his palace at S'rinagar, was hurriedly cremated at the burning

place situated on the island at the confluence of the Mahasarit and

Vitasta,.1 It is certain that the island of May^sum (Skr. Maksika-

svamin) is meant here, at the western end of which the Mahasarit or

Tsunth Kul falls into the Vitasta.

■" ■ ■ • •

The stream flowing from the Dal is bounded on its northern bank

by an old embankment which stretches from the west foot of the

Takht-i Sulaiman close to the high bank of the Vitasta near the Second

Bridge. This embankment which is the most substantial at or around

STinagar and known only by the general designation of Suth (from

Skr. setu ), ‘dyke,’ is undoubtedly of very early date. It protects the

whole of the low-lying portions of the city on the right river-bank as

well as the floating gardens and shores of the Dal which would other¬

wise be exposed to annual inundations from the Vitasta. A tradition

still heard by Mr. Vigne ascribed the construction of this embankment

to King Pravarasena.2 * It is indeed evident that its construction was a

necessary condition for the safety of the newly founded city.

Several topographical indications warrant the conclusion that it

was this old dyke in wrhich the popular legend recorded by Kalhana

recognized the leg and knee of the demon. A glance at the map shows

that the eastern portion of the ‘ Suth ’ turns sharply at a right angle

and thus curiously resembles a bent knee. Ksurikdbala was the name

of the place where Pravarasena according to the legend was supposed

to have reached firm ground after crossing the stream. I have shown

that this name in the form of its Kasmiri derivative Kliudabal still

attaches to the city quarter which lies at the western end of the Suth.8

Finally it will be seen from the map that Kalhana’s words regard¬

ing the ‘ Setu ’ dividing the waters of the Mahasarit, describe exactly

the present embankment which has on one side the Tsunth Kul and

on the other side the various marshes and canals fed by the Mar. It

1 See viii. 339.

2 See Vigne, Travels , ii. p. 69.

s See note iii. 339-349.

U4

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Old limits of Prava¬

rapura.

has been shown above that this second outflow of the Dal also shared

the old name of Mahasarit.1 2

93. The name of the village Saritaka where the demon showed

to the king the proper site for his city, has

long ago disappeared. Its position, however,

is sufficiently marked by the mention of the

goddess S'cirikd. The latter, a form of Durga, has since ancient times been

worshipped on the hill which rises to the north of the central part of

Srinagar and is still called after her. The modern name of the hill,

Haraparvat, is the regular phonetic derivative of Skr. Sarikaparvata.

By this name it is designated in the latter Chronicles and Mahatmyas.8

Another passage of the Ttajataranginl shows that the term Vetala-

siitrapata , * the demon’s measuring line,’ clearly connected with the

above legend, was also in later times applied to the limits of the oldest

part of Pravarapura 3. But our materials do not enable us to ascer¬

tain these limits in detail. Kalhana it is true, has not failed to specify

them, as he mentions the temples of Vardhanasvamin and Visvakarman

as marking the extreme ends of Pravarasena’s city 4 *. Unfortunately

the position of neither of these structures can now be traced.

So much, however, is clear that the new city was at first confined

to the right bank of the river. Kalhana tells this distinctly, 6 and

those sites and structures which he particularly mentions in his de¬

scription of Pravarasena’s capital, are all found as far as they can be

identified, on the right bank. The account of Hiuen Tsiang and the

T'ang Annals show that even in the 7th century Pravarapura extended

mainly along the eastern bank of the river.6

Kalhana follows up his account of the foundation of the city with

a brief description of its splendours 7. He

tioKnaofapraavampuraP' noteS tte ^yagant story of its having

once counted thirty-six lakhs of houses, and

1 Compare § 65.

2 See note iii. 339-349. HarQ is the Kasmirl name of the goddess Sarika as

well as of the S'arikd bird (Maina) ; comp. Buhler, Report, pp. 16 sq.

Panjabis and other foreign visitors from India have by a popular etymology

turned the * Hill of S'arika ’ into the ‘ Hill of Hari (Visnu) ’ or the ‘ Verdant Hill.*

The latter interpretation could be justified only on the principle of lucus a non

lucendo ; for verdure is scarce indeed on the rocky faces of the Sarikaparvata.

Dr. Bernier already, Travels, p. 398, was told this popular etymology, probably by

his friends from Delhi.

8 See vi. 191 note.

4 iii. 357.

6 iii. 358.

® See above, §§ 8, 10.

7 iii. 357-363.

1899.]

THE OLD AND NEW CAPITALS

145

refers to the regularly arranged markets with which its founder had pro¬

vided it. The city of his own time still boasted of “ mansions which

reached to the clouds ” built, no doubt, mostly of wood just as the mass

of private houses in modern S'rinagar.1

When he mentions “ the streams meeting, pure and lovely, at plea-

su re-residences and near market streets,” he means evidently the

numerous canals from the Dal and Anchor lakes which intersect the

suburbs and also pass through the heart of the city. They and the

river still serve as the main thoroughfares for the market traffic, and

all principal Bazars are built along their banks.2 The S'arikaparvata

receives due mention as “ the pleasure-hill from which the splendour

of all the houses is visible as if from the sky.” Nor does he forget to

praise the cool water of the Vitasta which the citizens find before their

very houses on hot summer-days.

Finally he refers to the abundance of magnificent temples with

which successive kings had adorned Pravarapura, and of which so many

are particularly mentioned in his narrative. Of the number and im¬

posing appearance of these structures we can even at the present day

form some idea if we examine their massive remains which meet us in

every part of modern S'rinagar. The high embankments which now

line the river’s course within the city, are mainly composed of carved

slabs, columns and other ancient stone materials. Their profusion and

imposing dimensions must even to a superficial observer suggest an idea

of the architectural splendour of ancient S'rinagar.

94. It can scarcely be the result of chance that Pravarasena’s

city has escaped the fate of so many Indian

Advantages of the capitals, of being superseded by later founda-

site of Srinagar. tions

There had indeed not been wanting attempts on the part of later

rulers to transfer the capital to other sites which they had chosen for

their own cities. The great Lalitaditya, then Jayapida, Avantivarman,

1 Both Mirza Haidar and Abu-1-Fazl apeak with admiration of the many lofty

houses of S'rinagar, built of pine wood. Thia material was used, then as now, as

being cheap and more secure against earthquakes. According to Mirza Haidar ‘‘most

of these houses are at least fire stories high and each story contains apartments,

halls, galleries and towers” ( Tdrikh-i-Rashidi , p. 425). That the mass of private

dwellings in S'rinagar were already in Hindu times constructed of wood, is shown

by Rdjat. viii. 2390. The many disastrous fires recorded point to the same con¬

clusion.

2 Useful and convenient as these canals undoubtedly are, it is rather difficult

to concede to them now the epithets of ‘ pure and lovely.’ They add, however,

greatly to the pictaresqueness of the city and certainly make the want of carriage

roads less felt.

J. i. 19

146

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

and S'amkaravarman, had successively endeavoured to effect this object.

The great ruins of Parihasapura, Jayapura and Avantipura show

sufficiently that the failure of the first three kings was not due in any

way to deficient means or want of purpose.

Of Lalitaditya the Chronicle distinctly records that he proposed,

Nero-like, to burn down the ancient capital to assure the predominance

of his own creation, Parihasapura. And the long list of splendid edifices

erected at the latter place during his own reign shows plainly that for

a time at least that monarch’s pleasure had succeeded. Yet each one

of these temporary capitals speedily sank into insignificance, while

Pravarapura continued to be the political and cultural centre of Kasmlr

down to the present day.

We can safely attribute this exceptional position of Srinagar to the

great natural advantages of its site. Occupying a place close to the

true centre of the Valley, S'rinagar enjoys facilities of communication

which no other site could offer. The river along which the city is built

provides at all seasons the most convenient route for trade and traffic,

both up and down the Valley. The two lakes which flank STinagar,

offer the same facilities for the fertile tracts which lie immediately to

the north. The lakes themselves furnish an abundant supply of

products which materially facilitate the maintenance of a large city

population. The great trade route from Central Asia debouches

through the Sind Valley only one short march from the capital.

Nor can we underrate the security which the position of S'rinagar

, , _ _ assures both against floods and armed attack.

Natural defences of . e , v , , , , ,,

Srinagar The neck or high ground which irom the north

stretches towards the Vitasta, and separates

the two lakes, is safe from all possible risk of flood. It is on this ground,

round the foot of the S'arika hill, that the greatest part of the old Pra¬

varapura was originally built. The ancient embankment which connects

this high level ground with the foot of the Takht-i-Sulaiman hill

sufficed to secure also the low-lying wards fringing the marshes of the

Dal. A considerable area, including the present quarters of Khanayar

and Ranlvor (Skr. Rajanavdtika), was thus added to the available

building ground on the right bank and protected against all ordinary

floods.

The frequent sieges which S'rinagar underwent during the last

reigns related by Kalhana, give us ample opportunity to appreciate also

the military advantages which the city’s position assured to its defenders.

With the exception of a comparatively narrow neck of dry ground in

the north, the S'rinagar of the right river-bank is guarded on all sides

by water. On the south the river forms an impassable line of defence.

1899.]

ANCIENT SITES OF S'RINAGARA.

147

The east is secured by the Dal and the stream which flows from it.

On the west there stretch the broad marches of the Anchor divided

from the Vitasta by a narrow strip of firm ground.

From the north, it is true, the city can be approached without passing

such natural obstacles. But the map shows that just to the north of the

S'arika hill inlets from the two lakes approach each other within a few

tliousaud feet. The narrow passage left between them could at all

times easily be guarded. It is curious to note that the successful attacks

on the city of which the Chronicle tells us, were delivered from the

north, treachery or the defenders’ weakness having opened this passage.1

The later and smaller portion of STinagar occupying the left river-

bank, does not share the same natural advantages as the old one. The

present level of the ground on which it stands appears to have been

raised gradually by the accumulated debris of centuries. We do not

know exactly when the extension of the city in this direction began.

The number of ancient sites on this side is comparatively small. The

royal residence was transferred to it only in the reign of Ananta

(a.d. 1028-63). There too we find a natural line of defence. It is the

Ksiptika or Kutakul which flows round the western edge of this part of

the city and is also often mentioned in the accounts of the later sieges.

Section IV. — Ancient sites of S'rinagara.

95. Having thus reviewed the origin and the general position of

Hill of Marika ^ie capital, we may proceed to a brief

survey of the more important ancient sites

which our available materials permit us to trace in it. We can conve¬

niently start on our circuit from the Hill of S'arika to which the

legendary account of the city’s foundation had taken us.

The goddess S'arika which has given to the hill its name, has been

worshipped since ancient times on the north-west side of the hill. Certain

natural markings on a large perpendicular rock are taken by the pious

to represent that kind of mystical diagram which in the Tantrasastra

is known as S'rlcakra ? This ‘ Svayambhu’ Tirtha is still a much fre¬

quented pilgrimage place for the Brahmans of the city and has been so

probably since early times.8 The S'arikamahatmya now in use relates

1 Compare for Uccala’s entry into S'rlnagar, vii, 1539 sqq. ; that of Sussala, viii.

944 sqq. ; compare also note viii. 1104-1110.

* Compare Rdjat. note i. 122, regarding the worship of such diagrams.

1 Compare Jonar, (Bo. ed.) 472, 767.

148

ANCIENT GEOGRAPHY OF KAS'MlR

[Extra No. 2,

that the hill was carried to its present position by Durga who had taken

the shape of a S'arika bird. The goddess is supposed to have thus closed

a gate of the Daityas dwelling in hell. This legend is alluded to already

in the Kathasaritsagara.1

Another ancient designation of the Haraparvat is ‘ Hill of Pra-

dyumna ’ ( Pradyumnapitha,-giri,-6ikhara , etc.), often found in the

Chronicles and elsewhere.2 The Kathasaritsagara accounts for the

origin of this name by a story which connects the hill with the love of

Usa and Aniruddha, the son of Pradyumna. Kalhana mentions a

Matha for Pasupata mendicants vdiich King Ranaditya built on the hill.

The eastern slopes of the latter are now occupied by extensive buildings

connected with the famous Ziarats of Muqaddam Sahib and Akhun

Mulla Sbah. It is probable that these Muhammadan shrines have taken

the place of Hindu religious buildings, as at so many old sites of

Kasmir.

Close to the foot of the southern extremity of the hill lies a rock

which has from ancient times received worship as an embodiment of

Ganesa, under the name of Bhimasvamin. A legend related by Kalhana

connects this ‘ Svayambhu ’ image with Pravaraseua’s foundation of

S'rinagar.8 From regard for the pious king the god is there said to

have turned his face from west to east so as to behold the new city.

The rock is covered by the worshippers with so thick a layer of red

lead that it is not possible to trace now any resemblance to the head of

the elephant-faced god, still less to see whether it is turned to the west

or east. In fact, if we are to believe Jonaraja, the rock image has

subsequently changed its position yet a second time. This Chronicler

relates that Blnmasvamin from disgust at the iconoclasm of Sikandar

Butsliikast has finally turned his back on the city.4 This last turn

would, no doubt, most satisfactorily account for the present amorphous

look of the sacred rock.

There is nothing in the Chronicles that would lead us to assume

that the hill of S'arika was ever fortified in Hindu times. The great

bastioned stone-wall which now encloses the hill and the ground

around its foot (Nagar-nagar), was built by Akbar as an inscription

still extant over the main-gate proclaims.6 The fort which now crowns

the summit of the hill, is of even more modern origin.

■j *

1 See lxxiii. 107 sqq.

8 See Rajat. iii. 460 note.

f See iii. 352 note.

4 See Jonar. (Ro. ed.), 766,

6 Compare Fonrth Chron, 939 sqq.

1899.]

ANCIENT SITES OF S'RINAGAR A

149

Temple of Pravares-

vara.

96. A short distance to the south-east of the Bhimasvamin rock,

and outside Akbar’s fortress, lies the Ziarat of

Bahau-d-din Sahib, built undoubtedly with the

materials of an ancient temple. The cemetery

which surrounds it contains also many ancient remains in its tombs

and walls. At the south-west corner of this cemetery rises a ruined gate¬

way built of stone-blocks of remarkable size, and still of considerable

height. This structure is traditionally believed by the S'rinagar

Pandits to have belonged to the temple of 8'iva Pravaresvara which

Kalbana mentions as the first shrine erected by Pravarasena in his new

capital.1

An old legend related by Kalbana and before him already by

Bi 111 ana, makes the king ascend bodily to heaven from the temple of

Pravaresvara. Bilhana speaks of the temple as “ showing to this day

a gap above, resembling the gate of heaven through which King

Pravara bodily ascended to heaven.”2 Kalbana, writing a century later,

also saw at the temple of Pravaresvara “ a gate resembling the gate of

heaven.” Its broken stone roof was supposed to mark the king’s pas¬

sage on his way to S'iva’s abode.

This tradition still attaches to the roofless stone-gate above de¬

scribed, which may indeed be the very structure seen by Bilhana and

the Chronicler. As far as its architecture is concerned, it might well

belong to the earliest monuments of S'rinagar. It owes its preservation

perhaps to the exceptional solidity of its construction and the massive¬

ness of its stones. Stone-blocks measuring up to sixteen feet in length

with a width and thickness equally imposing were no convenient

materials for the builders of Muhammadan Ziarats, Hammams, etc., who

have otherwise done so much to efface the remains of ancient structures

in S'rinagar. The position of the ruin is very central and might well

have been chosen by the founder of Pravarapura for a prominent shrine

in his new city.

Not far from Bahau-d-dln Sahib’s Ziarat to the south-west stands

the Jami‘ Masjld, the greatest Mosque of

S'rinagar. Around it numerous ancient re¬

mains attest the former existence of Hindu

temples. Proceeding still further to the eoutli-west in the midst of a

thickly built city quarter, we reach an ancient shrine which has remain¬

ed in a comparatively fair state of preservation probably owing to its

early conversion into a Ziarat. It is now supposed to mark the resting

place of the saint styled Pir Haji Muhammad. It consists of an octa-

Position of Rana-

svamin temple.

1 See Rajdt. iii. 350 note.

8 See ViTtram. xviii, 23.

150

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

gonal cella of which the high basement and the side walls are still well-

preserved. The quadrangular court in which it stands is enclosed by

ancient walls and approached by ornamented gateways.

The position of this shrine has suggested to me its possible identity

with the ancient temple of Visnu Ranasvamin which Kalhana mentions as

founded by King Ranaditya.1 This temple must have enjoyed consider¬

able celebrity till a comparatively late period. Manklia refers to it as

an object of his father’s devotion and Jonaraja in his comments on the

passage speaks of Visnu Ranasvamin as one of the chief shrines of

Pravarapura.2 3

In his own Chronicle Jonaraja indicates this temple as the furthest

point up to which Zainu-l-‘abidin carried the canal flowing through

Jainanagari} The latter locality corresponds to the Srinagar quarters

of Sangin Darwaza and Naushahr, and the canal itself is the one now

known as Lacham Kul .4 * It brings the waters of the Sind River via Am-

burlier to the northern suburbs of S'rinagar, and after flowing past the

Jami‘ Masjid empties itself into the Mar canal near the bridge called

Kad1 Kadal. In the corner formed by the two canals stands the ruined

temple above described. If it could be shown that the present termina¬

tion of the Lacham Kul is the same which Jonaraja knew in the

middle of the fifteenth century, the identity of those remains with

the Ranasvamin temple might be considered as certain.

97- Crossing the Mar to the south we reach the city quarter

known as Bradhnar , occupying the right bank

of the river between the Fourth and Fifth

Bridge. It derives its name from the ancient

Bhattarakamatha which is repeatedly referred to in the Rajatararigini

as a building of considerable size and strength.6 Bilhana too notices it

specially in his descripton of STinagar. Like other Mathas built

originally to serve the purposes of a Sarai, it was used on occasion as a

place of defence. Queen Didda sent her infant son there at the time of

a dangerous rising.

Bhattarakamatha ;

Diddamatha.

1 Raj at. iii. 453 sq. note.

2 See Srileanthacar . iii. 68.

3 See Jonar. 872.

* Compare Rajat. iii. 453-454 note. The Lacham Kul is mentioned by

Abu-1-Fazl, ii. p. 355. It probably took its name (equivalent to #Laksmikulya) from

Laksmi, the queen of Shahabu-d-din, in whose honour the quarter of Laksminagari

was found in the vicinity of the present Sangin Darwaza ; see Jonar. 407 sq.

6 See Rajat. vi. 240 note ; viii. 2426 ; Viler am, xviii. 11. For the derivation of

Brad1 from Bhattaralea comp. Brqdnamlal < Bhattaranadvald , below. That Bhatta¬

rakamatha was the old name of this locality, is known to the tradition of the

Fandits 5 see Buhler, Report, p. 16.

1899.]

ANCIENT SITES OF S'RINAGARA.

151

The Chronicle shows us often the Mathas of Srinagar utilized as

places of refuge in the times of internal troubles, occasionally also

turned into prisons.1 We may hence conclude that they were substan¬

tially built, probably like modern Sarais in the form of detached quadran¬

gles, and thus better adapted for defence than other city-buildings.

That Mathas more than once left their names to the city-quarters

in which they stood, is shown by the designation of other wards. Thus

the large quarter of _Didamar which forms the western end of the city on

the right river-bank, retains the name of the Diddamatha.2 It was

built by Queen Didda for the accommodation of travellers from various

parts of India. As a local name Diddamatha meets us often in the later

Chronicles. Above Didamar we find near the Sixth Bridge the quarter

of Bctlandimar .3 * It represents in all probability the Baladhyamatha

of the later Chronicles which Jonaraja mentions as having been built

by Baladhyacandra under King Rajadeva in the 13th century.41

A little to the north of the Sixth Bridge lies the Mahalla known by

, , , , the name of Khcindabavan. It has received

Skandabhavana ; ., „ , ,, . , „

Nadavana its appellation from the ancient Vihara ot

Skandabhavana, a foundation of Skandngupta

whom Kalhana mentions among the ministers of Pravarasena II. ’s

successor Yudhistliira.5 The site of the Vihara has been traced by me

in the close vicinity of the Ziarat of Pir Muhammad Basur. Certain

ancient remains there were locally known and worshipped till the

middle of the present century as a Tlrtha sacred to Skanda.6

The ground immediately to the north-east of Khandabavan is now

an open waste space used partly for Muhammadan graveyards. It

seems to have been unoccupied already in old times. For it was chosen

as the burning place for the widows of the murdered king Sussala

when a rebel force hovering around the capital rendered the usual

burning ground on the island of Maksikasvamin inaccessible.7

The quarter of Nctrvor still further to the north is probably iden¬

tical with the old Nadavana, mentioned by Kalhana as the site of a

Vihara built by one of King Megliavahana’s queens. I have shown

in my note on the passage that the modern name goes back to a form

1 Compare Rdjat. vi. 223; viii. 374, 1052, 2309.

2 See Rdjat. vi. 300 note.

3 Its old name could not be shown on the map owing to want of room.

* See Jonar. 82.

6 See iii. 380.

8 Compare Note K, vi. 137, also for the temple of Parvaguptesvara which stood

close by.

7 Rdjat. viii. 1441 sg.

152

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

Bridges of old Srl-

nagara.

#Nadavata.1 The termination vat a ‘ garden,’ frequent in Kasmir local

names, may safely he taken as the equivalent of vana in Kalhana’s form

of the name.

98. Before we continue our survey further up the river, it will be

useful to make a brief reference to the bridges

which connect the two river-banks within the

city. S'rinagar lias now seven bridges across

the Vitasta. Their number has remained unchanged for at least five

hundred years.

Already Sliarifu-d-din had heard that of the thirty boat-bridges

constructed across the great river of Kasmir, there were seven in the town

of S'rinagar. The boats were bound together by chains, and through the

bridges a way could be opened for the river traffic.2 Sharifu-d-din’s

notice is of interest because it shows clearly that down to the end of the

Hindu period permanent bridges across the Vitasta where unknown in

Kasmir.

I had been led to the same conclusion by an examination of the

Rajataranginl passages bearing on the subject.3 Kalhana distinctly

says of the two bridges the construction of which he specially records,

that they were built with boats. Elsewhere this inference may be

drawn from the rapidity with which the bridges are broken at the

approach of the enemy or in danger of fire.4 *

The first bridge of this kind is ascribed by Kalhana to Pravara-

sena II. who built the ‘ Great Bridge ’ {Brhatsetu)

in his new capital. “ Only since then is such

construction of boat-bridges known.”6 This ‘ Great Bridge ’ is

subsequently mentioned in connection with a great conflagration

which destroyed the city in the time of Sussala (a.d. 1123). This

fire arose at the southern end of S'rinagar, and Kalhana mentions

that the smoke first rising from Maksikasvamin: May^sumliad scarcely

been noticed from the ‘ Brhatsetu ’ when the fire was already spreading

over the whole city.6 Kalhana evidently refers to the 1 Great Bridge ’

as a comparatively distant point from Maksikasvamin. Considering

that the river forms an almost straight reach from this locality to the

present Fourth Bridge, it appears to me likely that Pravarasena’s bridge

was somewhere in the vicinity of the latter. The position is in the

Brhatsetu.

1 Compare Rajat. iii. 11 note.

2 See Tdrtkh-i-Rashldtf p. 431.

s See note iii. 354.

* See Rajat. vii. 909, 1539 ; viii. 1182 ; S'riv. i. 308, 720 ; ii. 70, 122.

6 Rajat. iii. 354.

6 Compare Rajat. viii. 1171-72 note.

1899.]

ANCIENT SITES OP R'RINAGARA.

153

very heart of the city. It is just here that Zainu-l-‘abidin subse¬

quently constructed the first permanent bridge over the Vitas ta named

after him Zaina Kadal (Jainakadali) ,l

Another old boat-bridge had been established by Harsa just oppo¬

site to his palace.2 The latter as we shall see was situated on the left

bank somewhere near the present Second Bridge (Haba Kadal). The

bridge proved fatal to Harsa’s fortunes, because it enabled the rebels

to make their final and successful assault on the palace.

There can be little doubt that the first permanent bridge across the

Vitasta was of wood and showed the same peculiar cantilever con¬

struction which the Kasmlr bridges have preserved to this day. The

latter have attracted the attention of all modern travellers and have

often been described.8 But it is curious that none of them can be traced

back beyond the time of Zainu-l-‘abidin. The explanation may lie in

the fact that that stone-architecture in which the engineers of the

Hindu period were so proficient, did not permit of the construction of

bridges with a sufficient span. For their Muhammadan successors

working chiefly in wood it was easier to overcome this difficulty.

