enti Re

Ps eas eae 42h

fue. woyet Peay ieee lex

FH ETAT OR Pie eh ATA nS,

sereak Prete Sr fi too fe Riek aee ts Nee ay

Tye Malay WCE dations’

i Ph] an Pee oo th eee eae TT

eee ties ae *.

He Nbed Wh Fe i Roth RAE

| BABE R Sdn fea e04

ime tes ePri eRe We a eee eas

a 5 Rot RAVEN EAT SE BU AT UG) sO here tt 5K:

ane d winitaine ERR TNE M6 Saks GIs ats Ry Bam eRe ac Med Ra mb |

aba Lo cat BE RAMEN ANG Me matAns bos ihe Bn Revehe mies reeahehch BAC AR,

5 EASA Tt helt,

oh. an “er Tat eee Ae eA EMS Doingasa ony

pete Yo Dome Re Ron Ae ener tinea MFC NEE EM oh Cine ACCOR!

Noe a ae

shouts Aa ey ee Tet at

r eh ey le at Whcsracnate

stig he ota rigits ag ners ed) Rus te Baas asi} Yeease ae

Ce eee) ee ee % Pa] ‘ 7 (SM NeRers Ast Peemcoen.

é hoe ee PMB NEG EBS OS 48 GERD

Uae CS A Mog 8 Oh ee 12d unvio-ten ab see instesrtatay

tte Hehe edlacderte, wma Ae ey aariane A, AT AES

cr S313 :

re isa le “y — a

hem Ty Py a STM a” Ae ae pa a wee

$ Min tease anh Hit Reef RA hy mm, SSM aga: Ay Wheat

i] . res ong! ’ 4 stoee gente

wey fe deal Bid, Pedals he r =) , eos

avo : :

neath eara

te Penna

Putt Gaan nig Rongarme ne ines

¢ ‘art i Fate ng ae

PLY i wee i) A Me ein ee >

AES 3 , 58% ua oh nasqustees RET, 7 *

Mig torn one RT bey ee YY . Mintle tee -

pay teuihatateg OT CURIA A ty ties A AN OY Aa tr@etyncacita a sar boat siusans

Yea Cote $5 MeN Cee ak PDN veatedd ate

; Sen cir eadeshom Sat tn ofa ben ea

ED Myre Ne Meera tr Se Sek A Thee ty Voevier sg AAAS NND eA Degtn

Yes taas haved ta ANNE Fy Se Abe oar ed aks ee

VaR the he. ANIA Ye VE vane

a fee bern gares 4g yy Pega thy bet

Kane sitees

nN estas

PRES TOE Pe

degen time tafe

Hida feteite Va

thy Vt Hate nay

TBM tee

ass case are

a ee Yt eee

op ADAP OE

Pee

De dase 3 ¢

eaeeaetes (te Na TDMA mag cry

Ve hetal hy cored ted, vod we Am Vy ny ty tote

sve he ye ae “83

Aen erie Me pees nF ery

Rodd a nernaten Mein a

INS ie aden

Age

Soe» ad cai ern

Rater seas ace

pest

BY Ae signing oo

aed

sbie Dae ss Aree,

aaa a taep ie Be woo

waaee eae

D7 p44: Shee

dtrse wv

Fisyaaiyatege het Whe Wega s rts Pe Re RRRI ey oy

POR EM, dat dee meaeyaeay ah sr NIP es ls WTAE &

Pus

ea yee RR

Aueye etal iat

LSS Sis Sasa gues oH to LUY

BPecorrins Pilg genset ean

eo aware B ary 5 os

Pies Seiaeae ee ue pagel

Fea we

*

- ?

ie Pi te VER ney vat eat ij . see seer

ee Cee eae £ 2 ame Nas / a pean 12 JA“ pier guga vers we eSNy Bye

" i aia VPC (a aay ESE PS UO Gop EARS we awest

yi An Byelaly ie ‘i Serie Wan kt Pe

PVA Bie yeDE yao bs weil

AAP RSIS Pap GREP EE 3 eateries

i hoe be : ati eae Give dels apege

. treat

ee:

ol Ree eae ah

a flys

eit ! ‘

eoecerauqe and ea Pecan eat of

ee weer ADEE OS AALS,

5 ence eas Pai He 8 Nantecnag

Eye PU ee Bee ee leks eee iyeay

Phere

BN 5 dae SNGRNDY SS ZO Fi ee Fe eB, gO AM a fhe en Mg ae

5 ee” 5 Nr 2 FE ae 2 Se XN

77) ne Z 79) a n 4

ile NVINOSHIINS Sa 1YVYaGIT LIBRARIES SMITHSONIAN INSTITUTION NOILNL

on hs Be

Be ul 5 us e us

aa Ye om c = 3 c =

ca “A = | 0 4 fo

O ~ SS bac eel oO a ro) ea

| = a zz or 2 oy q

LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYVYUSIT LIBRA

= i za —- a

_ S ad = a

} Gera v3 w -_ ae oO

5 tif 2 2 Z ss a

> Bey, “4 > ~ oh =) >

= ‘yg 2 oa 5 = 2

é - GY = Ss a = —

: Ge ix : ; : -

— _ op) =

NOILNLILSNI NVINOSHLINS S3IYVYUGIT LIBRARIES INSTITUTION

) Zz 2) aa ur wn” = ie n

< = Press Ce eS < WY =

Sf tof $ = oA ; z= WY AS at Zz =

: t1fe ho O se O KMX & Se © ai:

YY By 2 WE 8 ae ce

| = z BOS = =

Mj = > = Mees: = >

7) Peg 7) SRY eee ” ie 4

LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS SA3IYWYEIT LIBRA

/ = = oa 2) =

3 w .o tu Bi o

: oc = oc = a =

= : Cc oe = oe pe

) ae) Ye: Oo — O = Oo

: —! za a! z 8 — a

NVINOSHLINS S3SI4¥VYUGIT LIBRARIES SMITHSONIAN INSTITUTION

ae z a z ks z

a \ nes 20 5 GUL a) 5

> \ Wc p = Uf? fu» >

ENN RS. — ot gy és =

2 \\ a F) ram, Zo , 7 “ad

a Oe ma me a ”

a) ee = wn = wn =

LIBRARIES SMITHSONIAN INSTITUTION NOILONLILSNI NVINOSHLINS S3I1uvugia_LIBRA

a (Op) Fade “es 12) z W = a

: ; = x " = < = < ;

. 4 Zz a = = z \

) IN ” Sane p wo ” Y

es D 2 2 2

tae \ 3 = S = z = \

— ea —_ —-y, = N“ ‘

x 2 a 2 B 2 @ ;

NOILALILSNI_ NVINOSHLINS LIBRARIES SMITHSONIAN INSTITUTION Pi

) zZ " 2 Ww) : 2

wo Wd LJ Yy, Mad

: za cc = cc Ya, = 0

| 5 a Y' fu 7 A <

7 = ae = <x (fora

a om = oc ‘ng fee Lr &

| oO eo 5 me 5 ie

| Zz af z 3 4 Zz = |

LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31YVYUsEIT LIBRA

| é 1 = a z JE |

| 5 A ff P e “0 = x

} ee, ey > pe q = >

- LYY at piel = —

) 9 “iff m 2 m 2 2

— — Ww =

NOILNLILSNI NVINOSHLINS S3IYVYEIT LIBRARIES

= w z ” = 7)

ne = Shes = < =

z = ee = z =

O Ser) a a8 ro) Se

) (Spy WY , (7p) w” map) ios

) De oO” % pa O TE O

= = Z <= a z

= > = Oh = >

a) bisa) Nae De cae 7) za

ate o7 Sipe *WNY OC YY CF +

> —— = ~ >" =

: ” PAS Pas ap)

SJIYVYHUGIT LIBRARIES

7p) = Zp)

ul Ae Ww

= o

<=

= ow

rs) ay

z= ee)

MITHSONIAN INSTITUTION NOILNLILSNI

S3IuYVUGIT LIBRAR!

INSTITUTION

Se sara Vedi? LIBRARIES

yy

s

4

NVINOSHLIWS

SMITHSONIAN

YW

*

SMITHSONIAN INSTITUTION NOILNILILSNI

SMITHSONIAN

NOILNLILSNI

SMITE

INSTITUTION NOILNLILSNI

LIBRARIES

NVINOSHLINS SSIYVYE!IT LIBRARIES SMITHSON]

NVINOSHLINS S3IYVUal]

INSTITUTION

SMITHSONIAN

AN

Saluvydl

INSTITUTION NOILNILILSNI

NVINOSHLIWS

LIBRARIES

2

>

2

_ NVINOSHLI}

Fae sae

ey :

oo SS

< OS

O \S

2

rad

Oo

te

=)

eae,

me

Ww

=

NVINOSHLI!