Among the most characteristic features of the river-scene as it

now presents itself within S'rinagar, are the numerous wooden bathing

cells moored before all city Ghats. They have been there already in

Hindu times. For Kalhana mentions more than once the sndna-

kosthas of the river.4 From a humorous sketch of city-life which

Kalhana draws for us, we can see that they formed, then as now, the

favourite meeting-place of the idle and curious.6

99. Resuming our walk up the river-bank we pass the remains of

more than one old temple near the present

St|rInagaraerS Ziarats of B ^ ^ah (Zainu-l-‘abidin), .Shall

Hamadan and elsewhere. But we have no

data for their identification. An old site is marked by the present

Ghat S5mayar, below the Second Bridge, which represents the

Somatirtha of the Rajatarahgim.6 The place is still visited as a

Tirfcha, and some old Lingas are found by the river-side. The quarter

in which the Somatirtha lies, is known as Sudarmar. It owes its name

to the Samudramatha built by Samudra, the queen of Ramadeva, in

1 See S'rlv. i. 231 sq., 296.

2 Rajat. vii. 1549.

3 See, e.g., Vigne, Travels , ii. 23; Lawrence, Valley, p. 37.

4 Compare Rajat. viii. 706, 1182, 2423. Also Ksemendra, Samay. ii. 38, know

the term sndnahosthaJca which lives in the present Ks. iran&kuth.

6 See Rajat. viii. 706-710.

6 See Rajat, viii. 3360 note.

J. r. 20

154

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

the 13th century. The numerous passages in which the Samudramatha

is mentioned by the later Chronicles, makes this identification quite

certain.1

A little higher up, if we can trust local tradition, stood the

ancient temple of Vardhamanesa mentioned already in King Samdhi-

mat’s reign. The site so designated by the Purohitas of the adjoining

Mahalla is close to the Malayar Ghat. I have referred already in a

previous note to the curious manner in which an ancient Linga sup¬

posed to be that of Vardhamanesa was recovered a few years ago from

a neighbouring Mosque and a Mahatmya composed for the newly

established shrine.2 *

The confluence of the Tsunth Kul or Mahasarit with the Vitasta

we have also had occasion to notice. 8 It is the Tirtlia now known as

MarIsamgama. Beyond it lies the great island of Maytsum, the ancient

Maksikasvamin, now chiefly occupied by the houses and camps of Euro¬

pean residents and visitors. From the way it is referred to by Kalhana,

it appears that it was already partly inhabited in old times.4 * Follow¬

ing up the right bank of the Mahasarit above the junction we reach

the quarter of Kliudabal already identified with the Ksurikabala of

King Pravarasena’s story.

Here begins the old embankment or Setu, noticed in connection

with the latter.6 To the north of this embankment stretches an exten¬

sive marsh fed by canals coming from the Dal and known as Brar‘lnambal.

It is the Bhattdranadvald of the Chronicle into which the body of one

of Harsa’s ministers was thrown after his execution.6

At the eastern end of the Setu where it joins the rocky foot of the

Takht-i Sulaiman hill, there has been for at least a century a gate

through which the Tsunth Kul flows out from the lake. It is closed at

times of flood when the Vitasta rises higher than the level of the Dal.

It is highly probable that this gate is very old and contemporary with the

construction of the embankment itself. Beyond it lies the suburb of

Drugajan. This is identified in an old gloss of the Rajataranginl with

Dorgagalika, where according to tradition the blind King Yudhisthira I.

was imprisoned after his abdication.7

1 See Jonar. Ill ; S'rlv. iv. 121, 169, 290 j Fourth Chron. 504, 618.

3 See above, § 31 note 2.

8 See above, § 65.

4 See Bajat. iv. 88 note.

6 Compare § 92.

* See Bajat. vii. 1038. Nambal, from Skr. nadoald, is the regular Ks. word for

‘ marsh.’ Brqr * is a direct phonetic derivative of Skr. bhattara ‘god.’

1 See Bajat. ii. 4.

1899.]

ANCIENT SITES OF S'RINAGARA.

155

Left river bank.

Leaving tlie Seta where it makes its great bend and going north

across low ground flanked by marshes, we reach the quarter of Nov^pur,

The bridge which leads here over the Mar or Maliasarit, is repeatedly

mentioned as Naupurasrtu by STivara, in connection with later sieges

of STinagar.1 By breaking it, the south-eastern parts of the city were

rendered more secure. Continuing our route to the north we come

to the great suburb of Rancor. It is traversed by numerous canals

coming from the Dal. Kalhana mentions it repeatedly by its ancient

name of Rajanavatika as a place largely inhabited by Brahmans. Their

solemn fasts ( prciyopavesa ) gave no small trouble to King Sussala in his

worst straights.2 Raidvor lias continued to the present day a favourite

place of residence for city Brahmans.

100. We have now completed our circuit of the ancient city as

far as it lay on the right bank of the river

and may proceed to the smaller and later por¬

tion which occupies the left bank. Just opposite to the ‘ Marisamgama’

stands the Shergarlii, the modern palace of the Dogra rulers. Its

site was apparently first chosen by the Path an governors for their

fortified residence.

Immediately below the palace the Kutakul or Ksiptika branches

off from the river. We have already noticed its value as a line of

defence for this part of the city.3 The quarter of Kathiil which lies

next between the Kutakul and the river is of ancient date. It is men¬

tioned as Kasthila by Kalb ana and other writers, Bilhana speaking

of it particularly as a locality inhabited by Brahmans.4 *

At the northern end of the Kathiil quarter and close to the pre¬

sent Second Bridge, we must assume the palace

of the later Hindu kiugs to have stood. Its

position is indicated by an interesting passage

of the Rajatarangini which informs us that King Ananta (a. d.

1028-63) abandoned the palace of the former dynasties and transferred

the royal residence to the vicinity of the shrine of Sadasiva.6 The

new site was adhered to by subsequent kings probably till long after

Kalhana’s time. The mention of the Sadasiva shrine and the fre-

Site of Royal

Palace.

1 See S'nv. iv. 122, 243.

2 See Jtajat. viii. 756, 768, 899. For the phonetic relation of Rani < Skr.

Rajana, see viii. 756 note; vor is common in Ks. local names and derived from

Skr. vdtikd ‘garden.’

3 See above, § 67.

4 See Rajat. viii. 1169 note, and Vikram. xviii. 25.

6 Compare Rdjat, viii. 186-187, and for detailed proof of the identification, the

note thereon.

156 ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

quent references to the Ksiptika as flowing near to the royal palace

(rajadhani) enable us to fix the position of the latter with fair accu¬

racy. In the note on the above passage I have shown that the Sada-

siva temple stood opposite to the Samudramatha which occupies the

right river bank just below the Second Bridge. Exactly in the posi¬

tion thus indicated we find now an ancient Linga, on the river Ghat of

Purusayar, which the tradition of the local Purohitas knows by the

name of Sadasiva.

It is in this neighbourhood, then, that the palace stood which had

witnessed so many tragic scenes related in the last two Books of

Kalhana’s Chronicle. Its great height is specially referred to by

Bilhana. This suggests that it was in part at least built of wood, just

like a later palace described by Mirza Haidar.1 “Sultan Zainu-l-‘abidin

built himself a palace in the town which in the dialect of Kashmir is

called Rcijdan [i.e., Skr. rajadhani']. It has twelve stories, some of which

contain fifty rooms, halls and corridors. The whole of this lofty struc¬

ture is built of wood.” This construction of the palace would well

explain the rapidity with which it was burned down by the pretender

Uccala on his final attack upon Harsa.2 We can thus also understand

why there are no particularly striking remains at the site which could

be attributed to the ruins of this royal residence.

The last-named incident gives Kalhana occasion to mention also

some other data regarding the royal palace. Close to it was a garden

in which Harsa and his ill-fated son Bhoja enjoyed a deceptive rest

before the rebels’ last assault.3 The gardens near the palace are also

elsewhere mentioned. Harsa had their trees cut down because they

obstructed the view, and at a later time the besieging Damaras fed their

camp-fires with wood brought from the same gardens.4 * Even at the

present day there are numerous old gardens across the Ksiptika close

to the site where the palace once stood. In front of the palace was the

boat-bridge already mentioned which the king had himself constructed,

and which was the scene of his last desperate struggle.6

Where the old palace stood which was abandoned by King Ananta,

we cannot say with accuracy. It is, however, probable that its site was

in the old part of Pravarapura on the right bank. Kalhana mentions it

twice as purdnarajadhani (‘the old palace’), but gives no particulars.6

Its deserted ground was built over with a Matha in Kalhana’s own time.

1 See Tdrikh-i-Rashidi, p. 429.

2 See Rdjat. vii. 1565 sq., 1583.

s Rdjat. vii. 1538 sqq.

4 Rdjat. vii. 1223 ; viii. 1056.

* See Rdjat. vii. 1539, 1549.

& See Rdjat. viii. 837, 2417.

1899.]

THE ENVIRONS OF S'RINAGARA.

157

The embankments on the left side of the river as well as the

walls of Ziarats show ample remains of ancient buildings. But we have

no means of identifying any particular sites. At the western ex¬

tremity of this part of the city, however, we may locate with some

probability the temple of Ksemagaurisvara, built by Queen Didda’s weak

husband Ksemagupta. Bilhana in his description of S'rinagar mentions

it as an imposing building, the 1 2 * 4 Mandapas * of which extended to a

4 Samgama ’ of the Vitasta.1 I have shown elsewhere that the confluence

meant is probably that of the Vitasta with the Dugdhasindhu or

Chatg^kul which lies opposite to the quarter of Diddamatha.2

Section V. The Environs of S'RINAGARA.

101. Having completed our survey of old S'rinagara we may now

proceed to examine the ancient sites of its environs. They are almost

all situated to the north of the Yitasta within the Pargana now known

as Phakh , and designated as Phakhuvd in S'rlvara’s Chronicle.8 It

comprises the tract lying between the east shore of the Anchor, the

range towards the Sind Valley and the hills which enclose the Dal on

the east and south. Owing to the facility of communication across the

lake and the manifold attractions of its shores, Phakh seems to have

always been a favourite resort for the inhabitants of the capital. This

fact is fully illustrated by the numerous places of ancient date which

we find dotted around the lake.

Starting from its southernmost corner in the immediate vicinity

, of the city we come first to the hill popularly

(Gopadri ) known as lakht-i-feulaiman. Its bold pyrami¬

dal form and the old temple which crowns its

summit, make this hill a most conspicuous object in the land-scape of

S'rinagar. The present name of the hill, meaning 4 Solomon’s throne,’

is undoubtedly of Muhammadan origin. Its alleged derivation from

Samdhimat, the saintly hero of a well-known legend recorded in the

Rajatarangini, is nothing but an invention of the Bachbattas of S'ri¬

nagar.4

1 Vikram. xviii. 23.

2 Compare Rdjat. vi. 172-173 note.

8 S'riv. iv. 306. The Lokaprakasa writes Phagvd while the modern Mahatmyas

of Isalaya or I4^bar and Suresvarl affect the form Phdlaka.

* The name Takht-i-Sulaiman is common enough in the local nomenclature of

Muhammadan countries; compare, e.g ., the peak of this name in the Sulaiman

158

ANCIENT GEOGRAPHY OF KASHMIR. [Extra No. 2,

That the ancient designation of the hill was Gopadri is shown

beyond all doubt by an interesting passage of Kalh ana’s Chronicle. It

relates how the troops of the pretender Bhiksacara when thrown back

from the city which they had endeavoured to enter after crossing the

Mahasarit, i.e., from the south-east, took refuge on the ‘ Gopa hill’ or

Gopadri.1 There they were besieged by the royal troops until a diversion

made by Bhik§ficara enable them to retreat to the higher hills in the

east by the low neck which connects these with the Takht-i Sulaiman.

Kalhana in the First Book of his Chronicle informs us that King

Gopaditya built a shrine of S'iva Jyesthesvara on the Gopadri.2 It is

difficult not to connect this notice in some way with the extant temple

which occupies so prominent a position on the summit of the hill. Gene¬

ral Cunningham, it is true, on the strength of an alleged tradition had

proposed to identify this temple with the Jyestharudra shrine which

Kalhana mentions as a foundation of Jalauka, Asoka’s son, in the

ancient STiuagari.3 But Prof. Biihler has already shown that there is

no genuine tradition regarding the temple among the S'linagar

Brahmans.4

It is certain that the superstructures of the present temple belong

to a late period.6 But the massive and high base on which this

temple is raised, and certain parts of the structure are no (loubt of a far

earlier date. These may well have formed part of a building which

in Kalhana’s time, — rightly or wrongly, we have no means to judge, —

was looked upon as a shrine of Jyesthesvara erected by King Gopaditya.

There is no other ancient ruin ou the hill. Nor would the configuration

of the latter have admitted at any other point but the summit, of the

construction of a shrine of any dimensions. It is of interest to note

that the tradition of Abu-l-Fazl’s time distinctly attributed the temple

standing on ‘ Solomon’s hill ’ to the time of Gopaditya.6

Koh, S. of the Gumal Pass. The derivation from Samdhimat, referred to by Prof.

Buhler, Report, p. 17, is not supported by any evidence whatever and unknown

even to the most modern Mahatmyas.

1 See Rajat. viii. 1104-10 note. That the Taklit-i -Sulaiman was called by its

ancient name Gopadri, had been surmised already by Pt. Govind Kaul at the time

of Prof. Biihler’s visit; see Report, p. 17. But the decisive evidence of this passage

was not known to him.

2 See i. 341 and note.

a Rajat. i. 124 ; Anc. Geogr., p. 95 ; also above, § 90.

* See Report, p. 17.

6 See the remarks of Fergusson, History of Indian Archit., p. 282, against Gen.

Cunningham’s and Major Cole’s assumptions who represented the extant temple as

one of the earliest buildings in Kasmlr.

6 Ain-UAkb., ii. p. 383.

1899.]

THE ENVIRONS OF S'RINAOARA.

159

102. In my note on Rajat. i. 124 I have shown that an old tradi¬

tion which can he traced back to at least the

sixteenth century, connected the Takht hill with

the worship of S'iva Jyestharudra or, by another

Tlrtha of Jyesthe§-

vara.

form of the name, Jyesthesvara ( Jyestliesa).1 And we find in fact a

Liiiga known by this name worshipped even at the present day at

the Tlrtha of Jyether, scarcely more than one mile from the east foot of

the hill.

Tliis Tirtha which undoubtedly derives its name from Jyesthesvara,

lies in a glen of the hill-side, a short distance from the east shore of the

Gagri Bal portion of the Dal.2 3 Its sacred spring, designated in the

comparatively modern Mahatmya as Jyesthanaga , forms a favorite

object of pilgrimage for the Brahmans of STinagar. Fragments of

colossal Lingas are found in the vicinity of Jyether and show with

some other ancient remains now built into the Ziarats of Jyether and

Gupakar, that the site had been held sacred from an early time.

It is in this vicinity that we may look for the ancient shrine of

Jyestharudra, which Jalauka is said to have erected at STinagarl. But

in the absence of distinct archaeological evidence its exact position

cannot be determined. It is highly probable that whatever the

origin and the date of the temple on the Takht hill may be, it was

connected with the worship of Jyestharudra at Jyether. No other

Tirtha is known in the immediate neighbourhood.

The distance of the shrine from the Tirtha is scarcely greater than

that of Lalitaditya’s temple at Mar tan d from the sacred spring in

honour of which it was erected. And in both places the distance of

the temple is easily accounted for by the more prominent position

which was thus secured for it. There is yet another analogy in the

case of the two shrines. Both have long ago ceased to be places of

popular worship. But the sacred springs, to the presiding deity of

which they were dedicated, continue to attract pilgrims though shorn

of all splendour of temples and images.

Kalhana in the same passage which mentions the erection of

haras ’ on Brahman settlers from Aryade^a.8

The combination of the two local names suggests that by the Gopa-

1 Compare Fourth Chron. 592, 853, 806.

2 For Jyesthefoara > Jyether we have exact analogies in Kapa(e6varci > Kdther,

Amarefvara > Amburher , etc.

3 See i. 341. Agrahara is the regular term designating a Jaglr or piece of land

bestowed on individuals or religious corporations, etc, ; see note i. 87,

160

ANCIENT GEOGRAPHY OF KASHMIR.

[Extra No. 2,

graharis are meant the ferti le lands of the present Gupahdr , between

the north foot of the Takht hill and the Dal. The name Gupakar may

be, in fact, the direct phonetic derivative of the term used by Kalhana.*

Our surmise is supported by the reference which Kalhana in the

verse immediately following makes to the village BhuksIravatika.

This place is identified by the old glossator A?j with Buch'vor, a small

hamlet situated on the narrow strip of land at the rocky north-west

foot of the Takht hill. The modern name is clearly derived from

Kalhana’s form. Gopaditya is said to have removed to this confined

and secluded spot Brahmans who had given offence by eating garlic.

The combined mention of Gopadri, Gopagrah ara and Bhuksira-

vatika in Rajat. i. 341 sq. suggests that Kalhana has reproduced here

local traditions collected from the sites immediately adjoining the hill.

Whether the connection of these localities with King Gopaditya’s reign

was based on historical fact, or only an old popular etymology working

upon the word Gopa found in the first two names, can no longer be

decided.

Continuing our route along the eastern shore of the Dal we come,

at a distance of about one mile from Gupakar, to the large village

of Thid, prettily situated amid vineyards and orchards. It is the Theda

of the Rajatarangini, mentioned as one of the places which the pious

King Samdhimat or Aryaraja adorned with Math as, divine images,

and Lingas.2 Abu-1-Fazl speaks of Thid as “a delightful spot where

seven springs unite ; around them are stone buildings, memorials of

by-gone times. ”s The remains here alluded to can no longer be traced,

but the seven springs ( S aptapuskarinl ) which are also referred to in

the Haracaritacintamani (iv. 40 sqq .), are still pointed out.

The cluster of villages which we reach about one and a half miles

beyond Thid, and which jointly bear the name Bran, can be safely

identified with BhImadevI which Kalhana notices along with Theda.

The Nilamata knows the sacred site of Bhimadevi in conjunction with

the Suresvari Tirtha which we shall next visit, and in the Haracarita¬

cintamani it is named with the seven springs of Theda. The Tirtha

of Bhimadevi is no longer known, but may be located with some pro¬

bability at the fine spring near Ddmpor marked now by a Muhammadan

shrine.

I Gupakar may go back to a form # Gup^gar, with assimilation of g to the

preceding tenuis. In Ks. the hardening of g to fe is by no means unknown, see

Dr. Grierson’s remarks, Z.D.M.G., 1., p. 3. * GupQgdr could easily be traced back to

Gnpdgrahara through Pr. forms like # Gupagrdr.

* See Rdjat. ii. 135 note.

8 Atn^i-Akb., ii. p. 361.

1899.]

THE ENVIRONS OF S'RINAGARA.

161

103. A sacred site of far greater fame and importance is that of

Tirtha of Suresvari. the present village °f U'h-r which lieS ab°Ut

two miles further north on the Dal shore and

a little beyond the Mughal garden of Nishat. The site was known in

ancient times as Suresvariksetra (‘the field of Suresvari’).1 It was sacred

to Durga- Suresvari who is still worshipped on a high crag rising from

the mountain range to the east of Isobar village. The seat of the

goddess is on a rugged rock some 3000 feet above the village, offering no

possible room for any building. The numerous shrines erected in her

honour were hence built on the gently sloping shore of the lake below.

The Tirtha of Suresvari is often referred to in Kalhana’s Chro¬

nicle and other Kasmirian texts as a spot of exceptional holiness. It

was particularly sought by the pious as a place to die at. The pilgrim¬

age to Suresvari is connected with visits to several sacred springs in

and about Isobar. One of them, S'atcidhara, is already mentioned by

Ksemendra.2 It is passed in a narrow gorge some 1500 feet below the

rock of Suresvari.

Isobar derives its present name from the shrine of Isesvara which

King Samdhimat-Aryaraja according to the Rajatarangini erected in

honour of his Guru Isana.3 An earlier form, Istbror, which is found in

an old gloss of the Chronicle and evidently was heard also by Abu-1-Fazl,

helps to connect Isobar and Isesvara. 4

Isobar is still much frequented as a pilgrimage place. The chief

attraction is a sacred spring known as Guptaganga which fills an ancient

stone-lined tank in the centre of the village. This conveniently acces¬

sible Tirtha is the scene of a very popular pilgrimage on the Vaisakhi

day and has fairly obscured the importance of the mountain seat of

Suresvari. A ruined mound immediately behind the tank is popularly

believed to mark the site of the Isesvara shrine. Numerous remains

of ancient buildings are found around the sacred springs and elsewhere

in the village. They probably belong to the various other temples

the erection of which is mentioned by Kalhana at the site of Suresvari.5

Passing round the foot of the ridge on which Suresvari is

worshipped, we come to the small village of

■*. , Harvan which the old glossator ot the tvajata-

i ripurGsv^rSi. $ #

rangini identifies with Sadarhadvana the

1 Compare for Sureivarl and the site of Isobar, note v. 37.

2 See Samay. ii. 29.

8 See Rdjat. ii. 134 note.

* -bar is a modern contraction for -brdr, from Skr. bhattaraka , which in Kasmlr

local names has often taken the place of its synonym - Uvara ; comp, e.g., Skr.

Vijayeivara > Ks. VijQbrdr.

5 See Rdjat. v. 37, 40 sq , ; viii. 3365.

J. i. 21

162

ANCIENT GEOGRAPHY OP KASHMIR.

[Extra No. 2,

wood of the six Arliats ’). Tliis place is mentioned by Kalhana as the

residence of the great Buddhist teacher Nagarj un a.1 The name Harvan

may well be derived from Sadarhadvana, but in the absence of other

evidence the identification cannot be considered as certain. On the

hill-side south of the village I observed already in 1888 fragments of

ornamented bricks. Since then remarkable remains of ancient brick-

pavements have come to light on occasion of excavations made for the

new Srinagar waterworks.

Proceeding further up the valley of the stream which comes from

the Mar Sar lake, we reach, at a distance of about three miles from the

Dal, the village of Triphar. Evidence I have discussed elsewhere,

makes it quite certain that it is the ancient Tripuresvara (Tripuresa).2

The latter is repeatedly mentioned as a site of great sanctity by Kalhana

as well as in the Nilamata and some Mahatmyas. But it has long ago

ceased to be a separate pilgrimage place. A little stream known as the

Tripuragahga near Triphar is, however, still visited as one of the stations

on the Mahadeva pilgrimage.

Ksemendra in the colophon of his DaSavataracarita refers to the

hill above Tripuresa as the place where he was wont to find repose and

where he composed his work. In Zain-ul-‘abidin’s time Tripuresvara

seems yet to have been a Tirtha much frequented by mendicants.3

Tripuresvara too possessed its shrine of Jyesthesvara, and to this King

Avantivarman retired on the approach of death.4 A legend related by

the S'arvavatara connected the site of Tripuresvara with the defeat of

the demon Tripura by S'iva and with the latter’s worship on the

neighbouring peak of Mahadeva. I have not been able to examine

the site and am hence unable to state whether there are any ancient

ruins near it.

The whole mountain-ridge which stretches to the south of Triphar

and along the Dal, bore in ancient times the name of S'ridvara.5 On

the opposite side of the Valley rises the great peak of Mahadeva to a

height of over 13,000 feet. Numerous references to it in the Nilamata,

S'arvavatara, and other texts, show that it was in old times just as now

frequented as a Tirtha.

We may now again descend the valley towards the north shore of

the Dal. On our way we pass close to Harvan the village of Tsatsa where

the convenience of modern worshippers has located a substitute for the

1 See Rajat. i. 173 note.

8 Compare Rajat. v. 46 note.

8 See S'riv. i. 402.

4 See Rajat. v. 123 note.

8 See Rajat, viii, 2422.

THE ENVIRONS OF S'RlNAGARA.

163

1899.]

Hiraiiyapura.

ancient Tirtha of the goddess S'arada (see below § 127). Leaving aside

the famous garden of Shalimar of which our old texts know nothing,1 *

we come to a marshy extension of the Dal known as Telabal. The

stream which flows through it and which forms a branch of the river

coming from the Mar Sar, bore the old name of Tilaprastha ,a

104. The road which takes us from Telabal to the mouth of the

Sind Valley is the same which was followed

by the pretender Bhiksacara and his rebel

allies on a march to Suresvari described in the Rajatarangini.3 The

narrow embankment on which they fought and defeated the royal

troops, leads across the Telabal marshes.

At the south foot of the ridge which runs down to the opening of

the Sind Valley, we find the village of Rgnyil , the ancient Hiranyapura.4 *

The place is said by Kalhana to have been founded by King

Hiranyaksa. As it lies on the high-road from the Sind Valley to

S'rinagar it is repeatedly mentioned also in connection with military

operations directed from that side against the capital. The victorious

Uccala when marching upon S'rinagar, had the Abhiseka ceremony

performed en route by the Brahmans of Hiranyapura. It seems to have

been a place of importance, since it figures in a fairy-tale related in the

Kafchasaritsagara as the capital of Ka^mir.6 A spring a little to the

south of the village is visited by the pilgrims to the Haramukutaganga

and bears in Mahatmyas the name of Hiranyaksanaga.

From near Ranyil several old water-courses radiate which carry the

water of the Sind River to the village lying

between the Anctfiar and the Dal lakes. One

of these canals passes the village of Zukur.

A tradition recorded already by General Cunningham identifies this

place with the ancient Juskapura. Kalhana names the place as a

foundation of the Turuska (i.e. Kusana) King Juska who also built a

Vihara there.6 The Muhammadan shrines and tombs of the village

contain considerable remains of ancient buildings.

Juskapura ;

Amaresvara.

1 The first reference to this somewhat over-praised locality which I can find,

is in Abu-1-Fazl who mentions the waterfall or rather the cascades of ‘ Shalahmar ’ ;

see ii. p. 361. The Vitasta-, Isalaya-, and Mahadeva-Mahatmyas which are of very

modern origin, show this faet also by their references to * S'dlamara’ and the whimsical

etymologies which they give for the name (Mdrafald, etc.). We might reasonably

expect that Jonaraja and S'rivara in their detailed accounts of the pal would have

mentioned the place if it had then claimed any importance,

3 See Rdjat. v. 46 note ; S'riv. i. 421.

& See Rdjat. viii. 744 note,

4 For detailed references see Rdjat. i. 287 note.

6 See Kathdsar. lxv. 215 sqq.

6 See Rdjat, i, 168 note ; Anc, Qeogr, p. 101.

164

ANCIENT GEOGRAPHY OF KAS'MTR. [Extra No. 2,

To the west of Juskapura and on the shore of the Anchor lies the

large village of Amburher. It is the ancient Amare6vara often men¬

tioned in the Rajatarahgini in connection with military operations to

the north of SVinagar.1 This is easily accounted for by the fact that

the place lay then as now on the high road connecting the Sind Valley

with the capital. It took its name from a temple of S/iva Amaresvara

which Suryamati, Ananta’s queen, endowed with Agraharas and a Matha.

The ancient slabs and sculptured fragments which I found in 1895 in

and around the Ziaiat of Farrukhzad Sahib, may possibly have belonged

to this temple.

Continuing on the road towards SMnagar for about two miles

further we come to the large village of Vicar Nag prettily situated in

extensive wallnut groves. A fine Naga near the village forms the

object of a popular Yatra in the month of Caitra. It is supposed to be an

epiphany of the Ailapattra Naga who is mentioned also in theNilamata.

An earlier designation seems to be Muktamulakanaga which is given to

the locality by S'rlvara and in the Tirthasamgralia.2 * To the west of

the village and near an inlet of the Anctdar are the ruins of three

ancient temples now converted into Ziarats and tombs.8

Only a quarter of a mile to the east of Vicar Nag and on the other

side of the old canal called Lacham Kul

(*Laksmikulya) stands the hamlet of Anta-

bavan. In my “Notes on Ou-k'ong’s* account of Kasmir” I have proved that

Antabavan derives its name from the ancient Vihara of Amrtabhavana

which Amrtaprabha, a queen of Meghavahana, is said to have erected.4 *

Ou-k'ong mentions the Vihara by the name of Ngo-mi-t' o-po-wan which

represents a transcribed Prakrit form *Amitabhavana or Araitabhavana.

An ancient mound with traces of a square enclosure around it, which is

found between the canal and the hamlet, may possibly belong to the

remains of this Vihara.

A

Proceeding to the east of Antabavan for about a mile we come to

the large village of Sudarabal situated on a

deep inlet of the Dal, known as SudarVkhun.

The name of the village and the neighbouring portion of the lake make

it very probable that we have to place here the sacred spring of

Sodara.6 It formed the subject of an ancient legend related by

Amrtabhavana.

TIrtha of Sodara.

1 See Rdjat. vii. 183 note.

2 See S'riv. iv. 65. On his authority the name Muktamulakanaga ought to have

been shown on the map.

s Compare for a view of these remains, Cole, Ancient Buildings , p. 31.

4 See Rdjat. iii. 9 note, and Notes on Ou-k'ong , pp. 9 sqq.

* See Rdjat. i. 125-126 note. Ks. -ba? in Sudarabal means merely ‘ place.’

ib99.]

NORTHERN DISTRICTS OF MADAVARAJYA.

165

Kalhana. King Jalauka had built at S'rinagari his shrine of Jyestha-

rudra whose original place of worship was at Bhutesvara, below

Mount Haramukuta, He then wished to have by the side of the new

shrine also the Sodara spring which adjoins the site of Bhutesvara.1

To fulfil the king’s pious desire “ there broke forth from a waterless

spot a spring which was alike to Sodara in colour, taste and other

respects.” A golden cup thrown into the original Sodara spring appear¬

ed after two and half days in its Avatara near STinagaii. This miracle

removed all doubts as to their identity.

Close to the mosque of Sudarabal and by the lake shore are two

pools fed by perennial springs. These according to a local tradition

were in old times visited by numerous pilgrims. Now all recollection

of this Tirtha has been lost among the Brahmans of STinagar. But the

name of a portion of the village area, Battvpor, points to a former

settlement of Battas or Purohitas. It is curious, too, that we find only

half a mile from the village the Ziarat of Hazrat Bal, perhaps the most

popular of all Muhammadan shrines in the Valley. It is supposed to

be built over the remains of the miracle-working Pir Dastagir Sahib.

Is it possible that the presence of this rather ubiquitous saint at this

particular spot had something to do with the earlier Hindu Tirtha P

Section VI. — Northern Districts of Madavarajya.

105. Our circuit through the Phakh Pargana has brought us back

to the purlieus of the capital. We must leave them now once more

and start on our tour through the outlying districts. We may direct

it first to the upper half of the Valley, the ancient Madavarajya.

This again is divided by the Vitasta into two portions, one to the north

and east, the other to the south and west of the river. We shall begin

with the Parganas on the right bank, starting from S'rinagar.

The Pargana which adjoins $Tinagar from the south-east, is now

District of Khaduvl known as Vihi. It extends from near Parana-

dhisthana to the spur of Vastarvau, near

Vantip5r ( Avantipura), aud comprises a wide semi-circular tract of

fertile Karewa lands. In ancient times the district took its name from

the village of Khaduvi, the present Khruv .2 The Damaras of the

Khaduvl district are repeatedly mentioned by Kalhana along with those

of Holada, the modern Vular Pargana.

1 For Sodara , the present Naran Nag, see notes i. 123 ; v. 55-59.

2 Compare Raj at. viii. 733 note.

166

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

The site of Pandrethan or Puranadhisthana has already been fully

noticed. About two miles higher up the river lies Pand^chuk village,

with some ancient remaius and the traces of a stone bridge-head, pro¬

bably of late date. The old name of the place is unknown. We pass

next by the river the village of Simpor. This may retain the name of

►Simhapura, founded by King Jayasimha in Kalhana’s time.1

Less than two miles to the north-west of Simpor lies the village of

Zevan , the ancient Jayavana. It was correctly identified already by

Prof. Biililer on the basis of the happy and exact description given of it

by Bilhana.2 The poet mentions in this “ place of high-rising monu¬

ments ” the “pool filled with pure water, sacred to Taksaka, lord of

snakes.” This pool still exists in the Taksaka Naga which is visited

annually by the pilgrims to Harsesvara.8

The mention made by Kalhana in his history of Narapura of the

pilgrimage to the Taksaka spring proves that in old times it must have

enjoyed great reputation as a separate Tirtha. It is in fact the only

Kasmir Naga which is distinctly mentioned in the Tirtha list of the

Mahabharata (iii. lxxxii. 90). Abu-1-Fazl records the interesting fact

that this spring was populary held to be the place whence the culti¬

vation of saffron flourishing in this neighbourhood originated.41 In

Akbar’s time the cultivators, undoubtedly Muhammadans, still wor¬

shipped at this fountain at spring time. It was customary to pour

cow’s milk into it to secure a good omen for the success of the crop.

We see that the Taksaka Naga long retained his importance with the

cultivators as a local divinity.

About two miles to the north-east of Zevan we come on gently

Khonamusa rising ground to the village of Khuntmoh. It

is, as already stated above, the ancient Khona¬

musa, famous as the birthplace of Bilhana. The latter in the Vikramaii-

kadevacarita gives an enthusiastic description of the charms of his rural

home.6 The saffron fields which Bilhana mentions extend close to the

lower of the two separate hamlets which form the village. In the

upper hamlet is a sacred spring called’ Damodaranaga with some sculp¬

tured funeral. Steles and a few fragmentary inscriptions.

On the hill-side above the village issues another Naga which under

1 See Rdjat. viii. 2443 note.

8 Compare Report , pp. 5 sq. ; Rdjat. vii. 607 note ; Vikram. xviii. 70.

s See Rdjat. i. 220 note.

* See Am-i-ATcb., ii. p. 358.

6 For a detailed and accurate account of the position and remains of Khonamu$a,

see Prof. Bugler’s Report , pp. 5 sq. The identity of Khun^moh with the Khona¬

musa of Rdjat. i, 90, was first pointed out by Gen. Cunningham, Anc, Geogr., p. 98.

1899.]

NORTHERN DISTRICTS OF MADAVARAJYA.

1G7

Padmapura.

the name of Bhuvaneharl is visited on the pilgrimage to Harse^vara.

The latter Tirtha lies on the summit of the high ridge which rises to

the north of the village. It consists of a ‘ Svayambhu’ Linga situated

in a small cave and enjoys considerable popularity. I have not been

able to trace its name except in the local Mahatmya and the Tirtha-

samgraha.

The chief place of the Yihi Pargana is now the town of Pampar,

the ancient Padmapura, about four miles south¬

west of Khun^moh.1 It was founded in the

beginning of the 9th century by Padma, the powerful uncle of the

puppet king Cippata-Jayapida. Padma is said by the Chronicle to have

also built a temple of Visnu Pctclmasvamin. To this may possibly

belong the scanty remains of an ancient temple which have been de¬

scribed by General Cunningham.3 Close by is the Ziaratof Mir Muham¬

mad Hamadani, with some fine ancient columns and ornamented slabs

which are likely to have been taken from this temple. Also the other

Ziarats of the town show similar remains. Padmapura, owing to its

central position in a fertile tract, seems to have always been a place of

importance and is often mentioned by Kalhana and the later Chroni¬

clers.

Proceeding north-eastwards of Padmapura we pass first Balvhom, a

large village, which in the Lokaprakasa and Tirthasamgraha figures as

Balasramci. Under a large Deodar near it Baladevi is now worshipped

in the form of an old stone-image. Numerous ancient Steles, showing

miniature reproductions of temples, are found in the neighbouring

rivulets and canals ; they were apparently used in recent times as

stepping-stones which would account for their preservation. At the foot of

a rocky spur which descends from the mountain-range to the north, lies

the picturesque village of Uyan , once mentioned by Kalhana under the

name of 0 van a. 3 It has a large sulphurous spring visited by the sick.

About two miles further east we reach the large village of Khruv ,

the ancient KhaduvI which, as we have seen, gave to the district its

former name. There is an abundance of fine springs in and about

Khruv ; Abu-1-Fazl mentions them as objects of worship and estimates

their number at 360. 4 Above the village a so-called Svayambhu-cakra

or mystical diagram is shown on a rock.5 It is held sacred to Jvala-

1 For a detailed notice see Rdjat. iv. 695 note. The old name of the place is

well-known to S'rinagar Pandits; Vigne too, Travels, ii. p. 31, recognized it correctly.

2 See J. A. S. B., 1848, p. 274.

3 See Rdjat. vii. 295.

A Ain-i-Akb., ii. p. 358.

6 Compare for such diagrams, also designated Devicajcra or MdtrcaJcra, Rdjat. i.

122 note.

168

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

&anara.

muklii-Durga and largely visited by pilgrims. I am, however, unable

to trace any old reference to this Tirtha.

Only a mile to the south-east of Khruv is the village of Sdr, until

recently the seat of a flourishing iron-industry.

Kalhana mentions it by the name of S'anara as

an Agrahara founded by King S'aclnara.1 Whatever the historical

value of this notice may be, which Kalhana took from Padmamihira,

the evidence detailed in my note on the passage proves that the present

S'ar is intended. An intermediate form of the name is preserved in the

S'nar of an old gloss. The Ziarat of Khwaja Khizr which stands here

near several small springs, is built with remains of a Hindu temple;

among them, is a Lihga-base some six feet square.

About two miles south-west of S'ar are found the well-preserved

ruins of a temple near the village of Ladu (not marked on Survey map).

They have been described by Bishop Cowie, 2 3 but I am unable to trace

any old reference to this shrine in the texts I have examined. It is

remarkable for having a circular cella, the only one known to me in

Kasim r. A small square cella to the east of this temple has been

annexed to a neighbouring Ziarat.

Near the south end of the Vihl Pargana and on the river lies the

village of Lcitvpor. An old gloss of the Rajatarangini identifies it with

Lalitapura, a place founded in honour of King Lalitaditya by his

architect.8 The King according to the Chronicle was not pleased with

the attention ; in any case no importance seems to have attached to this

place. There are no old remains above ground, but the local tradition

still tells of King ‘ Lalitadit * as the founder of a large town on the

neighbouring Udar.

106. Passing round the foot of Mount Vastarvan we enter the

Pargana of Vular, the ancient Holada. This

identification is supported, apart from the

clear phonetic evidence, by all passages of the

Rajatarahgini which mention Holada.4 * Its feudal barons played a

great part in the troubled times of the later Hindu reigns.

Its most important place in old times was undoubtedly the town of

Avantipura, founded by King Avantivarman (a.d. 855-883). 6 Its posi¬

tion is marked by the present village of Vaut'por on the Vitasta. The

District of Holada ;

Avantipura.

1 See note i. 100.

2 See J. A. S. B., 1866, pp. 97 sqq.

3 See Rajat. iv. 186.

4 See Rajat. i. 306 note.

6 See Rajat. v. 45 sq. note. Its identity with Yantipdr was first pointed out by

Dr. Wilson in liis note on Moororoft, Travels , ii, p, 244.

1899.]

NORTHERN DISTRICTS OF MADAVARAJYA.

169

conspicuous ruins of this place attracted already the attention of the

early European visitors. General Cunningham did not fail to recognize

in them the remains of the two great temples of Avantisvamin and

Avantisvara which Avantivarman had built here.1 Of the two great

ruins one is at Van^por itself, the other and larger one half a mile

further down the river close to the hamlet of Jaubrar (map ‘ Jabair.’)

Owing to the complete destruction of the central shrines it is im¬

possible to ascertain now which was dedicated to Visnu and which to

Sfiva. The fine enclosing quadrangles of the temples have also suffered

badly. That of Avantisvamin was used as a temporary fortification in

Kalhana’s own time and underwent a severe siege.2

The site on which Avautivarman’s town was built, had apparently

enjoyed some sanctity before these temples were founded, and bore the

old name Visvaikasara. The great extent of the town is indicated by the

traces of ruined buildings which cover the foot of the hills east of

V antfpor for a considerable distance. The frequent references to

Avantipura in the Chronicles show that the town retained some impor¬

tance long after the death of its founder.

We hear but little of other old sites in Holada. The great town of

Mihirapura which King Mihirakula is said to have founded in it, can no

longer be traced.3 Khuli, a village situated a short distance to the

south-west of Tral, the modern headquarters of the district, may

possibly be the Khola of the Chronicle, one of Gopaditya’s Agraharas.41

Of Tral I am unable to trace any old notice.

The identification of the village of Buis , about two miles south of

Khuli, with the old Bhayaccheda is also uncertain.6 It is based on a

gloss of Rajanaka Ratnakantha, the writer of the Codex Archetypus of

the Rajataranginl. Still further south lies the village of Kai, probably

the old Katika named by Kalhana as a foundation of Tunjina I.6 This

identification is made in the old gloss on the passage and supported by

the phonetic evidence of the modern name.

Of old remains in Vular the interesting temple of Narastan at the

northern extremity of the district (34° 3' lat. 75° 10' long.) deserves

notice.7 Unfortunately I am unable to find any clue as to its old name

1 See for a full description J. A. 8 . B., 1843, pp 275 sqq. ; also ib., 1866, 121 sqq.

8 See Rajat. viii. 1429 sq., 1474 sqq.

5 See Rajat. i. 306.

A See Rajat. i. 340.

6 Compare iii. 381 note.

^ Rajat. ii. 14.

1 See Mr. Lawrence’s notice, Valley , p. 172. The attached photograph shows

the site after my excavations. Regarding the result of the latter, see Vienna

Oriental Journal , 1891, p. 345 sqq.

J. i. 22

170

ANCIENT GEOGRAPHY OF K AS' MIR. [Extra No. 2,

or history. Excavations made by me at the site in 1891 brought to light

interesting sculptures, but no evidence as to its name. The large

village of Sutur (map ‘ Sootoor ’) to the south-west of Narastan may

possibly account for the entry of Satrava in the Lokaprakasa’s list of

Parganas.

107. The eastern boundary of Vular is marked by the high spur

which descends from the north towards the

District_of confluence of the Vitasta and Gambhira. The

para.

adjoining district to the east is one of consider¬

able extent. It comprises besides the whole right or western side of

the Lidar Valley also the low-lying tract between the Vitasta and the

lower course of the Visoka. The modern name of this great Pargana is

Dachunpor which appears in S'rivara’s Chronicle as Daksinapara. This

clearly means ‘ the right bank ’ [of the Ledarl or Lidar]. Another form,

of the same significance, is Daksinapdrs'va found in the Lokaprakasa and

Marta ndamahatmya. To this designation corresponds the term Vdma-

pdrsva, now KhovurpSr, which as we shall see, is applied to the left side

of the Lidar Valley.1

The junction of the Vitasta with the Gambhira, i.e., the united

Visoka and Ramanyatavi, has already been mentioned above as a

Tirtha.2 Not far from it lies the village of Marhom , on the Vitasta,

mentioned by Jonaraja under its old name of Madavasrama.3 The first

part of the name is identical with that of Madavardjya.

About two miles south-east of Marhom and not far from the

Vitasta, we find the village Vag^hom, with a sacred spring known by the

name of Hastikarna. This name seems to have applied formerly to the

place itself which we find twice thus referred to by Kalhana.4 It is

possibly the Hastikarna, where Blioja, Harsa’s son, was treacherously

murdered.

About one mile to the south of Hastikarna the Vitasta makes a

great bend. The peninsula thus formed is

occupied by a small Udar or alluvial plateau

which owing to its height and isolated position

is a very conspicuous object in the landscape. It was once the site of

Temple of

Cakradhara.

1 General Cunningham, Anc. Geogr., p. 94, assumes that Ks. dachun ‘ right ’ is

‘ now used to denote the “ north,” and kawar, (recte khovur) or “ left,” to denote

the “ south.” ’ This assumption, however, as well as the explanation given for

the alleged change of meaning are based on some misunderstanding.

2 See § 64.

3 See Jonar. (Bo. ed.), 132.

4 See Rdjat. v. 23 note ; also vii. 1650. Another Hastikarna , mentioned by

Sri vara, i. 441, seems to have been near S'rinagar on the west.

1899'.]

NORTHERN DISTRICTS OF MADAVARAJYA. 171

one of the oldest and most famous shrines of the Valley, the temple

of Visnu Cakradhara.

The plateau is still known as Tsakadar Udar.1 Brahman tradition

is aware of the derivation of this name from Cakradhara. It was first

brought to the notice of European scholars by Prof. Buhler who had

duly recognized the antiquarian importance of the site.2 The shrine of

Cakradhara is often mentioned as a Tirtha of great sanctity.3 It was

also closely connected with the legends regarding the burned city of

Narapura , localized as we shall see in its close vicinity. But the only

detailed notice of the temple we owe to a historical incident which

occurred there during the civil wars of Sussala’s reign.4 *

The royal troops having been forced to evacuate the neighbouring

town of Vijayesvara or Vij^bror, the inhabitants of the latter place and

the neighbouring villages took refuge in the temple of Cakradhara. This,

by its position on the high and steep Udar, was naturally well-adapted

for defence. The temple filled by the crowd of fugitives and routed

soldiers, was soon besieged by the rebel troops of Bhiksacara. The

temple courtyard was protected by massive wooden ramparts and gates.

When these had been set on fire by the assailants a mighty conflagration

ensued in which the whole mass of people inside perished. Kalhana

vividly describes this great catastrophe which he believes to have

provoked divine vengeance and thus to have brought about the downfall

of the pretender.

The account here given is of topographical interest. It showTs that

the temple actually stood on the flat top of the Udar, and also explains

the scarcity of stone-remains in this locality. The absence of conspicu¬

ous ruins had already been noticed by Prof. Buhler. When visiting

in 1895 the Tsakadar Udar, I found only traces of a quadrangular

enclosure, about forty yards square. They are marked by hollows at

the northern end of the Udar which is separated from the rest by a

dip iu the ground. These hollows may possibly be the last indications

of the wooden ramparts which enclosed the shrine.

The temple seems to have been subsequently restored, and*

Jonaraja mentions the statue of Cakradhara among those chief divine

images which Sikandar Butshikast destroyed.6 Jayadratha in his

1 See Rajah i. 38, 201 notes.

2 See Report , p. 18.

3 See Rdjat. vii. 258, 261, 269 j Jonar. (Bo. ed.), 763 j S'rikanthac. iii. 12 ; Nila*

mata, 1170.

4 See Rdjat, viii. 971-995. The date of the burning of Cakradhara seems to

have been the 12th S'ravana sudi, a.d. 1121.

6 See Jonar, (Bo. ed.), 763.

172

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

Haracaritacintamani devotes a separate canto, vii , to the relation of

the legend which localized the disc-wielding god at the Tirtha of

Cakradhara. The latter is still referred to in a general way in the old

Vijayesvaramahatmya (No. 87, Poona MSS.). Now, however, Cakradhara

is no longer visited by the pilgrims to Vijayesvara though the Purohitas

of the latter place still retain a recollection of the former sanctity of

the site.

Legend of Narapura.

108. There can be no doubt that at the foot of the Cakradhara

Udar there stood once an ancient town of con¬

siderable importance. From the low ground

towards the river on the east and from the river-bed itself, ancient coins

going back to Creek and Indo-Scytkian rule are annually extracted in

considerable quantities. Popular tradition still asserts that this site was

once occupied by a great town. This tradition existed already in the

time of Kalhana who records it in the interesting legend of the burned

city of Narapura.1 This is told at great length in a poetic episode of

the First Book.

King Nara is said to have founded a splendid capital, called after

himself Narapura, on the sandy bank of the Vitasta close to the shrine

of Cakradhara. “ There in a grove was a pond of limpid water, the

habitation of the Naga Susravas.” A young Brahman who had found

occasion to assist the Naga and his two daughters when in distress, was

allowed to marry in reward one of the latter. He lived in happiness

at Narapura until the beauty of the Naga lady excited the passion of

the wicked king. When Nara found his advances rejected with scorn,

he endeavoured to seize the beautiful Candralekha by force. The

couple fled for protection to their father’s habitation.

The Naga then rose in fury from his pool and “burned the king

with his town in a rain of fearful thunderbolts.” Thousands of people

were burned before the image of Visnu Cakradhara to which they had

fled for protection, Ramanya, the Naga’s sister, came down from the

mountains carrying along masses of rocks and boulders. These she

dropped, as we have seen, along the bed of the Ramanyatavl or

Rembyar^, when she found that Susravas had already wreaked his

vengeance. The Naga himself feeling remorse at the carnage he had

caused, removed to a lake on a far-off mountain. There “ he is to the

present day seen by the people on the pilgrimage to Amaresvara.” 2

“ To this day,” thus closes Kalhana’s narration, “ that tale is remembered

by the people when they behold close to Cakradhara that town des¬

troyed by fire and that pond which has become a dry hollow.”

1 See Bdjat. i. 201-274.

* Compare regarding the lake of the Naga Sufravas on the route to Amburnath,

above, § 59.

1899.]

NORTHERN DISTRICTS OF MADAVARAJYA.

173

Vijayesvara.

Whatever the origin of the legend here told may have been, it is

clear that popular tradition in Kalhana’s time looked upon the barren

ground which stretches along the river between Tsakadar and the

present Vij^bror as the site of an ancient city. The ruins which in the

12th century were pointed out as the remains of the burned Narapura,

may have supplied the immediate starting point of the legend.

What these remains were we cannot say. As the ground referred

to is subject to annual inundation it is possible that the remains meant

have since disappeared under alluvial deposits. The habitation of the

‘ Susram ’ Naga was still shown to me in a hollow, generally dry, close

to the south-east foot of the Udar. The name of Narapura and its king

are no longer remembered. But the main features of the legend as

heard by Kalhana, still live in the local tradition.

109. The ancient town which once stood in the position indicated,

was evidently succeeded by Vijayesvara, the

present Vij^bror. The latter place, situated

less than two miles above Cakradhara, received its name from the

ancient shrine of S'iva Vijayesvara (Vijayesa, Vijayesana).1 This

deity is worshipped to the present day at Vij^bror. The site has evi¬

dently from early times been one of the most famous Tirthas of Kasmir.

It is mentioned as such in the Rajatarangini and many old Kasmirian

texts.2 The tradition regarding Asoka’s connection with it supplies

historical proof for its antiquity. According to Kalhana’s account

which may well have been based on genuine local tradition or even

inscriptional evidence, Asoka had replaced the old stuccoed enclosure

of the temple by one of stone. The great king was also credited with

having erected within this enclosure two temples called Asokesvara.

This old temple which is often mentioned by Kalhana and was the

scene of many a historical incident, has now completely disappeared.