=

ke

rs

Oo

Y

ae

ts

=

Ww

SMITHSON]

z

oO

io)

Be

a

=

Y

2 a 2 2 Z ‘

i fy’ 4 ro af i | as fp wey

ee = ac =. a tty JY

“yy = = : = « “iy

a3) FZ =

NVINOSHLINS SSIY¥VYdEIT LIBRARIES INSTITUTION NOILOALILSNI_NVINOSHLI

_~ = Z z z z a

= 0 = iD E )

» WK = a = i> = =

NSE = - = 5 :

SS 7) wn

Pos z a Zz A 2 o

“SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3INvUGIT_LIBRARIES

Zz. “* ” = ep) =

a S z s Zhi | & z

Zz \. 5 yr = Up Zz =

2 OS AE 3 a @: A j

“A SYS -. S Ae =

S & ~~ ¥S YS O Se oO he f. 8 : O

E AS =, FE: 2 G fe E Z

e [ [ae s 5 Ae aie

-NVINOSHLINS S3iuvVud!iT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI_ NVINOSHL

pa ” uJ wo ud ep)

= 4 a = a a

= = Zs oS = a

re er fe é S oc &

= 5 co - = ao re)

me Zz dj z = Zz

INSTITUTION NOILNLILSNI NVINOSHLINS S31Y¥VYEl) LIBRARIES

a ee ic = ire Fa

= S = S - 2

be 5 a ES D a

> ris > rad > en

e - = es fa =

= ” ia os m oes

D z 5 ee - =

SAIYVUGIT LIBRARIES SMITHSONIAN INSTITUTIG® NOILNLILSNI NVINC™.

2) z ce 7) oa “ %) =

z ok Un ie = ~\ = es

_ ; Ao ty ae = = Se = 4

= : hh fy Oo RAN = 5 = 5

8 Oye = WK S 2 3 | ~

Zz “iy E OS ae : = =

: ih ie : [Sik

J Si Ww iat

) POMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYVYSIT LIBRARI Ea te

a ; é = AEESON; » iW Anevan 2 = LOX . uJ ans¥M4 >» a ELIS ud

4 Pi, 4 oe \ Mod y) Z, x, Z Re Be DE f

een fy A /° 4\ PU aN z. fod LTT Dy a Nie ex m~ wy es

ANNALS OF THE ANNALE VAN DIE

SOUTH AFRICAN MUSEUM SUID-AFRIKAANSE MUSEUM

VOLUME 94 BAND 94

Er — rs

foe

an

JAN 03 1986

are Cygne 2 erage aig his OG Tea oe

1 wit a a ‘54

- ye i i - r

; Ye ee | ee ‘ ‘an

e n 1 . |

f. 5he i 4

peed F i ; Road) eye : ae ; , — 0 ets a

af >. ! : 1 nt mn

- oetth

i tpl Va mot 1h 0 bd

i ; y t

" a iy

=) iy 5 | m

5

—* : ¥ ta { ;

= y

) : val

” 0 ge

é Fs

j : Rive

Ly u

2) ym ME

By ‘

wo a p

a! b "

y

; y

x

i

1

oY

5 1s

“i ;

if

sa

t

in

i i 4 \ =) =

_ i ed Fl

7 1 7 Ss

1

. " '

: ” 1

We k

7 i 7 u

Beh ie i F

ius ,

ity

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

VOLUME 94 BAND

Hee PRUSPEES OF THE DIE SERUSEEBES, VAN DIE

SOUTH AFRICAN MUSEUM SUID-AFRIKAANSE MUSEUM

CAPE TOWN KAAPSTAD

1983-1984

*

P| |

S iy

s

\ e

ay

a }

5

eae

=

—

i

t

f

2

1 ‘ ,

ne a

ty! ee

r bcs)

LISi- OF CONTENTS

Davison, P.

Lobedu material culture: a comparative study of the 1930s and the 1970s. (Published

AUT al SA) ete We ne eames a Sealed Pale a, Baler 10 cytes Jalchig © ag RPRE A MRE eRe ets RAIA oc

ENGLAND, K. W. & Rosson, E. A.