According to the tradition of the local Purohitas it stood at a site close

the river-bank and nearly opposite to the bridge over the Vitasta.

When I first visited Vij^bror in 1889 I still found some ancient slabs

and fragments at this spot. It was then some 15 feet below the level of

the surrounding ground,3 and has since been partly built over. Stone

materials are said to have been removed from here for the new temple

1 Compare for detailed references, Bujat. i. 38, 105 notes.

2 The legend of the Tirtha is given at length in the x. Prakasa of the Haracari *

tacintdmani.

3 General Cunningham who saw these remains in 1847, rightly attributes them

to the temple of Vijayesa, but calls the place ‘ Vijayapara.’ He justly points to the

difference of level as an indication of the antiquity of the structure ; see Anc.

Geogr.} p. 98.

174

ANCIENT GEOGEAPHY OF KAS'MIB. [Extra No. 2,

of Vijayesvara which was built by Maharaja Ranbir Singh some thirty

years ago higher up on the river-bank.

It is probable that a temple so much frequented had undergone

more than one restoration in the course of the fifteen centuries which

lie between the time of Asoka and the end of Hindu reign in Kasmir.

Some time before a.d. 1081, while King Ananta was residing at the

Tirtha of Vijayesvara, the temple was burned down in a general con¬

flagration, caused by his son Kalasa. The latter, however, subsequently

restored the shrine. The old Linga of S/iva Vijayesvara seems to have

fallen a victim to the iconoclasm of Sikandar Butshikasfc.1

The town of Vijayesvara is ascribed by Kal liana to King Vijaya. 2 3

But nothing else is recorded of this ruler, aud this may cause a doubt

as to his historical existence. It is significant that the town is desig¬

nated either simply Vijayesvara or as Vijayaksetrar, which is abbre¬

viated from Vijayesvaraksetra. The modern name Vijvbror is the Ks.

equivalent of Vijayesvara, -bror (from Skr. bhattdraka ‘ god ’) having

replaced the more specific - is'vara , the usual designation of S'iva. 8

That the town had acquired importance at a comparatively early

date, is indicated by the mention of a thousand Agraharas said to have

been granted here by King Mihirakula to a settlement of Gaudhara

Brahmans.4 * It was large enough to accommodate the whole court and

army of King Ananta when the latter removed his residence to Vija¬

yesvara.6 * The narrative of the civil wars which fills the last Book of

Kalhana’s Chronicle shows the importance of the town by frequent

references to the military operations of which it was the object.6 One

of these passages proves that there was a bridge over the Vitasta here

already in the twelfth century, just as there is one still.

Vij^bror has remained a town of some importance and still boasts

of a considerable number of Brahmans, mostly Purohitas of the Tirtha.

The latter being conveniently situated on the pilgrims’ way to Martanda

1 See Jonar. (Bo. ed.), 762 ; for an earlier mention of this emblem see ib ., 127.

2 Rdjat. ii. 62.

3 Compare Rdjat . i. 38 note ; also ii. 134. In the same way I&brdr represents

Iseivara ; with the feminine-bmr for bhattdrikd we have SundVbrqr for Samdhyddevi ,

Budqbrar for Bhedddevl , etc.

The forms ‘ Bijbiara,’ ‘ Bijbihara,’ 1 Bijbehara,’ etc., under which the local name

figures in European books, are all based on a faulty Panjabi pronunciation. A

fanciful etymology of the name which sees in the first part of the word vidyd

‘ learning ’ and in the second ‘ Vihara ,’ has found favour in the guide-books and may

be mentioned here for curiosity’s sake.

* See Rdjat. i. 317.

6 See Rdjat. vii. 336 sqq.

6 Rdjat. viii, 746 sqq., 969 sqq., 1140, 1509 sqq., eto.

NORTHERN DISTRICTS OF MADAVARAJYA.

175

1899.]

and Amaranatha is much frequented even at the present day. The

Mahatmyas of Vijayesvara do not fail to name a considerable number

of minor Tirthas to be visited along with the main site now marked

by the new temple above referred to. But apart from Gakradhara and

Gambhirasamgcima I am unable to trace any of these in the older texts.

Turning to that portion of the Dachunpor district which lies in the

Lidar Valley we have but few old localities to notice. The village "of

Livar , some ten miles to the north-west of Vijayesvara, is the Levara

of the RajataraiiginI, mentioned as an Agrahara established by King

Lava.1 Kular , about four miles higher up the Valley, is identified by

an old gloss with Kuruhara, said to have been an Agrahara of Lava’s

son Kusa.2

Close to Pahalgam where the Lidar Valley divides into two

branches, lies the hamlet of Mamal. A small temple of the usual

Kasmir style built by the side of a fine spring is visited by the pilgrims

to Amaranatha. It is designated in the Mahatmya called Amaresvara-

kalpa as Mammesvara. It is in all probability identical with the shrine

of this name mentioned in the Rajatarangini.3

110. As we have already before noticed the several sacred sites

of the Amaranatha pilgrimage, we may now

turn back and descend to the left or eastern

portion of the Lidar Valley. It forms the

modern Pargana of Khovurpor. The latter name meaning ‘ left side ’

reproduces the earlier designation Vamaparsva, of the same significance,

found in Jonaraja’s Chronicle, the Lokaprakasa and elsewhere.1' In the

upper portion of the Pargana I am not able to identify any particular

old locality, though ancient remains in the form of sculptures of some

interest are found near several Nagas of this tract, e.g., at Lokutipur

and Sali (Papaharananaga) .

The large village of Hutamar is undoubtedly an old site. Its

modern name seems to identify it with the S'aktamatha which Ksemen-

dra names as one of the stations in the peregrinations of his heroine

Kankali. The chief mosque of the place is built with the remains

of a Hindu temple and preserves in its walls some sculptured fragments

of remarkable beauty.6

District of

Vamaparsva.

I See Raj at. i. 87.

8 Rdjat. i. 88.

* See viii. 3360.

* Jonar. (Bo. ed.) 79,1232.

& See Samay. ii. 43. The change of S'dkta > Hut * is in accordance with the

phonetic laws of Kasmiri ; mar is the regular derivative of matha, see above, § 56.

[When preparing my map, I had not noticed th« local name of Ksewendra’s text ;

it is hence not shown on the map].

176

ANCIENT GEOGRAPHY OF K AS' MIR. [Extra No- 2,

Shrine of

Bhimakesava.

About one mile below Hutamar and on the bank of a branch of

the Lidar, lies the liamlet of Bumfizu, which

contains an ancient structure of considerable

historical interest. The Ziarat of Baba Bam^din

Sahib is nothing but a well-preserved temple, converted, with a liberal

use of plaster, into the supposed resting place of a Muhammadan saint.

1 have shown elsewhere that there is good reason to identify this shrine

with the Bhimakesava temple which Bhlma S'ahi , king of Kabul,

the maternal grandfather of Queen Didda, is said to have erected during

the rule of her husband Ksemagupta (a.d. 950-958). 1

The legendary of the Ziarat relates that the saint was originally a

Hindu and bore before his conversion to Islam the name of Bhlma

Sadhi. It is easy to recognize in this name an adaptation of Bhlma

S'ahi. Also the name of the locality Bum^zu which the Martanda-

rnahatmya renders by Bhimadvipa , is clearly derived from the old name

of the shrine. Bhlma is an abbreviation of Bhimakesava to which Ks.

zn , ‘island,’ has been added with reference to the several islands

formed here by the Lidar immediately in front of the hamlet.

Kalhana tells us a curious anecdote regarding the fate of Bhlma

S'ahi’s temple in King Harsa’s time who confiscated the great treasures,

with which it was endowed.2 Close to the present Ziarat of Bam^din

Sahib is a small cave in the cliff containing a well-preserved little

temple which is still used for Hindu worship. Another smaller shrine

outside has been turned into the tomb of Rlshi Ruknu-d-dln Sahib.

111. About one mile south of Bum^zu we reach the Tirtha sacred

to Martanda which has from early times to the

present day enjoyed a prominent position

among the sacred sites of Kasmlr. It is marked by a magnificent

spring traditionally represented as two, Vimala and Kamala. An ancient

legend connects them with the birth of the sun-god Martanda.3 The

Tirtha is visited at frequent intervals by crowds of pilgrims and is

well-known also in India proper.

The popular name of the Tirtha, Bavan, is derived from Skr.

hhavana , ‘[sacred] habitation.’ This somewhat general appellation

seems to have come into use already at an early date, as S'rivara employs

it,4 and is in itself an indication of the great popularity of the Tirtha. A

Tirtha of Martanda.

1 See Rdjat. vi. 178 note. For an accurate description of the temple, see Bishop

Cowie’s paper, J. A. S. B., 1866, pp. 100 sq.

2 See Rdjat. vii. 1081 sqq.

3 Compare for a detailed account of the Tirtha, Rdjat. iv. 192 note. The Vimala

Naga is named by the NUamata , 963 ; S'rlv. i. 377, etc.

4 S'rlv. i. 376, 387.

1899.]

NORTHERN DISTRICTS OF MADAVARAJYA.

377

more specific designation is Maisabavan, Skr. * Matsyabliavana ’ ; this

owes its origin to the abundance of sacred fish which swarm in the

large basins filled by the spring.1

The ancient remains at the sacred spring itself are very scanty.

All the more imposing are the ruins of the great temple which King

Lalitaditya erected at a short distance in honour of the presiding deity

of the Tirtha.2

They are situated a little over a mile to the south-east of

* Bavan,’ near the northern edge of the Udar which stretches towards

Anatnag. It can scarcely be doubted that the site was chosen with

a view to the prominent position it assured to the great temple.

Kalhana duly praises “ the wonderful shrine of Martanda with its

massive walls of stone, within a lofty enclosure.” Its ruins though much

injured by the ravages of time and earthquakes, form still the most

impressive specimen of ancient Kasmlr architecture. They have been

much admired by European travellers and often described. They are

the earliest ruins in Kasmir the date of which is fixed with approximate

accuracy.3

The name Martanda, in the form of M&rtand or Matan, still

attaches to the ruins though they have long ago ceased to be an object

of religious interest. King Kalasa had sought this great fane at the

approach of death and expired at the feet of the sacred image (a.d.

1089). Harsa, his son, respected this temple in the course of the ruth¬

less confiscations to which he subjected the other rich shrines of the

country. Subsequently in Kalhana’s time the great quadrangular court¬

yard of the temple with its lofty walls and colonnades was used as

a fortification. The destruction of the sacred image is ascribed to

Sikandar Butshikast.

Kalhana distinctly mentions the town “ swelling with grapes ” which

Lalitaditya founded near his temple ; but of this no trace remains now.

It is probable that at that time a canal supplied water from the Lidar

to the naturally arid plateau on which the temple stands. This canal

seems to have been repaired by Zainu-l-‘abidin whose irrigation works

on the Martand Udar are described at length by Jonaraja.4 The

1 Comp. Ain-i-Ahb., ii. p. 358.

2 See Rdjat. iv. 192 and for details my note on the passage. For a description

of the temple compare, e.g., Cunningham, J.A.8.B., 1848, pp. 258 sqq. • Cole,

Ancient Buildings, pp. 19 sqq. Fergusson, Ind. Architecture, pp. 285 sqq.

3 Lalitaditya’s rule falls in the first half of the eighth century. Gen. Cunning¬

ham’s assumption that the temple was built by the earlier King Ranaditya, and only

the enclosure by Lalitaditya, rests on a misinterpretation of the Rajatar. passages

iv. 192 and iii. 462.

4 See Jonar. 1245 sqq .

J. i. 23

178

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

plateau lias since become once more an arid waste though the course of

the old canal can still be traced above Hutamar.

The town of Martanda had left its name to the small Pargana of

Matan which comprised this plateau as well as the villages situated

along the foot of the hills further east. It is referred to as Martandcidesa

by Jonaraja.1 * Abu-1-Fazl notices the large temple of Matan and the

well or pit close by, which a Muhammadan legend represents as the

place of captivity of the ‘angels Harut aud Marut.’ 8

Section VII. — Southern Districts of Madavarajya.

112. At the foot of the western extremity of the Martanda plateau

lies the town of Islamabad or by its Hindu

Anantanaga. name Anatnag . The latter is derived from

the great spring of the Anantanaga which issues at the southern end

of the town. The Naga, though no Tirtlia of particular repute, is

mentioned in the Nilamata, Haracaritacintamani and some Mahatmyas.3

Of<the town, however, I cannot find any old notice, and it is in all pro¬

bability, as its Muhammadan name implies, a later foundation. To the

north of the town and on the way to Bavan is the Gautamandga , named

by the Nilamata and the Martandamahatmya.

The modern name of the small district which comprised besides

Anatnag the tract immediately south and west of it, is Anyech. This

is represented in some Mahatmyas of recent composition by Anekdksa.

This name occurs also once in Srivara’s Chronicle, but the locality there

meant is not certain.4 *

The valley of the Arapath or Harsapathd which opens to the east

of Islamabad, forms the Pargana of Kut^har.

This name is in all probability connected with

that of the ancient Tirtha of Kapate^vara,

situated on the southern side of the valley close to the village of Kother

The name of the latter is undoubtedly a derivative of JZapatesvava, as

the analogy of Jyether < Jyesthesvara, Triphar < Tripuresvara, etc.,

clearly shows.

Tirtha of

Kapatesvara.

l Jonar. 1310.

8 See Aln-i-Akb., ii. p. 358. For the Muhammadan story, see also Vigne i. p. 361.

8 See Nilamata , 902 ; Vitastdf-Trisariidhydmdhdtmya, etc., also Haracar, x. 251 sqq.

(Anantabhavana).

4 S'rlv. iii. 184.

6 See for a detailed account, Rdjat, i. 32 note.

1899.]

SOUTHERN DISTRICTS OF MADAVARAJYA.

179

The place of pilgrimage is the sacred spring of Papasudctna sin-

removing ’), situated a short distance above Kother. In it Siva is

believed to have shown himself in the disguise ( kapata ) of pieces of

wood floating on the water. The legend is related at length in the

Nilamata, and the author of the Haracaritacintamanl devotes to it a

separate canto which has now become the official Mahatmya of the

Tlrtha.1 The importance of the latter is shown by the fact that

Kalhana mentions it in his Introduction first among the sacred sites of

K as mir.

Before him already Alberuni had heard of the story that pieces

of wood sent by Mahadeva appear annually “ in a pond called Kudai-

shahr to the left of the source of the Vitasta, in the middle of the month

of Vaisaklia.” 2 Kudaishalir ) , is an easily explained corrup¬

tion for i.e ., *Kavades'var , a prakritized form of the name. The

map shows that the description of the position of the Tlrtha is accurate

enough with reference to the Nilanaga as the Vitasta’s traditional

source. The date named by Alberuni is identical with that prescribed

for the Kapatesvara Yatra.

The sacred spring rises in a large circular tank, enclosed by an

ancient stone-wall with steps leading into the water. According to

Kalhana’s account this enclosure was constructed about a century before

his own time at the expense of the well-known King Bhoja of Malava.

The latter is said to have taken a vow to always wash his face in

the water of the Papasudana spring which he caused to be regularly

supplied to him in jars of glass.3 In my note on the passage I have

shown that local tradition at Kother still retains a recollection of this

story though in a rather legendary form. A small temple which stands

to the east of the tank, and some other remains probably belong to the

period of Bhoja. Abu-1-Fazl too knows, “ in the village of Kotihar,

a deep spring surrounded by stone temples. When its water decreases

an image of Mahadeva in sandal wood appears.”

About four miles to the north-east of Kother and on a branch of

_ . _ _ the Arapath river lies the populous village

amahgasa. of S'angcis, the ancient S'amangasa. 4 * The modern

name can be traced back to S'amarigasa through a course of regular

phonetic conversion, one stage of which is preserved in the form

Svangas supplied by the old glossator of the Chronicle.6 Some old

1 Haracar. xiv.

2 See India, ii. p. 181.

8 See Rdjat. vii. 190 sqq.

4 See Rdjat. i. 100; viii. 651.

6 Compare Rdjat. i. 100 note and the analogy of S'andra > S'dr.

180

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

carved slabs built into the chief Ziarat of the place attest its antiquity.

A short distance above S'angas we come to another old place. It is the

present village of VntHus which on the authority of the same glossator

and the name itself we can safely identify with Kalhana’s TJtrasa . 1

Uccala and Sussala in their flight from Harsa’s coart found a tem¬

porary refuge with the Damara who resided there.

Turning back to the west we find in the middle of the valley the

village of Khondur. An old gloss enables ns to identify it with the

ancient Skandapura mentioned by Kalhana as an Agrahara of King

Gopaditya.2 * More important is Ach°bal , a large village at the west foot

of the ridge which lines the Kutaliar Pargana from the south. It is

mentioned in the Chronicle under the name of Aksavala. The beauti¬

ful springs of the place have often been described since Abu-l-Fazl’s

time* also by Bernier.8 The park around them was a favourite camping

ground of the Mughal court. The Nllamata calls the spring Aksi-

j palanaga.

113. The Kutahar Pargana is adjoined on the south by the dis-

_ _ _ . trict of Bring which coincides with the valley

Pargana of Bring. , -r . ° , T, . , \

or tne Bring stream. Its old name cannot

be traced ; the Lokaprakasa transcribes the modern designation by

Bliynga.

At the western end of the Pargana and about 5 miles to the south¬

west of Achabal is the village of Ldkabcivcin which an old gloss identifies

with the Lokaponya of the Rajatarangini.4 * The numerous passages

which mention the place agree with this location. The name Lokabavan

applies also to the fine Naga adjoining the village, and this explains

the second part of the present name -bavan (Skr. bhavana ).6 * King

Lalitaditya is said to have built a town here. A small garden-palace

erected in Mughal times near the spring is partly constructed of old

materials.

Ascending the Bring valley we come again to an old site at the

large village of Bidar. It is certainly the Bhedara of Kalhana who

notices here a wealthy Agrahara of King Baladitya.6 A ruined mound

in the village and some old sculptures at the neighbouring Brahman

village of Hangalgund are the only ancient remains now above ground.

1 Compare vii. 1254.

8 See Rajat. i. 340.

5 Compare Rdjat. i. 338. In the translation of the Am-i-Akb. the name appears

as ‘ Acch Dal’, ii. p. 358 ; see Bernier, Travels , p. 413.

♦ See Rdjat. iv. 193 note.

6 See above, § 111,

6 Rdjat. iii. 481,

SOUTHERN DISTRICTS OF MABAYARAJYA.

1B1

1899.]

TIrtha of

Ardhanarlsvara.

From Bidar we may pay a passing visit to a small Tirtha which

though I cannot find it mentioned in any old

text, may yet claim some antiquity. About

miles to the south-east of Bidar lies the village

of Naru in the low hills flanking the valley. It contains a small temple

of ancient date which was restored forty years ago by a pious Dogra

official. It stands by the side of a small Naga at which, according to

the local Mahatmya I acquired from the resident Purohita, Siva is

worshipped as Ardhanauisvara, that is, in conjunction with his consort

Parvati. Inside the temple is an ancient image of Visnu with a short

Sanskrit inscription said to have been found in a miraculous way at

the restoration of the temple. About half a mile to the south-west is

a sacred spring known as Svedanaga which seems to have risen origi¬

nally within a large temple. The remains of the latter lie in shapeless

heaps around the spring. The latter is still visited by pilgrims.

It appears to me likely that it is this spot which Abu-1-Fazl wishes

to describe in the following notice. After mentioning the Kukar Nag

and Sundshrar (see below) among the sacred places of Bring, he says:

“ At a little distance in the midst of a beautiful temple, seven fountains

excite the wonderment of the beholder. In the summer-time self-

immolating ascetics here heap up a large fire around themselves and with

the utmost fortitude suffer themselves to be burned to death.” 1 He

then mentions a lofty hill containing an iron mine to the north of this

spot. This can only be the hill above Sop, on the northern side of

the Valley and nearly opposite Naru, from which iron is still extracted

at the present time. There is no other Naga within Bring to which

Abu-l-Fazl’s description would apply so closely as to the Svedanaga.

The Kukar Nag , mentioned by Abu-1-Fazl for its good water incit¬

ing a healthy appetite, lies about a mile above Bidar. It is a spring

of very great volume, referred to in the Trisamdhyamahatmya as

Kukkutesvara.

Bring contains one of the holiest of Kasmir Tirthas in the sacred

spring of the goddess Samdhya, also called

Trisamdhya , the modern Sundvbrar .2 It is

situated in a side valley opening to the south of the village of Deval-

gom, circ. 75° 22' long. 33° 32' lat. The spring of Samdhya derives

its fame as well as its appellation from the fact that during uncertain

periods in the early summer it flows, or is supposed to flow, intermittent¬

ly, three times in the day and three times in the night. Owing to the

analogy thus presented to the three-fold recitation of the Gayatri

l See Am-i-Akb., ii. p. 356.

3 See Rdjcit. i. 33 note.

Tirtha of Trisamdhya.

182

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Nilanaga.

(Samdhya), it is held sacred to the goddess Samdhya. At the season

indicated it is visited by a considerable concourse of people.

The small spring, which is usually dry for the greater part of the

year, has owing to the curious phenomenon above indicated always

enjoyed great fame as one of the ‘ wonders ’ of the valley. Kalhana

duly mentions it immediately after Kapatesvara. The Nilamata too

knows it. Abu-1-Fazl describes it in detail, and Dr. Bernier made it

a special point to visit this ‘ merveille de Cachemire.’ 1 He has observed

the phenomenon with his usual accuracy. The ingenious explanation

he has recorded of it, shows how closely he had examined the topogra¬

phical features of the little valley.

Close to the Trisamdhya spring there is another Naga, sacred to

the Seven Rsis, but not sharing the former’s peculiar nature. There

are no ancient remains in the neighbourhood deserving special notice.

114. To the south of Bring lies the valley of the Sandran River

which forms the Pargana of Shahabad. This

name is of comparatively modern origin, as

Abu-1-Fazl still knows the tract as Ver .2 This designation still survives

in the designation Vernag, i.e ., ‘ the Naga of Ver,’ popularly given to

the fine spring which we have already noticed as the habitation of the

Nilanaga and the traditional source of the Vitasta. Abu-1-Fazl saw

still to the east of it ‘ temples of stone.’ These have now disappeared,

their materials having been used probably for the construction of the

fine stone-enclosure which Jahangir built round the spring. The deep

blue colour of the water which collects in the spring-basin, may possibly

account for the location of the Nilanaga in this particular fountain.

Kalhana’s reference to the “ circular pond ” from which the Vitasta

rises, shows that the spring had also in ancient times an artificial

enclosure similar to the present one.3 4

Reference has already been made to the sacred spring of Vith^vutur

only about one mile to the north-west of

Vernag. The small village near by is men¬

tioned by Kalhana as a town under the name of Vitastatra 4 Asoka

is said to have erected there numerous Stupas. Within the Dharmciranyci

Vihara there stood a lofty Caitya built by him, but of these structures

no remains can now be traced above ground. Vitastatra could never

have been a large town as the ground is too confined. But some

importance is assured to the site by the Banahal route which leads past

Vitastatra.

1 Compare Ain-i-Akb., ii. pp. 355 sq. ; Bernier, Travels, pp. 410 sqq.

8 See Aln-i-Akb., ii. pp. 361, 370.

3 See Raj at. i. 28.

4 See Rdjat. i. 102 note.

SOUTHERN DISTRICTS OF MADAVARAJYA.

183

1809.]

District of

Devasarasa.

9

it. This pass and its ancient name Banasala we have already

spoken of.1

Of other old localities Pancahasta, the present Panzath, has already

been referred to as the site of one of the traditional sources of the

Vitasta. Kalhana mentions it in connection with a Matha which

• •

S'uravarman, Avantivarman’s minister, built here.2 * A pretty valley

which opens to the south of Panzath, is now known by the name of

its chief village Ruzul. The latter is mentioned by Jonaraja as Rajo-

laka.8 About three miles higher up this valley is the Naga of Vasuki.

It is mentioned in the Nilamata and other old texts, but does not

appear to have ever been an important Tirtha. 4

115. The Pargana of Divasar which adjoins Shahabad-Ver on the

west, may be roughly described as comprising

the tract of alluvial plain drained by the

Vesau (Visoka). By its ancient name of

Devasarasa it is often mentioned in the Rajataraiigini and other Chro¬

nicles.5 Being extensively irrigated by canals drawn from the Visoka

it is very fertile. This accounts for the great part which the Damaras

or feudal landholders of Devasarasa played during the weak reigns of

the latter kings. No certain reference to a specific locality within this

tract can be traced in our old texts. But it seems probable that

Parevisoka, repeatedly named in Kalhana’s Chronicle, must be looked

for within Devasarasa ; the name means literally * beyond the Visoka.’ 6

The fertile valleys descending to the right bank of the Visoka from

that portion of the Pir Pantgal Range which lies between the Konsar

Nag Peak and the Molii Pass, form a small district of their own, known

in recent times by the double name Khur-Naravav. The first part of

this name is taken from the large village of Khur situated about two

miles from the Visoka, circ. 74° 56" 45" long. 33° 37' lat. It is marked

as ‘ Koori ’ on the larger Survey map. The name Kher! which we find

used by Kalhana and Sfiivara for the designation of the tract, is in all

probability the older form of Khur.7 It seems that in later Hindu times

the administration of Kheri, perhaps as a royal allodial domain, formed

a special charge. Kalhana often refers to the Kherikarya as a high

state-office. The Sikhs and Dogras who established Jagirs for members

1 Compare above, § 41.

2 See Rdjat. v. 24.

8 See Jonar. (Bo. ed.), 90.

* See Nilamata, 901.

6 Compare Rdjat. viii. 504 note.

6 Compare Rdjat. iv. 5 note.

7 Compare regarding the identification of Kheri, Rdjat. i. 335 note.

184

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Godhara-Hastisala.

of the reigning family in Kliur-Naravav, may thus have followed an

earlier arrangement.

The only localities in this little district that are known to ns by

their old names, are Godhara and Hastisala,

the present Gudar and Ast'hel.1 2 These two

villages are situated close together, on a branch of the Visoka near the

eastern limits of Khur-Naravav. Kalhana mentions the ‘ Agrahara'of

Godhara-Hastisala’ as a foundation of King Godhara. The old gloss

which transcribes these local names by Godhar-Astlhil enabled me to

identify the places intended.

A small stream which falls into the Visoka at Gudar is known by

the name of Godavari and forms a Tirtha of some repute among the

Brahmans of the neighbouring districts. In the Mahatmya of the

Tirtha the site of the village is called Godara , and its name connected

with the legend of the appearance of the Godavari. The local tradition

heard by me on the spot tells of a town which King Gudar is supposed

to have founded here. It is possible that the name of King Godhara,

which we know Kalhana took indirectly from Helaraja’s ‘ List of Kings,’

rests on no better foundation than this long-surviving local tradition.8

There are no ancient remains traceable above ground at Gudar, and

the locality is far too confined for a larger settlement.

The Naubandhana Tirtha and the Kramasaras or Konsar Nag

south of this district have already been previously noticed.

116. To the north of Divasar lies the considerable district of

Ad^vin extending from the western end of

Khur-Naravav to the lower course of the

Visoka. Its present name is derived from that

of the large village of Ad^vin, which lies on the left bank of the Visoka,

about three miles south-west of Vij^bror. In the form of Ardliavana

this name is found already in a passage of Jonaraja’s Chronicle, sup¬

plied by the new edition.3 The ancient designation of the district,

however, was Karala. This is used by Kalhana when speaking of the

Suvarnamanikulyd , the present Sun^man1 Kul, which has already been

referred to as irrigating part of Ad^vin.4

In the lower portion of the district and on the left bank of the

Visoka, we have the ancient Katimusa , the present village of Kaimuh .

District of Karala

(Ardliavana).

1 For details compare Rajat. i. 96 note.

2 Regarding the unhistorical character of the royal names which Kalhana

inserted on Helaraja’s authority, see Rajat. i. 86 note. They seem to be all of an

eponymic character.

3 See Jonar. (Bo. ed.) 1330.

4 See Rajat. i. 97 note, and above, § 78.

1899.]

SOUTHERN DISTRICTS OP MADAVARAJYA.

185

The place is mentioned by Kalhana as an Agrahara founded by

Tunjina I., and contains some old remains built into its chief Ziarat.1

Part of Ad^vin lies on an alluvial plateau. The northernmost

portion of this Udav seems to have been formed into a separate Pargana

after Zainu-l-‘abidin had constructed there extensive irrigation chan¬

nels. From the small town of Jainapuri founded by him the new sub¬

division took the name of ZainSpor or Jainapura.2 3 At the east foot

of the Zain^por Udar lies the village of Vach i (map ‘Woochi’) which

on the authority of an old gloss may be identified with Va^cika (or

Vascika), an Agrahara founded by Gopaditya.8

The Pargana which joins on to Ad^vin in the north-east, is now

known as JBot (map ‘ Batoo ’). Its ancient name is unknown. The

only old locality I can trace in it is the village of Sidciu, 74° 51' long.

33° 4P lat., the ancient Siddhapatha.4 * It has given its name to the

route previously mentioned which leads to the Budil and Konsar Nag

Passes.

It is curious that we find no old mention whatever of S'upiyan , a

considerable town, which is now the trade emporium for the Pir Pantsal

route. In this character S'upiyan has replaced the ancient Sfiirapura

or Hur^por, but the change must be a comparatively recent one.

S'urapura which we have already noticed as the Kasmir terminus

of the Pir Pantsal route, lies some seven miles

Surapura. “ _ _ t _ .

higher up on the Kembyar?'.6 It received its

name from the minister STiravarman who built it in the time of Avanti-

varman and transferred to it the watch-station or ‘ Dranga ’ of the

route. The position of the latter is marked by a spot known as Ilahi

Darwaza, a short distance above Hur^por.6 S'urapura must have been

a place of considerable extent as ruins of old habitations can be traced

on the river banks for over two miles below the present Hiir^por. It

evidently retained its importance down to Akbar’s time. For it is

regularly mentioned by all the later Chronicles in connection with

marches and traffic by the Pir Pantsal route. The ancient remains of

the place have been described by me in my notes on the lafter.7

Our previous account of the old localities on the way to the Pir

1 Compare Rdjat. ii. 55 note.

2 See Jonar. (Bo. ed.) 1144' sq. ; S'rlv. iii. 194; Fourth Chron. 360, 383.

3 Compare Rdjat. i. 343 note.

4 See Rdjat. viii. 557.

3 Regarding S'urapura and its old sites, compare Rajat. Note D (iii. 227) ; v. 39

note ; also J. A. S. B., 1895, pp. 381 sqq.

6 See above, § 42.

T See J. A. S. B., 1895, p. 385.

J. i. 24

18(5

ANCTENT GEOGRAPHY OF KA^MTR.

[Extra No. 2,

Pantsal Pass makes it unnecessary for us to proceed now further in

this direction. Descending, then, by the Rembyar?' we come on its left

bank to the village of Degam situated about one and a half miles to the

west of S'upiyan. It is the Degrama of the RajatarahginI and the site

of the Kapalamocana Tirtha.1 A t the spring of the latter S'iva is

supposed to have cleaned himself from the sin attaching to him after

the cutting-off of Brahman’s head ( kapala ). The Tirtha is old, because

the Haracaritacintamani mentions it twice.2 There are but few ancient

remains at the sacred site, and the extant Mahatmya is evidently not

of old date. It calls the village by the name of Dvigrdma and knows

the modern S'upiyan by the name of Surpdyana.

117. The villages which lie at the foot of the pine-clad spurs

descending into the valley west and north-west of S'upiyan, formed until

recent times a small distinct Pargana known as Suparsamun. Abu-1-

Fazl mentions it (Soparsaman), but I am not able to trace it in our

older texts.

To the north of this tract and of Bot extends the Pargana of Suhru.

Its old name is unknown. Here at the foot

of the hills, we have the ancient Kalyanapura,

represented by the present village of Kalampor, situated 74° 54' long.

33° 48' lat. It was founded by Kalyanadevi, a queen of Jayapida.3 4 *

Being on the high road from the Pir Panteal Pass to STlnagar, it was

repeatedly7 the scene of battles fought with invaders from that direction.41

At Kalyanapura there was in Kalhana’s time the splendid country-

seat of a powerful Damara.6 The large village of Drdbagdm , some three

miles north of Kalampor, is mentioned as Drabhagrama by S'rivara,

along with Kalyanapura, in the description of a battle which was fought

between the two places.6

High up in the valley of the Birnai stream which debouches at

Drabagam from the south-west, is the site of

an ancient Tirtha which though now completely

forgotten must have ranked once amongst the most popular in Kasmlr.

In Kalhana’s introduction there is named, along with Trisamdhya,

Svayambhu, S'arada and other famous sites, “ the hill of Bheda ( Bheda -

giri) sanctified by the Gangodbheda spring.” There the goddess Sarasvati

Kalyanapura.

Tirtha of Bheda.

1 Compare Rdjat. vii. 266.

2 See Haracar. x. 249 ; xiv. 111.

3 See Rdjat. iv. 483 note.

4 See Rdjat. viii. 1261 sqq. ; S'riv. iv. 466 sqq.

6 See Rdjat. viii. 2348 sqq.

® ®ee iv. 467. For a miniature temple extant at Drabagam, compare

Bishop Cowie’s note, J. A. S. B., 1866, p. 117.

1899.]

SOUTHERN DISTRICTS OP MADAVARAJYA.

187

was believed to have shown herself as a swan in a lake situated on the

summit of the hill. This Tirtha has long ago ceased to be visited by

pilgrims, and all recollection regarding its position has been lost to Pandit

tradition. Fortunately the old Mahatmya of the sacred lake has surviv¬

ed in a single copy. With the help of some indications furnished by

it and an opportune notice of Abu-1-Fazl, I was able to make a search

for this ancient Tirtha which ultimately led to its discovery at the

present BucHbrar in the valley above indicated.

For the detailed evidence regarding this identification I must refer

to my note on Kalhana’s passage.1 Here a brief reference to the topo¬

graphical peculiarity of the site will suffice. The Mahatmya describes

the lake sacred to the goddess SarasvatI-B/ie</a as situated on the

summit of a hill, and Gangodbheda as a spring flowing from it. At

Biidvbrdr, a small Gujar hamlet, which occupies the position marked by

Bhedagiri on the map, T found an ancient stone-lined tank fed by a

spring on the top of a small hillock. The latter rises about seventy feet

above the level of the narrow valley in which it is situated. From

the side of the hillock issues a spring which is the natural outflow of

the tank and exactly corresponds to the description given of Gangod-

bheda. The name Bud^brdr is the direct derivative of Bhedadevl, ‘ the

goddess Bheda,’ the popular designation of the Tirtha found in the

Mahatmya; -brdr < Skr. bhcittarika is the equivalent of deni as in

Sund^brar, Har^brar and other names.

The water of the spring which fills the tank, is said to keep warm

in the winter. This accounts evidently for the story told in the

Mahatmya that snow never lies on the ground around the sacred tank.

Also Abu-l-Fazl’s notice of the Tirtha mentions this particular feature :

“ Near Shukroh (Sfirkru) is a low hill on the summit of which is a

fountain which flows throughout the year and is a place of pilgrimage

for the devout. The snow does not fall on this spur.”2

Also Sh'ivara helped to guide my search in the direction of

Bud^brar and to confirm the subsequent identification. He mentions

the route through Bhedavana , ‘ the forest of Bheda,’ as the line of retreat

taken by the troops who after their defeat in the above-mentioned

engagement near Drabagam were fleeing towards Bajauri.8 A glance

at the map shows that the thickly wooded valley of BucRbrar is meant

here. For a force beaten near Drabagam it affords the most direct and

*

safest retreat to the Pir Pantsal Pass and hence to Rajaurl. The route

leading through the valley joins the 4 Imperial Road ’ at Dubji and is

shown on the map.

1 See Rdjat. i. 35, Note A.

* See Ain-i-Akb.y ii. p. 362.

8 Compare S'riv. iv. 496 and the preceding narrative.

188

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

Returning' once move to the plain we have yet to notice two other

old localities of S'ukru. Bilau (map ‘ Belloli *) about four miles north¬

east, of Drabagam is probably the ‘ village of Bilava ’ once mentioned

by Kalhana.1 Within a mile of it lies the village Suntsamil which we

may safely identify with the Suvarnasanura of the RajataranginI, in

view of the resemblance of the names and the repeated mention of the

latter place together with Kalyanapura.2

118. East of S'ukru towards the Vitasta stretches the Pargana of

S'avur (map ‘Showra’). The earlier form of

Districts of Savur name cannot be traced. Its northern part

and Chrath. . # - . ,

is formed by the alluvial plateau known as the

Naunagar TJdar. This latter is twice referred to as Naunagara in

Kalhana’s Chronicle.3 The village of Payer which lies at the foot of

the Udar at its north-western end contains a well-preserved little temple

often described by European travellers.4 * Nothing is known regarding

the original name of the locality.

To the north of S'ukru we have the district of Chrath (shown by name

on the larger survey map). It extends from the hills above Ramuh in

a north-easterly direction to the left bank of the Vitasta. Its old name

is restored in Pandit Sahibram’s Tlrthasamgraha as * Srirastra , but

I do not know on what authority. Ramuh, first correctly identified by

Prof. Buhler with Kalhana’s Ramusa,6 is a considerable village on the

high road from S'upiyan to S'rinagar. It is first mentioned as an

Agrahara, founded by a queen of Tunjina I. A small spring at the

northern end of the village, called Dhanandga, is visited as a Tlrtha

and contains some fragments of ancient sculptures. The temple

erected by the Brahman family of the Dars which now holds Ramuh

as a Jagir, does not seem to mark an old site.

A short distance to the north of Ramuh rises an alluvial plateau

which is crossed by the road to S'rinagar. It is known as Gus ■

TJdar , from the village of Gus situated at its eastern foot, about two

miles from Ramuh. The place is mentioned as Gusika in STivara’s

1 See Rdjat. vii. 1016.

2 See Rdjat. vii. 1519 note; sun % ‘gold’ is the regular Ks, derivative of Skr.

suvarna.

3 See Rdjat. vii. 358.

* Compare, e.g., Cunningham, J. A. S. B., 1848, pp. 254 sqq. I am unable to

explain why the place figures in all European accounts as Payech, Pa Yech, etc.

Vigne, ii. 41, first uses this form which is locally quite unknown, and does not fail

to explain it by one of his naive etymologies.

6 Rdjat. ii. 55 ; Report, p. 7. Medial s becomes in Ks. regularly h ; comp.

Katlmusa > Kaimuh , Khonamusa > KhunOmoh , etc.

1899.]

SOUTHERN DISTRICTS OF MADAVARAJYA.

189

Chronicle which also knows the plateau by the name Gusikodtjara.1

At the other end of Chrath towards the Vitastfi lies the large village

of Batanpor, 75° 1' long. 33° 55' lat., which in all probability represents

the Ratnapura of the Rajatarahgini.2 3 The latter was founded in

Kalhana’ s time by Queen Ratnadevi who also constructed there a fine

Matha.

With Chrath may be mentioned two localities on the left bank of

the Vitasta though in recent times they were counted with the riveraine

Pargana of Sairu-l-MawazF Bala. Guripfir) a small village opposite

to the foot of Mount Vastarvan, is identified by an old gloss with

Gopalapura which, according to Kalhana, was founded by Queen Sugan-

dha (a.d. 904-6). s

Lower down on the river is the large village Kak^por which forms

as it were the riverside station or port for g'upiyan. A note from the

hand of Pandit Rajanaka Ratnakantha who wrote about the middle

of the 17th century the Codex Archetypus of the Rajataraiigini, identi¬

fies Utpalapura with Kak?p5r.4 * Utpalapura was founded by Utpala,

an uncle of King Cippata-Jayapida, in the early part of the 9th century.

If this identification is correct, one of the ruined temples extant at

Kak^por and noticed already by Gen. Cunningham, may be the shrine

of Visnu TJtpalasvamin mentioned by Kalhana in connection with the

foundation of Utpalapura. Jonaraja also knows the latter place and

records a late restoration of its Visnu temple.6

119. North of Chrath we come to the district of Nagam which is

one of considerable extent. Its old name

Distncte^of Nagam Nagrama is often mentioned in the later Chro¬

nicles.6 The only old locality which I can

trace in it, is the village of Arlgdm, situated 74° 45' long. 33° 56' lat.

It is the Hadigrama of Kalhana, mentioned as an Agrahara of Gopa-

ditya and as the scene of several fights in the Chronicler’s own time.7

1 S'riv. iv. 532, 465, 592 sqq. ; - udddra is the Skr. original of the Ks. term udar>

see Rdjat. note viii. 1427.

2 See Rdjat. viii. 2434.

3 See Rdjat. v. 244 note.

4 See Rdjat. iv. 695 note. The learned copyist’s note is in a copy of the

Ksetrapdlapaddhati seen by me in 1895 in the possession of a Kasmlr Brahman resident

at Lahore.

3 See Jonar. (Bo. ed.) Ill sqq., 369, 1142.

3 Compare Jonar. (Bo. ed.) 661; S'riv. ii. 10; iii. 24, 430; iv. 349; Fourth

Chron. 258, etc.

1 See Rdjat. i. 340 note. The old glossator on this passage renders Hadigrama

correctly by Adegrdm.

190 ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Some remains of old buildings are reported to exist at this place ; I have

not seen it myself.

About five miles due south of AUgom we find a small lake known

as Nilanag, situated in a valley between low spurs descending from the

Pir Pantsal Range. It appears to have been formed by an old landslip

which blocked a narrow defile in the Valley. This lake does not appear

ever to have enjoyed any particular sanctity. But Abu-1-Fazl by some

curious misapprehension transfers to it the legends of the famous

Nilanaga (at Vernag). He adds to them what appears like a garbled

version of the story of the city submerged in the Mahapadma or Volur

lake.1

Nagam is adjoined on the north by the Pargana of Yech which

extends to the immediate vicinity of STlnagar. Its old name is given

as Iksika by STivara.2 In the centre of the tract lies an arid alluvial

plateau known as Ddmadar Udar, where an ancient popular tradition

surviving to the present day has localized the legend of King Damodara.

The story as related by Kalhana, represents the king as having built

a town on the Udar which latter was called

Damodara’s Udar. _ o t , , , .

after him Damodarasuda.6 In order to bring

water to it he had a great dam, called Guddasetu, constructed by super¬

natural agency. Once hungry Brahmans asked the king for food, just

as he was going to bathe. The king refused to comply with their

request until he had taken his bath. The Brahmans thereupon cursed

him so that he became a snake. Ever since the unfortunate king is

seen by people in the form of a snake “rushing about in search of water

far and wide on the Damodara-Suda.” He is not to be delivered from

the curse until he hears the whole Ramayana recited to him in a single

day, a task which renders his release hopeless.

The modern name Damadar Udar is the exact equivalent of

Kalliana’s Damodara-Suda , the old Skr. term suda meaning a ‘ place

where the soil is barren.’ The local name Guddasetu still lives in that

of the small village GudVsuth situated at the south foot of the Udar.

Just at this point the latter shows its greatest relative elevation and

falls off towards the valley with a steep bank over one hundred feet

high. The wall-like appearance of this bank probably suggested the

story of an embankment which was to bring water to the plateau. In

view of the configuration of the ground no serious attempt at irrigation

by means of an aqueduct could ever have been made in this locality.

1 Compare Ain-i-Akb., ii. p. 363. It is possible that of the two Nilanagas which

the Nilamata, 903, mentions besides the famous spring of that name, one was located

in the Nagam lake.

2 S'riv. iii. 25.

s Compare for detailed references, Bajat. i. 156 note.

1899.]

SOUTHERN DISTRICTS OF KRAMARAJYA.

191

The Udar stretches in a north-westerly direction, for about six miles

from the village of VahHor, with a breadth varying from two to three

miles. It bears only scanty crops of Indian corn in patches. Being

entirely devoid of water, it is a dry and barren waste, a haunt of jackals

as in the days when King Ksemagupta bunted over the ‘ Damodararanya, 1 2

The main features of the legend regarding it are well known to popular

tradition throughout Kasmir. The inhabitants of the neighbouring

villages also point to a spot on the Udar known as Satards Teng, as the

site of Damodara’s palace. A spring called Damodar-Nag in the village

of Lalgam, is believed to have served for the king’s ablutions.

To Yech belongs also the small village of SomarVbug on the left

bank of the Vitasta which according to the note of the old glossator

A3 marks the site of the temple of Visnu Samarasvamin mentioned by

Kal liana.3 Another old locality in Yech is probably marked by the

hamlet of Halathal to which Abu-1-Fazl refers. It is not shown on the

Survey map, and I have not been able to ascertain its exact position.

HalHhal is evidently a derivative of S'alasthala, the name given by

Kalhana to a locality where a fight took place in the time of King

Ananta.3 Abu-1-Fazl mentions ‘ Halthal ’ for its quivering tree. “ If

the smallest branch of it be shaken, the whole tree becomes tremulous.”

Section VIII. — Southern Districts op Kramarajya.

Districts of Dunts,

Biru, Mafich%lidm.

120. — To the west of Yech and reaching close up to the capital,

lies the Pargana now called Dunts (map

‘Doonsoo’). Its ancient name is uncertain;

possibly it is intended by the name Dvdvimsati

in the Lokaprakasa’s list of ‘ Visayas.’ In Abu-l-FazTs table of Parganas

Dunts (‘Dunsu’) is already counted with Kamraz. An old locality in

it is S'0por, a large village situated circ. 74° 45' long. 34° 1' lat. (map

‘ Shalipoor ’). We may safely recognize in it the Selyapura of the

Rajatarangini which is referred to as a place on the direct route from

the Tos^maidan Pass and the Karkotadranga to S'rlnagar.4 S

DLukh'ilitH (map ‘ Haklitri ’) can safely be identified, in view of the

name and the evidence of an old gloss, with S'uskaletra mentioned in the

1 Compare Rdjat. vi. 183.

2 See note v. 25. — The ending -bug is not rare in Kasmir village names. Accor¬

ding to Pandit tradition, it is derived from Skr. bhoga in the sense of ‘ property

granted for the usufruct [of a temple].’

S See note vii. 159; Ain-i-Akb ., ii. p. 363.

4> See Rajat. vii. 4 94 note j viii. 200.

192

ANCIENT GEOGRAPHY OF KASHMIR [Extra No. 2,

Rajatarangini as a place where Stupas were erected by King Asoka.1

I have not visited the village myself and am lienee unable to say

whether there are any remains in the vicinity which could be attributed

to Stupas. Kalhana locates at S'uskaletra the fierce battle in which

King Jayapida recovered his kingdom.

West of Dunts and towards the mountains of the Pir Pantsal lies

the Pargana of Biru. Its old designation Bahurupa is derived from the

spring of that name which is situated at the present village of Biru,

74° 39' long. 34° 1' lat., and is already referred to as a Tirtlia in the

Nilamata.2 Abu-1-Fazl knows the village and spring by an intermediate

form of the name, Biruwa , and mentions the miraculous power of the

spring to heal leprosy.3 Close to the village of Biru is Sunapah in

which we may, with an old glossator of the Rajatarangini, recognize

Suvarnaparsva, an Agraliara of Lalitaditya.4 *'

About four miles to the south-west of Biru we reach Khag , a

considerable place. It is undoubtedly the Khagi or Khagika men¬

tioned by Kalhana as an Agraliara both of King Khagendra and of

Gopaditya.6

Some miles north of Kliag an isolated spur known as Poskar

projects into the level plain from the slopes of the Pir Pantsal Range.

At its eastern foot is the Puskaranaga , referred to as a Tirtha in the

Nilamata and several older Mahatmyas, and still the object of a regular

pilgrimage.6 Of the route which leads down into Biru from the Tos?1-

maidan Pass, and of Karkotadranga, the watch station on it, we have

already spoken above.

Biru and Dunts are adjoined on the north by the Pargana of

MancMhdm which extends eastwards as far as the Yitasta. It is

probably intended by the name of Maksasrama found in a single passage

of S'livara and in the Lokaprnkasa.7 The village of Ratvsun, situated

74° 38' long. 34° 4' lat., is probably, as indicated by an old gloss, the

1 Compare Rdjat. notes i. 102; iv. 473 ; Ks. Eukh^lit^r is the direct phonetic

derivative of the Skr. form.

2 See Nilamata, 948, 1180, 1341 sq. The name Bahurupa is given to the tract

by Jonar. (Bo. ed.) 286, 840; S'riv. ii. 19, iii. 159 ; iv. 620, and ought to have been

shown on the map.

3 Ain-i-Alcb., ii. p. 363.

4 See Rdjat. iv. 673.

3 Compare Rdjat. i. 90, 340.

8 See Nilamata , 1021, 1347. There were several other Puskara TTrthas in

Kasmir. One was connected with the Suresvari pilgrimage and probably situated

in Phakk. ; see S'arvdv. v. 56 sqq.

7 See S'riv. iv. 351.

1899.]

SOUTHERN DISTRICTS OF KRAMARAJYA.

193

Aristotsadana of the Rajatarangini.1 From this form the modern

name of the village can be derived without difficulty. A temple is said

to have been erected there by a queen of JBaladitya.

On the Vifcasta some six miles below Srinagar is the small village of

Malur which on the authority of Rajanaka Ratnakantha maybe identified

with Malhanapura, a foundation of King Jayapida.2 ZaintJcoth, situated

near marshy ground about two miles south-east of it, preserves the

name of Zainu-l-‘abidin, its founder, and is mentioned as Jainakotta

by Jonaraja.3

121. The Pargana of Par^spor (map ‘ Paraspoor ’) which lies next

to Manch^hom, is one of small extent, but

an asapura. contains a site of great historical interest. It

has received its name from the ancient Parihasapura, which King

Lalitaditya had built as his capital.4 The identity of the name Par^spar

and Parihasapura is evident on phonetic grounds and was well-known

to the authors of the Persian abstracts of the Rajataranginl. Yet

curiously enough the site of Parihasapura had remained unidentified

until I visited the spot in 1892 and traced the ruins of Lalitaditya’s

great structures as described by Kalhana, on the plateau known as the

‘ Par^spor Ildar.’

This plateau rises south-east of Shadipur, between the marshes of

Panzinor on the east and those of Haratrath on the west. Its length

is about two miles from north to south, and its greatest breadth not

much over a mile. On the north this plateau is separated from the

higher ground of Trigam by the Badrihel Nala which, as I have shown

above, represents the old bed of the Vitasta previous to Suyya’s regula¬

tion.6 On the other sides it is surrounded by marshes which for a great

part of the year are still accessible by boats. Its general elevation is

about one hundred feet.

A broad ravine which cuts into the plateau from the south, and in

which the village of Divar (map 4 Diara ’) nestles, divides it into two

parts. On the south-western portion are the ruins of two large temples,

much decayed, but still showing dimensions which considerably exceed

those of the great temple of Martanda. On that part of the Udar which

lies to the north-east and towards the Badrihel Nala, there is a whole

1 Raj at. iii, 482.

2 Compare Rdjat. iv. 484.

3 Jonar. (Bo. ed.), 1248.

4 For a detailed account of the site of Parihasapura and its identification, com¬

pare Note F, Rajat. iv. 194-204. The large scale map added to Note / shows the

position of the several ruins in detail.

6 See § 70.

J. i. 25

194

ANCIENT GEOGRAPHY OF KAS'MlR.

[Extra No. 2,

series of ruined structures. Among these three great buildings attract

attention. As an indication of their size it may be mentioned that the

ruined mound which marks the central shrine of the northernmost

building has a diameter of nearly 300 feet. Though it consists now

only of a confused heap of massive blocks it still rises to a height of

over 30 feet from the ground. The enclosing quadrangle which can

also be traced, measures about 410 feet square. At some distance

from this group of ruins there is another smaller one at the south¬

eastern extremity of the plateau now known as Gurdan.

I must refer for a more detailed account of these ruins and their

relative position to the Note on Parihasapura, F , appended to my

translation of the Chronicle. Here it will suffice to point out that the

four great temples of Visnu Parihasakesava, Muktakesava, Mahavaraha,

Govardhanadliara as well as the Bdjavihara with its colossal image of

Buddha, which Kalhana mentions as Lalitaditya’s chief structures at

Parihasapura, must all be looked for among these ruins. Their ex¬

tremely decayed condition makes an attempt at detailed identification

difficult.

Still less we can hope to trace now the position of the numerous

shrines, Lingas, Viharas, etc., which are mentioned by Kalhana as

having been erected at the king’s favourite residence by his queens and

court.1 One of the great ruins of the northern group shows features

characteristic of a Vihara and may be the Rajavihara. Some clue is

also furnished by the name Gurdan attaching to the isolated ruins

above mentioned. Gurdan is the common Kasmiri form of the name

Govardhana, and hence points to these ruins being the remains of the

temple called Govardhanadhara.

The state of utter destruction in which the ruins of Parihasapura,

are now found, is easily accounted for by the

history of the site. Parihasapura ceased to be

the royal residence already under the son of

its founder. The Chronicle distinctly records of King Vajraditya that

he withdrew the various foundations which his father Lalitaditya had

made there.2 When a century later King Avantivarman effected his

great regulation of the Vitasta, the bed of the river and its junction with

the Sindhu was diverted to Shad1pur, nearly three miles away from Pari¬

hasapura.3 This change must have still more seriously diminished the

importance of the latter. The ruinous condition into which Parihasa¬

pura must have fallen only one and a half centuries after its foundation,

History of Parihasa¬

pura.

1 See Rdjat. iv. 207-216.

2 Rdjat. iv. 395.

8 See above, §§ 70, 71.

1899.]

SOUTHERN DISTRICTS OF KRAMARAJYA.

195

is shown by the fact that STimkaravarman (a.d. 883-902) carried away

from it materials for the constr action of his new town and temples at

Pattana (Patan).1 2

Some of the shrines of Parihasapura, however, survived to a later

period. Thus we find the colossal copper statue of Buddha at the

Rajavihara mentioned as one of the few sacred images which escaped

being melted down in the reign of King Harsa (a.d. 1089-1101). Also

a great religious festival established at Parihasapura by Lalitaditya

seems to have been held still in Kalliana’s time.3 In the rising which

led to the downfall of Harsa, Parihasapura was occupied by the pre¬

tender Uccala.3 The steep slopes of the plateau and the marshes around

made it a position of military value. When Uccala had suffered a

defeat some of the routed rebels threw themselves into the Rajavihara,

which was subsequently burned down. After this, Harsa carried away

and broke np the famous silver statue of Visnu which had been placed

by Lalitaditya in the temple of Parihasakesava.

The final destruction of the temples is attributed by Abu-1-Fazl

and the Muhammadan chroniclers to Sikandar Butshikast. The

former records the tradition that after the destruction of the lofty

temple of ‘ Paraspur ’ a copper tablet with a Sanskrit inscription was

discovered which predicted its destruction ‘after the lapse of eleven

hundred years ’ by one Sikandar.4 This prophecy post factum shows

that its author, whoever he may have been, was rather weak in histori¬

cal chronology. Parihasapura had been founded only about six and

a half centuries before Sikandar Butshikast’s time.5 At the beginning

of the eighteenth century the ruins seem still to have been in a somewhat

better condition than now. Both Muhammad ‘Azim and Narayan Kaul

mention them and speak particularly of fragments of a large monolithic

column. Tradition seems to have connected these fragments with the

pillar of Garuda which Kalhana mentions as having been set up by Lalita¬

ditya.6 * 8 The huge square block of stone still visible on the top of the

northernmost mound is perhaps one of them.

1 See Bajat, v. 161.

2 See Bajat. iv. 242 sq. For the temple of Ramasvamin which was seen empty-

in Kalliana’s time, compare iv. 275, 334 sq.

3 Bajat. vii. 1326 sqq.

4 See Ain-i-ATcb., i. p. 364.

6 Exactly the same tradition is now current among the Purohitas of Vij^bror

about the destruction of the Vijayesvara image. This alleged inscription is said to

have run : EkadaSasatam varsam Sikendaramahabala I bismdlla iti mantrena nci6yante

VijayeSvardh 1 1 The curious Sanskrit of this doggrel is an indication that its author

may probably have belonged himself to the noble guild of the Bachbattas.

8 Compare Wilson, Essay, p. 50; also footnote 16 to Note F, on Parihasapura.

196

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

TrigramI ;

Phalapura.

The ruins of Pariliasapura have served ever since S'amkaravarman’s

time as quarries for stone-material. Their position near to navig¬

able water-channels made them particularly convenient for this

purpose. Since 1892 when I first saw the ruins, till 1896 many large

stone-blocks have found their way as road metal into the new Tonga

Hoad which passes the plateau on the south. On my report steps were

subsequently taken by the Darbar to stop this vandalism and prevent

its recurrence.

122. We have already above when describing the old bed of the

Vitasta near Pariliasapura, had occasion to

refer to the village of Trigam , the ancient Tri-

grami. It lies about one and a half miles to

the north-east of the Par^spor ruins. The place is mentioned already in

Lalitaditya’s time in connection with an affray which took place at

Pariliasapura.1 The J3onasar (* Bliavanasaras P) lake to the west of

Trigam is visited as a subsidiary Tirtha on the Kapalamocana pilgrimage.

The ruined temple south of Trigam which I believe may be identified

with the Vainyasvamin temple, has already been mentioned in our

remarks on the site of the old confluence.

A ruined site which lies opposite to Vainyasvamin on the western

side of the Trigam swamp, may for reasons set forth elsewhere be taken

for the old Visnusvamin temple.2 This is named by Kalhana as having

been situated opposite to the Vainyasvamin shrine on the other side of

the old confluence. The passage of the Chronicle describes the temple

of Visnusvamin as belonging already to Phalapura, while Vainyasvamin

was counted with Pariliasapura.

From this and some other indications I conclude that Phalapura

was the designation of a small territorial subdivision which probably

extended along the present left bank of the Vitasta near Shadipur.3

The site at which I locate the Visnusvamin temple, was included in

recent times in the riveraine Pargana of Sairu-l-mawazi‘ Payin (map

‘ Salimozapaieen ’). This, we know from Abu-l-Fazl, was created

already before Akbar’s time and probably absorbed Phalapura as well

as other minor tracts.4 * Phalapura had received its designation from

a locality of that name which Lalitaditya had founded apparently before

Pariliasapura,6 just as the latter gave its name to the Par^spor Pargana.

1 See Rajat. iv. 323 sqq.

8 Compare Rajat. Note I, v. 977IOO, § 12.

s See Rajat. Note I, § 13.

4 See Ain-i-Ahb., ii. p. 367.

6 Compare Rajat. iv. 184, 673.

1899.]

SOUTHERN DISTRICTS OF KRAMARAJYA.

197

Descending by the left bank of the Vitasta for about five miles

Jayapura below Sliad^ur, we approach the site of king

Jayapida’s capital, the ancient Jayapura.1 It

is marked by the present village of And^rkoth. This consists of two

distinct parts. One lies on an island in the marshes opposite Sambal and

the other facing the former on the strip of land which separates these

marshes from the Vitasta. On the island there are conspicuous remains

of ancient temples which have been first examined and described by

Prof. Bidder.2 They are attributed by the local tradition to King

j Jayapid.’ The identity of And^rkoth with King Jayapida’s town is

also well-known to the Srinagar Pandits. As Shi vara still uses the term

Jayapura or Jayapldapura for the designation of the present AmDrkoth,

we can easily understand the survival of the tradition.

Kalhana’s description of the town indicates clearly the situation

of the latter and also accounts for its modern name. Jayaplda accord¬

ing to this notice had the castle ( kotta ) of Jayapura built in the middle

of a lake, after having the ground required for it filled up, as the legend

asserts, by the help of Raksasas. There he constructed a large Vihara

with Buddha images, a temple of Kesava (Visnu), and several other

shrines ; other sacred structures were erected by his ministers. Besides

Jayapura the king built on ground recovered from the lake another

place, called Dvaravati, in imitation of Krsna’s famous town by the

sea-shore. Kalhana notes that in his own time Jayapura was popularly

designated as the ‘Inner Castle’ (abhyantara kotta) while Dvaravati

was known as the ‘ Outer Castle * ( bahya kotta).

The present name AndArkoth (from Skr. * Antarakotta) is the direct

derivative of this popular designation of Jayapura. It has in the course

of time been extended also to the site on which originally Dvaravati

stood. In my note on the passage I have shown that Jayapura must be

identified with the island portion of And^rkoth, while the remains in

that part of the village which lies on the lake shore opposite, belong to

Dvaravati. These remains are far less extensive than those on the

island. This is in full agreement with the fact that Kalhana men-

1 For a detailed note on the position of the twin towns Jayapura- Dvaravati,

see Rdjat. iv. 501-511. For a map showing the site on a larger scale refer to Note

I, v. 97-100.

2 See Report, pp. 13 sqq. where the topography and ruins of And^rkdth are

described in detail. General Cunningham had already heard of the identity of

And^rkoth with Jayapida’s town hut he does not seem to have visited the place ;

Anc. Geogr., p. 101. Owing to the erroneous location of Parihasapura on the right

bank of the Vitasta opposite Sambal, there is a good deal of confusion in his notes

on the two capitals.

198

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

tions great religious buildings- only in Jayapura and not in Dvaravati.

The latter is, indeed, referred to only in connection with the foundation

of Jayapura and does not appear ever to have been a place of import¬

ance. We can thus understand why its original name Dvaravati and

its subsequent designation ‘ Outer Castle ’ have both completely dis¬

appeared. The distance between the island and the opposite lake shore

being only about four hundred yards at the narrowest point, the name

of the far more important ‘ Inner Castle ’ was naturally extended also

to this outlying suburb.

The term Icotta which Kalhana repeatedly applies to Jayapura, and

which is contained also in its popular designation, is justified by its

position surrounded on all sides by water.1 The limited extent of the

island precludes the belief of Jayapura ever having been a populous

place. But it retained a certain importance far longer than Parihasa-

pura and served occasionally as a royal residence even in late times.

Queen Kota, the last of the Hindu rulers of Kasmir, retired to Jayapura,

and there she was murdered by her husband, the adventurer Shalimir

(a.d. 1339). 2 Zainu-l-‘abidm restored the town which had fallen into

decay and built there a new palace on the lake-shore. s

We have no distinct information as to the old course which the

Vitasta followed in the neighbourhood of Jayapura previous to Avanti-

varman’s regulation. If our views on the subject as above indi¬

cated are right, the main channel of the river must then have passed

through the marshes west of Jayapura. Notwithstanding the change

subsequently effected, Jayapida’s town did not lose its convenient access

to river communication. The great canal known as Nor which, as we

saw, is in reality nothing but an old river-bed, runs but a short distance

to the south-west of And^rkoth. A branch of it which is much used by

boats even at the present day though not shown on the map, passes still

actually along the old Ghats on the south side of the And^rkoth island.

It seems probable that Jayapura owed its preservation from the fate of

Parihasapura in part at least to the retention of a convenient waterway.

In Abu-l-Fazl’s time And^rkoth gave its name to a separate small

Pargana.

123. From the marshy tracts south of the Yolur which we have

approached at And^rkoth, we may return once

more to Par^spor. Crossing the swamps

formed west of the Par^spor plateau by the

District of Bhangila ;

Pattana.

1 Rdjat. iv. 506, 512 ; vii. 1625. Srlvara, iv. 540, 545, applies to Jayapura the

expression durga, ‘ fort.’

2 See Jonar. 300.

8 See S'riv. i. 250 sqq.

1899.]

SOUTHERN DISTRICTS OF KRAMARAJYA.

199

Sukhnag and other hill streams, we come to the considerable district

of Bdngil. It is often referred in the Rajatarahgini and the other

Chronicles by its ancient name of Bhangila.1 No old localities belong¬

ing to it are mentioned in our texts, unless we may count with Bangil

the closely adjacent Patan situated on the shore of the Pambasar marsh,

circ. 74° 37' long. 34° 10' lat.

This large village occupies the site chosen by King S'amkaravar-

man (a.d. 883-902) for the town which was to bear his name.2 Kalhana,

however, informs us that Samkarapura “ subsequently lost its proper

appellation and became known only by the name Pattana, ‘ the town.’ ” 3

This somewhat general designation still survives in the present Patan.

Kalhana sees in this disappearance of the original appellation the just

retribution of fate for the king’s cruelty and other bad qualities. Yet

the old name must have long lingered on by the side of the popular

‘ Pattana.’ For Ksemendra mentions S'amkarapura, and Kalhana him¬

self speaks of the ‘ town of S'amkaravarman ’ when subsequently refer¬

ring to events of his own time.4 * Pandit tradition too has retained a

recollection of the founder of Pattana and its original name.

S'amkaravarman is said to have carried off “ whatever was of value

at Parihasapura,” in order to raise the fame of his own town. At the

same time Kalhana plainly tells us that “ what gave fame to that town

was only what is still to be found at Pattana, — manufacture of woollen

cloths, trade in cattle, and the like.”6

The only ancient remains of any pretension which can now be

found at Patan, are, in fact, the ruins of the two temples which were

erected there by S'amkaravarman and his queen Sugandha.6 These

shrines which bore the names of S'amkaragaurisa and Sugandhesa are

structures of no great dimensions and are without the fine quadrangular

courts which enclose all more important Kasmirian temples. They

have been fully described by General Cunningham and others. Kalhana

when mentioning these buildings ironically alludes to kings who like

bad poets take the materials for their works from others’ property.

This combined with the immediately following mention of S'amkaravar-

man’s exploitation of Parihasapura, makes it probable that the building

materials for these very temples were taken from the ruins of Parihasa¬

pura. This could have easily been done, owing to the convenient water-

1 See Rajat. vii. 49S note.

2 See Rajat. v. 156 note.

3 Compare v. 213.

4 Compare Samay. ii. 13 ; Rajat. viii. 2488, 3130,

6 Rajat. v. 161 sq.

3 Compare Rajat. v. 158 note.

200

ANCIENT GEOCRAPHY OF KASHMIR. [Extra No. 2,

route offered by the marshes which stretch between Par^spor and

Patan, a distance of only seven miles.

Though S'amkarapura owed thus to its founder but little that

could secure distinction, yet the site he had chosen for it was one likely

to retain some importance. Patan still lies on the direct road between

Srinagar and Baramula, reckoned at two daily marches, and has probably

always just as now been the half-way station between the two places.

Considering that Baramula is the starting point of the route to the

west, traffic and trade were thus sure to be attracted to S'amkaravar-

man’s town. We find it referred to as a local centre still in Kalhana’s

time, and it has remained to the present day a large and thriving place.

Patan figures as a separate Pargana in Abu-l-Fazl’s list. A

popular tradition has it that when Todar Mai, Akbar’s minister, was

arranging for the redistribution of Parganas, he inadvertently omitted

the Patau village at which he was just then encamped. To remedy the

mistake Patan with its immediate vicinity was made into an additional

Pargana.1 However this may be, we find Patan subsequently named

as the chief place of the Tilagam Pargana.2 At the last settlement it

became the headquarters of one of the new Tahsils.

The Pambasar lake which stretches to the east of Patan as far as

the ‘ Gond Ibrahim ’ and i Adin River ’ of the map, is referred to by

Kalhana under the name of Pampasaras. King Harsa seems to have

extended or regulated it.3 The Karewa ground to the west of Patan

with the deep valleys which intersect it, forms the Pargana of Tilagam.

It is mentioned in the Fourth Chronicle, 780, by the name of Taila-

grama.

About four miles to the north-west of Patan and on the high road

to Baramula lies Tapar, a considerable village,

ra apapura On the evidence of an old gloss and several

passages of the Chronicles, it can be safely identified with the ancient

Pratapapura.4’ The latter was founded by King Pratapaditya-Dur-

labhaka, the father of Lalitaditya, probably in the second half of the

seventh century. When visiting the place in 1892 I found close to the

road two ruined mounds covered with large slabs and architectural

fragments evidently marking the sites of old temples. Since then, I

am informed, most of these remains have been turned into road metal

by the native contractors employed in the construction of the new cart-

road to Srinagar.

1 See Bates, Gazetteer, p. 2.

2 See Mogrcroft, ii. p. 113; Vigne, ii. 16G.

$ See Rdjat. vii. 940 note.

* Compare Rdjat. iv. 10 note.

1899*]

SOUTHERN DISTRICTS OF KRAMARAJYA.

201

124. The district through which the Vitasta flows immediately

District of Kruhin before leaving the Valley, bears now the name

of Kruhin. The ancient form of this name is

unknown unless the Lokaprakasa’s ‘ Krodhanavisaya’ may be connected

with the tract. Kruhin extends along both sides of the river, but its

greater portion lies on the left bank.

Proceeding on the road towards Baramula and at a distance of

about six miles from the latter place, we pass on our right the village

of Kanispor. It is identified by an old glossator of the RajataranginI

and by the Persian Chroniclers with the ancient Kaniskapura. The

latter is mentioned in the RajataranginI as a town founded by the

Turuska king Kaniska,1 whom we know as the great Indo- Scythian or

Kusana ruler from the coins and Buddhist tradition. There are no

conspicuous remains above ground at Kanispor, but old coins and carved

stones are occasionally extracted from an old mound near the village.

We have already had occasion to speak of the important position

Varahamula occupied by the ancient twin towns Hug kapur a

and Varahamula. Built on the banks of the

Vitasta immediately above the gorge through which the river leaves

the Valley, they form the starting-point on the great route of com¬

munication to the west. It is unnecessary to refer here again to the

commercial and other advantages which have made this site one of

great importance from ancient times to the present day.

Varahamula, situated on the right river-bank, has left its name to

the present town of Varahmul , usually called Baramula by Panjabis

and other foreigners.2 The name Varahamula or Varahamula — both

forms occur in our texts — is itself derived from the ancient Tlrtha of

Visnu Adi-Varaha who was worshipped here evidently since early

times. Prom it the site of the town and its whole neighbourhood

received also the designation of VarahaJcsetra. Various legends related

at length in the Varahaksetramahatmya and often alluded to in the

Nilamata and the other Mahatmyas, connect this sacred site and the

Tirthas of the immediate neighbourhood with the Varaha or Boar

incarnation of Visnu. An abstract of these legends as well as an

accurate description of the scanty remains of ancient date to be found

at the several Tirthas, has been given by Prof. Biihler.3

1 Compare Rdjat. i. 168 note. General Cunningham’s suggested identification

of Kaniskapura with ‘ Kampor,’ on the road from S'rinagar to S'upiyan, is unsupport¬

ed by any evidence. The place is really called Khampdr and has no ancient

remains whatever.

* For detailed references regarding Varahamula and Varahaitsetra , see Rdjat.

vi. 186 note.

8 See Report, pp. 11 sqq.

J. i. 26

202

AKCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

Tlie ancient temple of Varaha which seems to have been one of the

most famous shrines of Kasmir, is repeatedly

Temple of Varaha. mentioned by Kalhana. According to the

tradition of the local Purohitas it stood near the site of the present

Kotitirtha , at the western extremity of the town and close to the river-

bank. Some ancient Liiigas and sculptures found at the Kotitirtha

may have originally belonged to the temple. The destruction of its

sacred image is noted by Jonaraja in the reign of Sikandar But^hikast.1

A short distance below this site where a steep spur runs down to

the river-bed, stood the ancient watch-station, still known as Drang ,

which has already been described. A bridge over the Vitasta existed

at Varahamula already in old times.2 *

It cannot be doubted that Varahamula is a very ancient place.

It enjoyed the advantage of being on the right river-bank, which is

followed by the old route down the Vitasta Valley. But on the other

hand the contracted nature of the ground which it occupies, between the

hill-side and the river, did not favour the development of a large town.

On this account we find that the twin town of Huskapura built on the

open pin in of the opposite bank was in ancient times the larger of the

two places.

Huskapura is mentioned by Kalhana as the town built by King

Huska, the Turuska, and is often referred to

Huskapura. . * . , \ ,. „ T,

m his subsequent narrative.6 Its name sur¬

vives in that of the small village of Uskiir , situated about two miles to

the south-east of the present Baramula. The identity of Uskiir and

Huskapura, correctly noted already by General Cunningham,4 * is well-

known to STinagar Pandits, and is indicated also by an old glossator of

the Rajatarangini. Kalhana in one passage distinctly includes Hus¬

kapura within Varahak^etra, i.e ., the sacred environs of the Varaha

Tirtha,6 and the same location is implied by numerous other references

in the Chronicle.

King Huska of the Rajatarangini has long ago been identified with

the Indo- Scythian ruler who succeeded Kaniska, the Huviska of the

inscriptions and the OOHpKI of the coins. The foundation of Huska¬

pura falls thus probably within the first century of our era. Hiuen

Tsiang, as we saw, spent his first night after passing through the

western entrance of the kingdom, in a convent of Hn-se-kia-lo or Hus¬

kapura. Alberuni too knows ‘ Ushkara,’ opposite to Baramula.

1 Compare Jonar. 600.

2 See Rajat. viii. 413.

8 For detailed references as to Huskapura : Uskiir, see Rajat. i. 168 note.

4 See Anc. Geogr., pp. 99 sq.

6 See vi. 186.

1899.] SOUTHERN DISTRICTS OF KRAMAllAJYA. 20S

Kalhana mentions Huskapura far more frequently tlian Varaha-

mula. The conclusion to be drawn herefrom as to the relative impor¬

tance of the two places in Hindu times, Is confirmed by the frequent

references which the Chronicle makes to religious buildings erected in

Huskapura. Of King Lalitaditya-Muktaplda it is recorded that he

built there the great temple of Visnu Muhtasvamin and a large Viliara

with a Stupa.1 * Ksemagupta who sought the sacred soil of Varahaksetra

in his fatal illness, had founded two Mathas at Huskapura.8

At present foundations of ancient buildings can be traced at

numerous points of the plain which stretches from the left river-bank

towards the low hills behind TJ§kiir. These remains as well as two

colossal Lihgas still in situ have already been noted by Bishop Cowie.3

About 400 yards to the west of the village are the much-damaged

remains of a Stupa, which had been found still intact by Bishop Cowie and

photographed in that condition by Major Cole (1870). Subsequently it

was dug into and partly levelled down “by some Sahib’s order,” as the

villagers told me. Of this excavation I have not been able to trace a

report. But General Cunningham refers to an ancient coin of the

Taxila type which was found in this Stupa and had come into his

possession.4 5

It is possible that this Stupa was identical with the one which

King Lalitaditya erected at Huskapura. Of the Vihara which Kalhana

mentions in connection with the king’s Stupa, I have shown elsewhere

that it was in all probability the same convent which Ou-k'ong refers

to under the name of Moung-ti Vihara.6 The Moung-ti of the Chinese

transcription seems to represent a prakritized form of the shortened

name MuJcta or Muktfi. The latter forms which are abbreviations

( bhimavat ) for Muktapida , occur also in the designations of other re¬

ligious buildings erected by that king (Muktakesava, Muktasvamin).

As we do not meet with the name of Huskapura in any of the

later Chronicles it may be assumed that its importance did not survive

the time of Hindu rule.

I See Rajat. iv. 188.

8 Rajat. vi. 186.

S See J. A. S. B., 1866, p. 123.

4 See Coins of Anc. India, p. 62.

5 Compare Notes on Ou-k'ong, pp. 6 sqq. ; Rajat. iv. 188 note.

204

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Section IX. — The Northern Districts op Kramarajya.

125. The ancient localities in the Vitasta Valley below Varaha-

miila have been noted by us already in connection with the route which

leads through it. We may therefore proceed now to those Parganas

of the ancient Kramarajya which lie to the north of the river and the

Volur lake.

The district which adjoins Kruliin in this direction, is known as

Hamal (map ‘Hummel’). Its ancient name

District of Samala. c - _ j. i . T . , £ •, . ,.

was Samala from which the former designation

is the direct phonetic derivative.1 S'amala is very frequently mentioned

in the last two Books of the Rajataraiigini, particularly on account of

its feudal chiefs or Damaras who played a prominent part in all the

civil wars of the later reigns. The pretender Bhiksacara in particular

had his most powerful adherents in S'amala and often took refuge with

them. The village of Vanagrama which is mentioned on one of these

occasions, is probably identical with the present Vangam , situated circ.

74° 25' long. 34° 19' lat.2 3 Kakaruha , another place in S'amala, referred

to in connection with Bhiksacara’s campaigns, can no longer be traced.

To the north of Hamal we reach the Pargana of Machipur

(map ‘ Mocliipoora.’) Its ancient name is no¬

where mentioned. In it lies the sacred site

of Svayambhu which owing to the apparently

volcanic phenomenon there observed has from early times been renowned

as a Tirtha. Kalhana in his introduction duly notes the ‘ Self-created

Eire’ ( Svayambhu ), which “rising from the womb of the earth,

receives with numerous arms of flame the offerings of the sacrificers.” s

The spot meant is still known as Svayambhu , or to the villagers as

Suyam. It lies on a low ridge about half a mile south-west of the

village of Nich^hom (not shown on map) and about one and a half miles

north of Tsakavadar (map ‘ Sheikwadda.’) Visiting it in 1892 I found

there in a shallow hollow the soil bright red like burned clay and

furrowed by narrow fissures. In certain years steam has been known

to issue from these fissures. The ground then becomes sufficiently hot

to boil the S'raddha offerings of the pilgrims who at such times flock to

the site in great numbers. The phenomenon which may be either truly

volcanic or, according to a modern authority, be caused by hidden seams

of coal taking fire, was last observed in the year 1876. Occurrences at

Tirtha of

Svayambhu.

1 See Rdjat. vii. 159 note.

2 See Rdjat . viii. 1438,

3 See Rdjat. i. 34, and for further references the note thereon.

1899.] THE NORTHERN DISTRICTS OF KRAMARAJYA.

205

the beginning of the present century are referred to by Mr. Vigne and

Dr. Falconer.1 Abu-1-Fazl too mentions the phenomenon at ‘Soyam.’2

Considering the rarity of the occasions when this manifestation of

the ‘ Self-created Fire ’ is observed and the pilgrimage performed, the

total absence of ancient remains cannot surprise us. There is, however, a

Mahatmya of the Tirtha, and the latter is also referred to in the

Nilamata. A pilgrimage which King Uccala (a.d. 1101-11) made to

Svayambhu gives Kalhana occasion to. acquaint us with some localities

of the neighbourhood.3 The king who was stopping in Kramarajya,

is said to have started for the village of Varhatacakra with a small

retinue to see the miracle there. On his way which took him past the

village of KambaleAvara, he was set upon in a deep mountain gorge by

robbers from whom he escaped only with difficulty.

I believe, the places mentioned in connection with this adventure,

can still be identified without difficulty. Varhatacakra is probably the

present Tsakavadar, tsa7ca being the ordinary Ks. form for Skr. cakra and

vadar the phonetic derivative of Varhata -.4 Cases of village names in

which the two component parts, being originally distinct names, can

alternate in their position, are by no means unfrequent in Kasmir.

Thus we have now Dara-Sad^por and Sadvpor-Dara, etc.

In Kambalesvara we may safely recognize the present village of

Krambhar , situated about six miles north-east of Svayambhu ; for the

ending -har as the derivative of Skr. -es'vara, compare Triphar < Tripur es-

vara , etc.5 The way from Krambhar to Svayambhu leads through the

valley of the Panjtar stream. The latter as I convinced myself by

personal inspection on a tour in 1892, passes above Rajpor a narrow

thickly- wooded gorge. The path which follows the tortuous course of

the stream at the bottom of the gorge, offers excellent opportunities for

an ambuscade such as described by Kalhana.

Badarkal, a small village, about four miles south-east of Krambhar,

has a small local Tirtha marked by a spring and some old Lin gas. It

is visited on the pilgrimage to Svayambhu and mentioned by the name

of Bhadrakali in the Mahatmya of the latter.

126. The Pargana of Uttar stretching along the foot of the range

District of Uttara towards the Kisanganga, forms the extreme

north-west of the Kasmir Valley. A passage

1 See Vigne, Travels , ii. p. 280 ; Lawrence, Valley , p. 42.

2 Ain-i-Akb., ii. p. 365.

8 Compare Bdjat. viii. 250 sq. note.

4 For medial Skr. r > Ks. d , compare e.g.> Bhattdraka[ma{ha > Brqd*[mar

for t > r, e.g. Kasthcivata > Kasi^var.

8 For detailed evidence on the phonetic points alluded to, see Rdjat. viii, 250 note.

206

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

of the Rajntarangini records its ancient name Uttara, and refers also to

Ghosa as a locality situated in it.1 The place meant is undoubtedly

the present Gus situated in the centre of Uttar, near the confluence of

the Kamil River and the stream coming from Lolau. It is the starting-

point for the S'arada pilgrimage and is mentioned correctly as Ghosa

in tho S'aradamahatmya.

About ten miles higher up the Kamil river lies the village of

Panz^am, circ. 74° T long. 34° 29f lat. I take its position from Major

Bates’ Gazetteer; the ‘Atlas of India’ map does not show the place.

It is in all probability identical with PancigramT, mentioned by Kalhana

in connection with the surrender of the pretender Bhoja.2 I have not

been able to visit this portion of the district, and Major Bates’ reference

to Panzigam attracted my attention only after the preparation of my

map.3

In the extreme north-east of Uttar and within a mile of each other,

we have the old villages of Drang and HayZhom, referred to by Kalhana

under their ancient designations of Dranga and Hayasrama.4 * The

latter place, as its name shows, marks the position of an old frontier

watch- station towards the Kisangariga. We have already seen that

there is a route leading past it to S'ardi, the ancient Tirtha of S'arada

situated on that river.

Dranga and Hayasrama are both mentioned hy Kalhana in connec¬

tion with the siege of the Shrahsila castle which took place in his own

time. A brief reference may therefore be made here both to this

stronghold and the neighbouring shrine of S'arada, though they are both

situated outside the limits of the Kasmir Valley.

127. The introduction of the Rajatarahgini mentions the temple

of the goddess S'arada amongst the foremost

Tho Tirtha of Sarada* m- i-i ^ ~\r~ / _ r t. ,

Tirthas of Kasmir.6 It was well known even

far beyond the frontiers of Kasmir. Alberuni had heard of it,6 and a

story recorded in a Jaina life of the great grammarian Hemacandra

proves that its fame had spread even to far-off Gujrat.

1 See Rajat. vi. 281.

8 See Rajat. viii. 3124.

8 There seems to be good reason to suspect that Taramulaka, a place repeated¬

ly referred to in connection with Bhoja’s last campaign, lay somewhere in or near

Uttar. Unfortunately this locality which is of importance also for other portions

of Kalhana’ s narrative, has not yet been identified ; see note vii. 1307.

4 For Hdija^rama, see Rajat, viii. 2937 note ; for Dranga , viii. 2507 note, also

Note B, i. 37.

6 The position and history of the temple of S'arada have been fully discussed in

Note B, i. 37.

® See India , i. p. 117.

1899.]

THE NORTHERN DISTRICTS OF KRAMARAJYA.

207

Notwithstanding this former celebrity the S'arada shrine is now

almost completely forgotten by the Pandits of S'rinagar and the great

mass of the Brahman population of the Valley. Fortunately, however,

tradition had been more tenacious in the immediately adjoining tracts

of Kamraz. Guided by it I was able to ascertain the position of the

ancient Tirtha at the present Sardi , situated circ. 74° 15' long. 34° 48'

lat., on the right bank of the Kisanganga.

My note on Rajat. i. 37 ( B ) gives a detailed account of the tour

which in 1892 led me to the Tirtha as well as a description of the

ancient temple still extant at the site. The situation of the shrine

corresponds exactly to Kalhana’s description. Immediately in front of

it the sacred stream of the MadhumatI falls into the Kisanganga, while

another confluence, that with the Sarasvati river coming from the

north, is also visible from the temple.

In Jonaraja’s time the shrine was still sufficiently popular to attract

a yisit even from Sultan Zainu-l-‘abidin.1 2 Soon afterwards appar¬

ently the miracle-working image of the goddess was destroyed. Abu-1-

Fazl, however, still notes the sanctity of the site and correctly indicates

its position on the bank of the MadhumatI.3 *

The subsequent neglect of this Tirtha must be ascribed chiefly to

the obstacles to the pilgrimage which arose from the troubled political

condition of the Upper Kisanganga Valley. The Bomba chiefs of the

latter had made themselves independent in the later Mughal and Pathan

times. Their predatory inroads often threatened the adjacent tracts of

Kasmir while their own territory became practically inaccessible to

peaceful pilgrims. It is only since the advent of the Sikhs that the

pilgrimage to S'arada’s seat was revived. It is probable that the diffi¬

culties here briefly indicated must be held to account for the several

substitute Tirthas of S'arada which are now to be found in various parts

of Kasmir proper.

My visit to the old ‘ S'aradasthana ’ also enabled me to identify with

£irah.sila Castle. certainty the site of the STrahsila Castle. The

latter had been the scene of a memorable

siege by King Jayasimha’s troops which Kalhana describes at length.5

The accurate topographical data furnished in this account prove clearly

that the castle occupied the top of the steep ridge which projects into the

Kisanganga valley about two and a half miles below the S'arada temple.

1 Jonar. (Bo. ed.) 1056-71. This visit apparently took place a.d. 1422.

2 Am-i-Alcb., ii. pp. 365 sq. Abu-1-Fazl places S'arada’s stone temple “ at two

days’ distance from Haehamun,” i.e. Hay^hdm.

5 Rajat. viii. 2492-2709. The position of S'iraTisild and the evidence for its

identity with the ‘ Ganes Ghatl ’ hill have been fully discussed in my Note L, viii. 2492.

208

ANCIENT GEOGRAPHY OE KAS'MIR. [Extra No. 2,

The several incidents of the siege, in particular those connected with

the attempted escape of the pretender Bhoja, became at once easily

intelligible on a close inspection of this site. The ridge bears now the

name of Ganei Qhatl , from a curious rock formation on its side which

resembles the head of an elephant and is accordingly worshipped as a

‘ Svayambliu ’ representation of the elephant-faced god. It is very

probable that the older name Sirahs'ila which means literally ‘ the rock

of the head,’ owed its origin also to this very rock.

128. Returning from our excursion to the Kisanganga and the

confines of the Dard country, we enter im-

Districts^of^Lolau, mediately to the east of Draug-Hay^hom the

Pargana usually called Loldb. Its proper

Kasmirl name is Lalau, derived from Skr. Laulaha.1 In the picturesque

valley which forms this district, no old Jocalities can be specified.

Lolau is adjoined on the south by the Pargana of Zain^gir

which comprises the fertile Karewa tract between the Volur and the

left bank of the Pohur River. It received its present name from

Zainu-l-‘abidin who is credited with having carried irrigation canals

from the Pohur to the Udar ground of Jainagiri.2 The earlier name of

this tract can no longer be traced.

The chief place in it is the town of Sopur , the ancient Suyyapuea,

the foundation of which by Suyya, Avantivarman’s engineer, has

already been mentioned.3 Sopur which lies a short distance below

the point where the Vitasta leaves the Volur, has retained its import¬

ance to this day, and is still a town of over 8000 inhabitants. It

has during recent times been the official head-quarters for the whole

of Kamraz. From a passage of S'rivara it appears that this had

been the case already at an earlier period.4 * Relating a great confla¬

gration which destroyed Suyyapura in Zainu-1-Cabidin’s time, this

Chronicler tells us that in it perished the whole of the official archives

relating to Kramarajya. The royal residence, however, escaped and the

town itself was again built up by the king in great splendour. Of

this, however, nothing has remained ; nor does the town now show

older remains of any interest.

The suggested identity of the village Zdlur (map ‘Zohlar’) in the

north-west part of Zain^gir with Jalora once mentioned as a founda¬

tion of King Janaka,6 is doubtful, resting only on the resemblance of

1 Compare Rdjat. vii. 1241 note.

2 See Jonar. (Bo. ed.) 1449-56 ; also S'riv. i. 562 sq. ; iii. 59, 78,

8 Compare for Suyyapura, Rdjat. v. 118 note.

4 Compare S'riv. i. 560 sqq ,

6 See Rdjat. i. 98.

1899.]

THE NORTHERN DISTRICTS OF KRAMARAJYA.

209

the names. The large village of Bumai (map 1 Bamhai ’), situated

74° 30' long. 34° 22' lat., may be Kalhana’s BhImatika. The name

Bumai can be traced back without difficulty to the older form ; but the

context of the single passage in which BhImatika is mentioned, does

not supply any evidence as to its location.1

Bound the north shore of the Volur lake there stretches in a semi-

t „ circle the district of Khuy^hom. Its ancient

District of . . n Tr .

KhuyaSrama name is given by Kalhana as Khuyasrama

while Shlvara and the Lokaprakasa, with a

slight variation, call it Khoyasrama.2 The old route which led up to

the Madhumati stream and over the Pass of Dugdhaghata or Dud^khut

into the Darad territory on the Kisanganga, has been already fully

described.3

In connection with a Darad invasion which was directed into

Kasmir by this route, we read of Matrgrama as the place where the

invading force encamped.4 This is certainly the present village of

Matrtgdm situated close to the foot of the Tragabal Pass, circ. 74° 43'

long. 34° 28' lat. It lies just at the point where the route along the

Madhumati debouches into an open valley, and is the first place where

a larger camp could conveniently be formed.

The tract on the north-east shore of the Volur appears in old times

to have formed a separate small sub-division called Evenaka. It is

once mentioned by Kalhana, and also referred to in the Tirthasamgraha.6

But the evidence is not sufficient for a certain location. To it

may possibly have belonged also the village of Sudtrkdth, circ. 74° 43'

long. 34° 18' lat., which S'rlvara refers to by the name of Samu-

drakota.6

129. We have now reached the vicinity of the Sind Valley which

forms the largest of the Parganas of Kasmir.

The district now known as Ldr comprises the

whole of the valleys drained by the Sind and its tributaries as well

as the alluvial tract on the right bank of that river after its entry into

the great Kasmir plain.

Its ancient name was Lahara, and by this it is mentioned in very

District of Lahara.

1 Compare Rdjat. vii. 6; as to the relation of Bumai < Bhimatika comp,

\_zu : Bhima\J>esava.

2 See Rdjat. viii. 2695-98 note.

3 See above, § 56.

* See Rdjat. viii. 2775.

& Compare Rdjat. viii. 2695-98 note*

6 See S'riv. i. 400.

J. i. 27,

210

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

numerous passages of the RajatarahginI and the later Chronicles.1 The

lands of the district seem to have been from early times in the hands of

great territorial nobles. One family of Damaras resident in Lahara was

powerful enough for its members to play the part of true kingmakers

during a succession of reigns following after Harsa.2 It is probable

that the great trade-route to Ladakh and Central Asia which passes

through the district, added already in old times to its wealth and

importance.

In the midst of the wide water-logged tract of the Sind Delta we

find the ancient Tirtha of Tulamulya at the village now known as

Tultmul, situated 74° 48' long. 34° J3' lat. The Purohita corporation

of Tulamulya is represented as a well-to-do and influential body already

under King Jayapida.3 The large spring of Tulamulya is sacred to

Maharajnl, a form of Durga, and is still held in great veneration by the

Brahman population of STinagar. It is supposed to exhibit from time

to time miraculous changes in the colour of its water, which are

ascribed to the manifestation of the goddess. Owing to its convenient

position the Tirtha attracts large numbers of pilgrims from the capital.

Abu-1-Fazl notices the place and its marshy surroundings.4 * About two

and a half miles to the east of TuDmul lies the village of Dud^rhbm, on

the main branch of the Sind which becomes here navigable. It is

repeatedly spoken of by STlvara under its old name of Dugdhasrama.6

Ascending the valley we come to the large village of Mail1 gam,

situated a short distance from the right bank of the river, 74° 52' long.

34° 17' lat. It is the Mayagrama of Kalhana’s Chronicle, mentioned in

connection with a campaign of Bhiksacara in Lahara.6 In the time of

King Samgramaraja (a.d. 1003-28) Mayagrama gave its name to a

separate fund ( Muyagraminaganja ) which Queen S'rilekha had estab¬

lished evidently with the revenue assigned from this village.7 Mafd-

gam-Mayagrama still owns a large area of excellent rice-fields. The

village itself contains no ancient remains ; but a short distance above

it, at the foot of the spur which descends from a high alp known as

1 Compare for the identification of Ldr and Lahara, note Bdjat. v. 51. The

authors of the St. Petersburg Dictionary were already aware of it ; see P. W.

s. v. Lahara.

i Compare regarding the political part played by Janakacandra, Gargaeandra

and their descendants, Bdjat. viii. 15 sqq., 354 sqq., 502 sqq., etc. For an earlier

instance of Damara power in Lahara, see v. 51 sqq.

3 See Rdjat. iv. 638 note.

4 Ain-i-Akb., ii. p. 364.

b S'riv. iv. 110, 136, 263.

# See Bdjat. viii. 729.

1 Compare Rdjat. vii. 126.

211

1899.]

THE NORTHERN DISTRICTS OF KRAMARAJYA.

Tlrtha of Ciramocana.

Mohand Marg , there is an ancient stone-lined tank filled by a fine

spring known as Vutasan Nag. This is visited as a Tlrtha by the

Brahmans of the neighbourhood and is also mentioned under the name

of JJccaihslrna Naga in the Haramukuta and several other Mahatmyas.

About a mile above the village the high-road leading up the valley

passes a shapeless mound of large slabs which undoubtedly belonged to

an ancient temple.

130. About four miles above Marflgam on the left bank of the

Sind we reach a site which has enjoyed sanctity

from an early period. Close to the village

of Prang (not shown on map) situated circ. 74° 55' 30" long. 34° 16' 45"

lat., a small branch of the Kankanai River ( Kanakavahini ) flows into the

Sind. This confluence is now visited by the pilgrims proceeding to the

Haramukuta lakes as one of the chief Tlrthas on the route. In the

modern Haramukuta Mahatmya it is designated as Karahkatlrtha.

Bat I have shown that it is in reality identical with the ancient Tirtha

of Ciramocana mentioned in the Rajataraiigini, the Nilamata and the

old Nandiksetramahatmya.1

The Kankanai or Kanakavahini which is always named together

with Ciramocana, is a sacred river as it carries down the waters of the

holy Gaiiga-lake below the Haramukuta Peaks.2 This explains the

importance attached to this ‘ Samgama.’ The Haramukuta Mahatmya

which shows its comparatively recent origin by many of its local names,

metamorphoses the old Kanavahini into Karahkanadi and consequently

also changes the name of its confluence into Karankatlrtha .3 King

Jalauka, the son of Asoka, whom the Chronicle represents as a fervent

worshipper of Sflva Bhutesa and of Nandisa, is said to have ended his

days at Ciramocana.

Our survey has already taken us to the sacred sites of Bhutesa and

Jyestharudra marked by the ruined temples

at the present Buth'ser high up in the Kanka-

nai Valley. They are closely connected with

the Tirthas of Nandiksetra below the Haramukuta glaciers which

have also been described.4 The village of Vangath, which is the highest

permanently inhabited place in the valley, lies about two miles below

Butlflser. It is named VasjSthasrama in the Mahatmyas and believed to

mark the residence of the Rsi Vasistha. Allusions in the Rajataraiigini

and Nilamata show that this legendary location is of old date.5

Tirthas of Bhutesa,

. Jyestharudra.

1 See Rdjat. i. 149-150 note.

2 See above, § 57.

3 Regarding the local nomenclature of this Mahatmya, see above, § 31.

4 See above, § 57 ; also Rdjat. notes i. 36, 107, 113 ; v. 55-59.

6 Compare Rdjat. viii, 2430 note.

212

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Upper Sind Valley.

At the mouth of Kankanai Valley, and about two miles to the

north-east of Clramocana, is the hamlet of Baravul which Kalhana

mentions as an Agraharaof King Jalauka under the name of Varabala.1

A large sculptured Lihga base which 1 found here in 1891, shows the

antiquity of the place.

131. Returning to the main valley we come, about three miles

above Clramocana, to the large village of

Kangan situated on the right bank on the Sind.

It is, perhaps, identical with Kankanapura which Queen Didda is said to

have founded in commemoration of her husband Ksemagupta, known

by the epithet of ‘ Kankanavarsa.’ a No old localities can be identified

with certainty in the Sind Valley until we reach the village of Gaganglr ,

situated two marches above Kangan, circ. 75° 15' long. 34° 18' lat. This

is undoubtedly the Gaganagiri of Jonaraja, and the Fourth Chronicle.3

The place is mentioned in both texts in connection with invasions which

were made into Kasmir over the Zoji-La Pass. The first was that of

the Bhautta Rincana, the second the famous inroad of the Mughal

leader Mirza Haidar (a.d. 1532).4 The account which the latter himself

has left us of his exploit, fully explains the special reference made

to Gaganagiri by the Hindu Chronicler.

About three miles above Gagangir two rocky spurs descend from

opposite sides into the valley and reduce it to

a narrow gorge (see map). The passage of

this defile was until recent improvements of

the road distinctly difficult, as large fallen rocks blocked the narrow

space between the right bank of the river and the high cliffs rising

above it. It is at this point of the valley which Mirza, Haidar calls

‘ the narrow defile of Lar/ that the Kasmir chiefs vainly attempted

to stop the brave Turks of the invader’s advanced guard.

Kalhana’s Chronicles shows that the defile here indicated had

witnessed fighting already at an earlier epoch. When King Sussala’s

forces had driven Gargacandra, the great feudal chief, from his seats in

Lahara, we are told that the Damara with his followers retired to the

mountain called Dhudavana. There he was long besieged by the troops

Defile of

Dhudavana.

1 See Rajat. i. 121 note.

8 See Rajat. vi. 301.

8 Compare Jonar. (Bo. ed.), 197, and Fourth Chron. 316. The old name of the

locality ought to have been entered in the map. The Bombay edition of the Fourth

Chron. wrongly reads gamananiryanta for gaganagiryanta of the MSS.

* See TdrtJch-i RasKidl, p. 423. Mr. Elias in his note on the passage has quite

correctly identified the defile meant by his author. The Fourth Chronicle names

the autumn of the Laukika year [460]8 as the date of the event which agrees

exactly with Mirza Haidar’s a.h. 939 Jamad II (December, 1532 a.d.).

1899.] THE northern districts of kramarajya.

213

“ of the king who was encamped at the foot of the mountain.” In my

note on the passage I have shown that the name Dhuddvanci survives in

Durun Ndr (map ‘Darnar’), the appellation of the high spur which

descends into the Sind Valley from the south between Gagangir and

Sunamarg.1 It is exactly at the foot of this spur that the river passes

through the gorge above described. The position taken up by the

king’s opponent is thus fully explained.

Gagangir being already 7400 feet above the sea, is the last per¬

manently inhabited place in the valley. Some twenty-five miles higher

up we arrive at the Zoji-La Pass. Here we have reached the limits of

Kasmir as well as the end of our survey.

1 S eeRdjat. viii. 595 sqq. — Durun is the direct phonetic derivative of Dhuddvana.

Ndr , the Ks. equivalent of Skr. nada, the Anglo-Indian ‘ Nullah/ is often found as the

second part in names of high hill-ranges in Kasmir; compare, e.g.t the ‘Soornar’

and ‘ Baibnar ’ of the map, east of the Haramukh Peaks.

Serial No.

2U

ANCIENT GEOGRAPHY OF KAS'MIR.

[Extra No. 2,

SUPPLEMENTARY NOTE AA.— § 29.

Mahatmyas of KasmIrian Tirthas.

The following is a complete list of the Mahatmya texts acquired by

me in Kasmir. The numbers in the fourth column refer to the manu¬

scripts representing these texts in my collection. Where the same text

is found in several Manuscripts, the number of the best copy has been

shown first. In the column of Remarks the Tirtba to which the Mahat¬

mya refers has also been indicated, as well as the paragraph of the

present Memoir in which it has been, discussed.

1899.]

M AH ATM Y AS OF KAS'MIRIAN TlRTHAS.

215

23 G

ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2

MAHATMYAS OF KAS'MIRIAN TIRTHAS.

217

1899.]

J. i. 28

218

ANCIENT GEOGRAPHY OF KASMIR.

[Extra No. 2,

SUPPLEMENTARY NOTE BB. — § 86.

The lists of KasmTr Parganas.

The following table shows the names of Kasmir Parganas as con¬

tained in the lists of Abu-1-Fazl, Moorcroft, Yon Hiigel, Yigne and

Major Bates. The list of the latter ( Gazetteer , p. 2) is the fullest and

also indicates the division of the Parganas between Maraz and Kamraz

as shown in the table.

The arrangement of the Pargana names conforms to the topogra¬

phical order followed in Chapter IV. of this Memoir. The second

column gives the Pargana names according to their present Kasmiri

pronunciation, without regard to the often curiously distorted forms

in which these names are presented by the earlier lists. The third

column shows the authorities in whose lists each particular name is

found. In the fourth column the Sanskrit name of the district has

been indicated, whenever known, together with the text in which it

first occurs. In the last column references have been given to the

paragraphs of the Memoir specially dealing with the historical topo¬

graphy of the several districts.

Besides the Parganas shown in the table, Abu-l-Fazl counts with

Kasmir the Parganas of ‘ Maru Adwin ’ (Madivadvan), ‘ BanihaV

(Banahal), and ‘ Dachhin Khawarah’ ( i.e ., Dachiin-Khovur, the valley

on both banks of the Vitasta below Baramula). In the same way

Moorcroft adds the Parganas of ‘ Durbid’ (i.e., Dvarbidi in the Vitasta

Valley, comp. § 53), ‘ Karnao ’ and ‘ Tahirabad’ ; of the position of the

last-named tract I am not certain.

[The list of Kasmir Parganas given by P. Joseph Tieffenthalkr,

Description de VInde, p. 77, is only a defective reproduction of Abu-1-

Eazl’s list. Ritter, Asien, ii. pp. 1136 sq. , has endeavoured to elucidate

219

1899.] THE LISTS OF KAS'MIR parganas.

Tieffen thaler’s queer names with whatever materials were then avail¬

able] .

List of Abbreviations.

AF. Abu-l-Fazl, Ain-i AJcb., ii. pp. 368-371.

Mo. Moorcroft, Travels , ii. p. 113.

Hii. Yon Huger, Kaschmir, ii. 206 sq.

Yi. Vigne, Travels , i. p. 272.

Ba. Bates, Gazetteer, p. 2.

Parganas of Maraz (Madavara.tya).

220

ANCIENT GEOGRAPHY OF KASMIR. [Extra No. 2,

Parganas of Kamraz (Kramarajya.)

1899.]

THE LISTS OF KASHMIR PARGANAS.

221

222

ANCIENT GEOGRAPHY OF KAS'MlR.

[Extra No. 2,

Additional Note. — § 14.

Mr. W. Merk, c.s.i., c.s., at present Deputy Commissioner of

tlie Hazara District, to whom I submitted the above-printed remarks

on AlberunI’s route to Kasmlr (p. 23) and on the town of Babrahan

mentioned by him, has kindly supplied me in a letter, dated 25th April,

1899, w7ith the interesting information that “ the basin formed by the

three Nullahs which unite at Chamhad is called the ‘ Babarhan ’ tract.”

Chamhad is shown on the map as a village in the Mian Khaki Nullah,

south-west of Abbottabad, circ. 34° 7' lat., 73° 7' long.

Mr. Merk believes that the position of Alberuni’s “ town Babrahan ,

half-way between the rivers Sindh and Jailam” is marked by the

present ‘ Babarhan ’ which practically retains the same name. There

is much to support this identification. The Nullah called Mian Khaki

on the Survey map forms a convenient route from the central plain

of Hazara, about Mirpur, towards the Siran Valley through which

the Indus could conveniently be gained at Torbela. The latter has

remained a favourite crossing place to the present day, being situated

just where the Indus debouches from the mountains.

The 8 Farsakh or about 39 miles which Alberuni counts from

Babrahan to “ the bridge over the river,” i.e., according to my explana¬

tion, the present Muzaffarabad, would well agree with the actual

distance between the latter place and Babarhan. In calculating

this distance it must be kept in view that the old road from the Indus

to Kasmlr, according to Mr. Merk’s information, descended to Grarhi

Hablbulla, on the Kunliar River, through the Doga Nullah, i.e., by a

more direct route than that followed by the modern cart-road vid

Mansahra.

I am further indebted to Mr. Merk for the very interesting notice

that the plain near Mirpur, about 5 miles north-northwest of Abbotta¬

bad, is popularly known by the name of JJrasli or Orash. There cau

be no doubt as to this local name being the modern representative of

the ancient Urasd. Its survival in that particular locality strikingly

confirms the conclusion indicated above in § 83, and also in my note

on Rajat. v. 217, as to the position of the old capital of Urasa. The

designation of this capital was undoubtedly JJrasd .

1899.]

223

TABLE OF CONTENTS.

CHAPTER I.— INTRODUCTORY.

Para.

1. Ancient topography in Kal liana’s Chronicle ...

2. Preparation of maps ...

3. Arrangement of Memoir

4. Earlier Publications (Wilson, Troyer, Cunningham,

Lassen, Biihler)

Page.

1

2

3

CHAPTER IT.— ACCOUNTS OF OLD KAS/MlR.

SECTION I. — CLASSICAL NOTICES.

5. Alexander’s invasion. — Ptolemy’s Kaspeiria ... ... 8

6. The Kaspeiroi of Dionysios and Nonnos ... ... 10

7. Kaspatyros of Herodotos. — Supposed Derivation of ‘Kas-

mira’ from #Kasyapapura ... ... ... 11

SECTION II. — CHINESE RECORDS.

8.

9.

10.

11.

Earliest Chinese notice

Visit of Hiuen Tsiang (a. d. 631)

Kasmir in T’ang Annals

Ou-k'ong (a. d. 759) ...

SECTION III. — MUHAMMADAN NOTICES.

12. Kasmir closed to Arab geographers

13. Alberuni’s interest in Kasmir

14. Alberuui’s account of Kasmir. — Route into Kasmir. — De¬

scription of the Valley. — Description of Pir Pantsal. —

The fortress Lauhur... ...

13

14

17

18

20

21

23

SECTION IV. — INDIAN NOTICES.

15. Deficiency of information in Non-Kasmirian texts

28

224

Pai’a.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

ANCIENT GEOGRAPHY OF KASMIR.

SECTION V. — THE KASMIR CHRONICLES.

[Extra No. 2,

Abundance of Kasmirian sources ...

Kalhana’s Rajatarangini

Kalhana’s notices of Tirtbas

Kalhana’s references to foundations

Topographical data in Kalhana’s historical narrative

Accuracy of Kalhana’s topography

Sanskrit form of local names in Rajatarangini. — Official

names of localities

Later Sanskrit Chronicles. — Jonaraja; S'rivara; The

Fourth Chronicle

Persian Tarikhs of Kasmir

Kasmir poets. — Ksemendra. — Bilhana. — Mankha

The Lokaprakasa

»

SECTION VI. — THE NILAMATA AND MAHATMYAS.

The Nilamatapurana. — Condition of text

The Haracaritacintamani

The Mahatmyas. — Their date

Origin and purpose of Mahatmyas. — The local Purohitas...

Popular etymology in local names of Mahatmyas

Abu-l-Fazl’s account of Tlrthas ...

SECTION VII. — LOCAL TRADITION.

Local tradition of the learned

P. Sahibram’s Tirthasamgraha. — Its local names

Popular local tradition

Page.

29

30

31

33

35

37

39

40

43

45

46

48

49

51

53

55

56

59

CHAPTER III.— GENERAL GEOGRAPHY.

SECTION I. — POSITION AND CONFIGURATION OF KASMIR VALLEY.

The name Ka^mira. — Etymologies of name ... ... 61

Extent and position of Kasmir ... ... ... ^3

Legend of Satisaras. — Lacustrine features of Yalley ... 65

Kasmir defended by its mountains ... ... 67

Watch- stations on mountain routes. — Guarding of Gates 68

SECTION II. — THE PIR PANTSAL RANGE.

Kasmir orography. — Eastern portion of Pir Pantgal

Range (Pass of Banasala ; Naubandliana Tirtha ; Sid-

dhapatha) ... ... ... ... 70

1899.]

CONTENTS.

225

Para. Tage.

42. Pir Pantsal Ponte. — S'urapura. — Kramavarta ... 72

43. Hastivanja. — Legend of Miliirakula ... ... 73

44. Pancaladharamatha. — Ksemendra’s description of pass ... 74

45. The name Pancala ... ... ... ... 75

46. ‘Pir,’ a term for pass. — Pancaladeva ... ... 76

47. Pass of Pir Pantsal. — Pusyananada ... ... 78

48. Central part of Pir Pantgal. — Mount TatakutI ... 79

49- Tos^maidan Route. — Karkotadranga. — Lohara ... 79

SECTION III. — THE VITASTA VALLEY.

50. Valley of the Vitasta ... ... ... 82

51. Vitasta Valley Route. — Connection with Urasa ... 83

52. Gate of Varahamula. — Yaksadara ... ... 84

53. Old frontier in Vitasta Valley. — Bolyasaka. — Dvaravati... 85

54. Left bank of Vitasta ... ... ... ... 86

SECTION IV. — NORTHERN AND EASTERN MOUNTAIN RANGES.

55. Range towards Karnau and S'ardi. — Drariga ... ... 88

56. Pass of Dugdhaghata. — Marg : Mathika. ... ... 89

57. Mount Haramukuta. — Tirthas of Nandiksetra, Bhutesvara 91

58. North-eastern range. — Route over ZojI-La ... ... 92

59. Eastern Range. — Tlrthaof Amaranatha ... ... 93

60. Watershed range between Sindhu and Vitasta. — Range

towards Cinab ... ... ... ... 95

SECTION V.— UPPER COURSE OF THE VITASTA.

61. Name of Vitasta. — Legendary source of Vitasta ... 96

62. Headwaters of Vitasta. — Harsapatha ; Ledari ... 98

63- Course of Vitasta in alluvial plain. — Importance of river

communication ... ... ... ... 99

64. The Gambhira. — The Visoka. — The Ramanyatavl ... 101

65. The Vitasta near S'rinagar. — The Mahasarit ... ... 103

66. The Dal lake ... ... ... ... 101

67. The Vitasta in STinagar. — The Ksiptika. — The Dugdha-

gaiiga ... ... ... ... ... 105

SECTION VI. — LOWER COURSE OF THE VITASTA.

68. The Sindhu. — Sindhu Delta. — Confluence of Vitasta and

Sindhu ... ... ... ... ... J07

69. Suyya’s regulation of the Vitasta ... ... ... 108

70. Change of confluence of Vitasta and Sindhu. — N5r canal 109

J. i. 29

226 ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2,

Para. Page.

71. Results of Suyya’s regulation ... ... ... Ill

72. Course of Vitasta towards Volur lake. — Manasa lake ... 112

73. Volur lake. — Mahapadma Naga. — Name Ullola ... 114

74. Legends of Mahapadma Naga ... ... ... 115

75. Lower affluents of Vitasta. — Suyyapura ... ... 116

SECTION VII. — SOIL AND CLIMATE OF THE VALLEY.

76. Alluvial Plateaus (Ildar). — Soil of Udars ... ... 117

77. Kasmir Climate. — Rigours of Kasmir winter ... ... 118

78. Cultivation of rice. — Irrigation. — Saffron-cultivation. —

Grapes ... ... ... ... ... 120

SECTION VIII. — ETHNOGRAPHY.

79. Old ethnography of Kasmir. — Absorption of foreign ethnic

elements. — Tribal sections ... ... ... 123

80. Races on Kasmir borders. — Khasas. — Darads. — Bhauttas... 124

CHAPTER IV.— POLITICAL TOPOGRAPHY.

SECTION I. — FRONTIERS OF ANCIENT KASMiR.

81. Territories S. E. of Kasmir. — Camba. — Vallapura. — Visa-

lata ... ... ... ... ... 126

82. Frontier territories to S. W. and W. — Rajapuri. — Lohara ;

Parnotsa. — Dvaravati ... ... ... 128

83. Urasa-Hazara. — Kisangariga Valley ; Kaimalia ... 130

84. Darad territory. — The Bhauttas ... ... ... 131

SECTION II. — ANCIENT POLITICAL DIVISIONS.

85. Kramarajya, Madavarajya ... ... ... 133

86. Administrative Districts. — Lists of Kasmir Parganas ... 134

87- Density of population in old Kasmir. — Number of villages 136

SECTION III. — THE OLD AND NEW CAPITALS.

88. SYinagara in Hiuen Tsiang’s time ... ... 137

89. Puranadhisthana, ‘ the Old Capital ’ ... ... 138

90. Asoka’s S'rinagari. — Transfer of name to new capital ... 139

91. Pravarasena’s capital. — Name of Pravarapura ... 141

92. Legend of foundation of Pravarapura. — Mahasarit ; Setu... 142

93. Old limits of Pravarapura. — Kalhana’s description of it... 144

94. Advantages of site of Srinagara. — Natural defences of

SYinagara. — City on left river-bank ... ... 145

1899.]

CONTENTS.

227

SECTION IV. — ANCIENT SITES OF SRINAGARA.

Para. Page.

95. Hill of S'arika. — Bhimasvamin Ganesa ... ... 147

96. Temple of Pravaresvara. — Position of Ranasvamin temple 149

97. Bhattarakamatha ; Diddamatha. — Skandabhavana ; Nada-

vana ... 150

98. Bridges of old STinagara. — Brliatsetu ... ... 152

99. Eastern quarters of S'rinagara. — Samudramatha. — Maksi-

kasvamin ; Setn. — Rajanavatika ... ... 153

100. Left river-bank. — Site of Royal Palace. — Gardens near

Palace ... ... ... ... ... 155

SECTION V. — THE ENVIRONS OF SRINAGARA.

101. The 4 Hill of Gopa.’ — Temple on Gopadri ... ... 157

102. Tirtlia of Jyestliesvara. — Gopagraliara. — Bhuksiravatika.

— Theda; Bhimadevi ... ... ... 159

103. Tirtha of S'uresvari; Isesvara temple. — Sadarhadvana ;

Tripuresvara. — S' rid vara ; Mahadeva ... ... 161

104. Hiranyapura. — Juskapura. — Amaresvara. — Amrtabhavana.

— Tirtha of Sodara ... ... ... 163

SECTION VI. — NORTHERN DISTRICTS OF MAD AVARAJYA .

105. District of Khaduvi. — Jayavana. — Khonamusa. — Padma-

pura. — Khaduvi; S'anara ... ... ... 165

106. District of Hola<]a. — Avantipura ... ... ... 168

107. District of Daksinapara. — Temple of Cakradhara ... 170

108. Legend of Narapura ... ... ... ... 172

109. Vij ayesvara. — Temple of S'iva Vijayesvara, — Lidar Valley 173

110. District of Varaaparsva. — S'aktamatha ; Bhimakesava ... 175

111. Tirtha of Martanda (Bavan). — Temple of Martanda ... 176

SECTION VII. — SOUTHERN DISTRICTS OF MAD AVARAJYA.

112. Anantanaga. — Kapatesvara Tirtha ; S'amangasa. — Aksa-

vala ... ... ... ... ... 178

113- Pargana of Bring. — Lokapunya. — Bhedara ; Ardhanari-

svara.- Tnsamdhya Tirtha ... ... ... 180

114. Ver Pargana (Shahabad). — Nilanaga; Vitastatra. —

Pancahasta ... ... ... ... 182

115. District of Devasarasa. — Kheri. — Godhara ; Hastisala ... 183

116. District of Karala (Ardhavana). — Jainapura. — S'urapura.

— Kapalamocana Tirtha ; Degrama ... ... 184

228

ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2,

Para. Page.

117* Kalyanapura; Drabhagrama. — Tirtlia of Bheda (Bheda-

giri) ... ... ... ... ... 186

118. Districts of S'avur and Chrath. — Ramusa; Gusika. —

Utpalapura ... ... ... ... 188

119. Districts of Nagam and Yech. — Hadigrama. — Damodara’s

Udar. — Samarasvamin ... ... ... 189

SECTION VIII. — SOUTHERN DISTRICTS OF KRAMARAJYA.

120. Parganas of Dunts, Biru, Manch^hom. — S'uskaletra. —

Bahurupa ; Kliagika ; Puskaranaga. — Malhanapura ... 191

121. Parihasapura. — Lalitaditya’s temples. — History of Pari-

hasapura ... ... ... ... 193

122. TrigramI ; Phalapura. — Jayapura ; Dvaravati ... 196

123. District of Bhangila. — Pattana. — Pampasaras. — Pratapa-

pura ... ... ... ... ... 398

124. District of Kruhin. — Kaniskapura. — Varahamula. — Temple

of Varaha. — Hnskapura ... ... ... 201

SECTION IX. — THE NORTHERN DISTRICTS OF KRAMARAJYA.

125- District of S'amala. — Tlrtha of Svayambhu. — Kambale-

svara ... ... ... ... ... 204

126. District of Uttara. — Ghosa ; Dranga ; Hayasrama ... 205

127. The S'arada Tirtlia. — The S'irahsila Castle ... ... 206

123. Parganas of Lolau and Zain^gir. — Suyyapura. — District of

Khuyasrama. — Matrgrama ... ... ... 208

129. District of Lohara. — Tulamulya. — Mayagrama ... 209

130. Tirtha of Ciramocana. — Tirthas of Bkutesa, Jyestharudra 211

131. Upper Sind Yalley. — Gaganagiri; Defile of Dhudavana ... 232

Supplementary Note AA: Mahatmyas of Ka^mirian Tirthas ...

Supplementary Note BB: The lists of Kasmir Parganas

Contents

List of Abbreviations

• •• • • • • • •

Addenda et Corrigenda

214

218

223

229

232

1899.]

LIST OF ABBREVIATIONS.

229

LIST OF ABBREVIATIONS.

Ain-i Akb.

Bates, Gazetteer.

Bernier, Travels.

Buhler, Report.

The Ain-i Akbari by Abu-1-Fazl ‘Allami, translated from the

original Persian, by Col. H. S. Jarrett. Vol. II. (Bibliotheca

Indica). Calcutta, 1891.

A Gazetteer of Kashmir and the adjacent districts of

Kishtwar, Badrawar, etc., compiled (for political and military

reference), by Charles Ellison Bates, Captain, B.s c., and

Brevet-Major. Calcutta. 1873.

Travels in the Mogal Empire A.D. 1656-1668, by Francois

Bernier, m.d. A revised and improved edition, etc., by

Archibald Constable. Westminster, Archibald Constable

and Co., 1891.

Detailed Report of a tour in search of Sanskrit MSS. made

in Kasmir, Rajputana and Central India. By G. Buhler.

Journal of the Bombay Branch of the Royal Asiatic Society,

1877.

Cole, Anc. Buildings.

Cunningham, Anc.

Geogr.

Drew, Jummoo.

Forster, Journey.

Fourth Chron.

Earucar.

Illustrations of Ancient Buildings in Kashmir. Prepared

under the authority of the Secretary of State for India. By

Henry Hardy Cole, r.e., Superintendent, Archaeological Survey

of India, N.-W. P. London, 1869.

The Ancient Geography of India, I. The Buddhist period,

including the campaigns of Alexander and the Travels of

Hwen-Thsang. By Alexander Cunningham, Major-General,

r.e. London, 1871.

The Jummoo and Kashmir Territories. A Geographical

Account. By Frederic Drew, f.r.g.s. London, 1875.

A Journey from Bengal to England, through the Northern

part of India, Kashmire, etc. By George Forster. Vols. I,

II. London, 1808.

ltajavallpataka of Prajyabhatta and S'uka. [ Fourth

Sanskrit Chronicle of Kasmir. Quoted according to the

editio princeps, Calcutta, 1835; compare below, p. 41].

The Haracaritacintamani of Rajauaka Jayadratha. Kavya-

mala, Bombay, 1897 [compare below, p. 48],

230

Hugel, Kaschmir .

India.

Jonar.

Lassen, Ind. Alt.

Lawrence, Valley.

Life.

Mah.

Moorcroft, Travels.

Nilamata.

Bajatar.

Ritter, Asien.

Samayam.

Si-yu-ki.

S'rikanthac.

S'riv.

Stein, Notes on Ou-

k'ong.

Stein, Notes on the

Pir Pant&al-

Tarikh-i Eashidt.

ANCIENT GEOGRAPHY OF KAS'MTR. [Extra No. 2,

Kaschmir und das Reich der Siek. Von Carl Freiherrn

von Hugel. Vols. I.-IV. Stuttgart, 1840.

Alberuni’s India. An account of the religion, philosophy,

literature, geography, etc., of India about A.D. 1030. An Eng¬

lish edition, by Dr. E. C. Sachau. Vols. I., II. London, 1888.

The Rajataranginl of Jonaraja. [Compare below, p. 41.

Quoted according to Calcutta edition, 1835, except for addi¬

tional verses found in Prof. Peterson’s edition, Bombay

Sanskrit Series, 1896].

Indische Alterthumskunde, von Christian Lassen. [Vols. I,

II., quoted in second edition]. Leipzig, 1866—71.

The Valley of Kashmir, by Walter R. Lawrence, i.c s.,

c.i.e., Settlement Commissioner, Kashmir and Jammu State.

London, 1895.

The Life of Hiuen-Tsiang. By the Shamans Hwui Li and

Yen-Tsung. With a Preface, etc., by Samuel Beal. London,

1888.

Mahatmya [compare below, p. 49. The several texts are

quoted from MSS.].

Travels in the Himalayan Provinces of Hindustan and the

Panjab ; in Ladakh and Kashmir ; etc., by Mr. William

Moorcroft and Mr. George Trebeck, from 1819 to 1825. Pre¬

pared for the press by Horace Hayman Wilson. Vols. I, II.

London, 1841.

The Nilamatapurana. [Compare below, p. 46. Quoted

from MSS.].

Kalhana’s Rajataranginl, or Chronicle of the Kings of

Kashmir. Edited by M. A. Stein. Sanskrit Text with criti¬

cal notes. Education Society’s Press, Bombay, 1892. [Notes

quoted will be found in my annotated translation of the Chro¬

nicle, in course of publication by Messrs. A. Constable & Co.,

London].

Die Erdkunde von Asien, von Carl Ritter. Berlin, 1833.

The Samayamatrka of Ksemendra. Kavyamala, Bombay,

1888. [Compare below, p, 44].

Si-yu-ki. Buddhist records of the Western World. Trans¬

lated from the Chinese of Hiuen Tsiang, by Samuel Beal.

Vols. I, II. London, 1884.

The S'rikanthacarita of Mahkha. Kavyamala, Bombay, 1887.

The Jaina-Rajatarangim of S'rivara. [Compare below, p.

42. Quoted, according to the editio princeps, Calcutta, 1835].

Notes on Ou-k'ong’s account of Ka9mlr. By M. A. Stein.

Proceedings of the Imperial Academy of Sciences, Vienna,

1896.

Notes on the ancient topography of the Pir Pantsal Route.

By M. A. Stein. Journal of the Asiatic Society, Bengal.

Calcutta, 1895.

The Tarikh-i Rashidi of Mirza Muhammad Haidar. An

English version, edited with commentary, etc., ‘by N. Elias j

the translation by E. Denison Ross. London, 1895.

231

1899.]

LIST OF ABBREVIATIONS.

Troyer.

Vigne, Travels.

Vikram.

Wilson, Essay.

Yule, Cathay.

Radjatarangitri. Histoire des Rois du Kachmir, traduite et

commentee parM. A. Troyer. Vols. I — III. Paris, 1840-52.

Travels in Kashmir, Ladak, etc. By G. T. Vigne, Esq.

Vols. 1, II. London, 1842.

The Vikramankadevacharita. Life of King Vikramaditya

of Kalyana, composed by his Vidyapati Bilhana. Edited, with

an Introduction, by Dr. G. Biihler. Bombay, 1875.

An Essay on the Hindu. History of Cashmir. By Horace

Hayman Wilson. Asiatic Researches, xv. Calcutta, 1825.

Cathay and the Way Thither ; being a collection of me¬

dieval notices of China, translated and edited by Colonel [Sir]

Henry Yule, c.b., r.e. With a preliminary Essay, etc. Vols.

I, II. London, 1866.

i

Note.

The transcription of Sanskrit, Persian and Arabic words in the

present Memoir conforms to the system approved by the International

Oriental Congress of 1894. In the phonetic rendering of Kasmiri

words the lines adopted by Dr. G. A. Grierson, in his analysis of Pandit

Isvara Kaula’s Kasmiri Grammar (/. A. S. B ., 1897-98) have been

closely followed.

232 ANCIENT GEOGRAPHY OF KAS'MlR. [Extra No. 2, 1899.]

CORRIGENDA.

• Page. Line.