A new sea anemone from South Africa (Anthozoa, Ptychodactiaria). (Published

| SSOP Foes Note Hig re ge hear nen rity ROL UN at a ce MEM Sa Mates a 20 BNE

HeemstraA, P. C. & KANNEMEYER, S. X.

The families Trachipteridae and Radiicephalidae (Pisces, Lampriformes) and a new

Species on A trom! southyAtncas (Rublisheddiune i942) sees eee eee

KANNEMEYER, S. X. see HEEMSTRA, P. C.

Prins, A. J.

A new ant genus from southern Africa (Hymenoptera, Formicidae). (Published

ING@wiembe ral SS) r ets cre Pes cc cccletn sme ie Src PRR teeter nga gh

Prins, A. J.

Morphological and biological notes on some South African arthropods associated

with decaying organic matter. Part 3. The families Dermestidae, Cantharidae,

Melyridae, Tenebrionidae, and Scarabaeidae (Coleoptera). (Published Sep-

LMT Cote SAR) tka toad a a teens eee anne tay gah yack a Ue ARR 2001 heh Mg eo ea

Rosson, E. A. see ENGLAND, K. W.

Volume 94 is complete in 5 parts.

Page

4]

305

118)

i ‘

: i

j ae Nie

,

i

4

:

i

nf ty

om

er

’

A

7 Ve i

, Sone

het 5 ‘

eo ae

‘

ae g jst Bag

’ } ni

i

\

= 52

7D patie 0 a ue @ ag

: et

NOVEMBER 1983 ISSN 0303-2515

"OF THE SOUTH AFRICAN

/ MUSEUM

CAPE TOWN

INSTRUCTIONS TO AUTHORS

1. MATERIAL should be original and not published elsewhere, in whole or in part.

2. LAYOUT should be as follows:

(a) Centred masthead to consist of

Title: informative but concise, without abbreviations and not including the names of new genera or species

Author’s(s’) name(s)

Address(es) of author(s) (institution where work was carried out)

Number of illustrations (figures, enumerated maps and tables, in this order)

(b) Abstract of not more than 200 words, intelligible to the reader without reference to the text

(c) Table of contents giving hierarchy of headings and subheadings

(d) Introduction

(e) Subject-matter of the paper, divided into sections to correspond with those given in table of contents

(f) Summary, if paper is lengthy

(g) Acknowledgements

(h) References

(i) Abbreviations, where these are numerous

3. MANUSCRIPT, to be submitted in triplicate, should be typewritten and neat, double spaced

with 2,5 cm margins all round. First lines of paragraphs should be indented. Tables and a list of

legends for illustrations should be typed separately, their positions indicated in the text. All

pages should be numbered consecutively.

Major headings of the paper are centred capitals; first subheadings are shouldered small

capitals; second subheadings are shouldered italics; third subheadings are indented, shouldered

italics. Further subdivisions should be avoided, as also enumeration (never roman numerals)

of headings and abbreviations.

Footnotes should be avoided unless they are short and essential.

Only generic and specific names should be underlined to indicate italics; all other marking

up should be left to editor and publisher.

4. ILLUSTRATIONS should be reducible to a size not exceeding 12 x 18 cm (19 cm including

legend); the reduction or enlargement required should be indicated; originals larger than

35 x 47 cm should not be submitted; photographs should be rectangular in shape and final

size. A metric scale should appear with all illustrations, otherwise magnification or reduction

should be given in the legend; if the latter, then the final reduction or enlargement should be

taken into consideration.

All illustrations, whether line drawings or photographs, should be termed figures (plates

are not printed; half-tones will appear in their proper place in the text) and numbered in a

single series. Items of composite figures should be designated by capital letters; lettering of

figures is not set in type and should be in lower-case letters.

The number of the figure should be lightly marked in pencil on the back of each illustration.

5. REFERENCES cited in text and synonymies should all be included in the list at the end of

the paper, using the Harvard System (ibid., idem, loc. cit., op. cit. are not acceptable):

(a) Author’s namie and year of publication given in text, e.g.:

‘Smith (1969) describes .. .’

‘Smith (1969: 36, fig. 16) describes . . .’

‘As described (Smith 1969a, 19695; Jones 1971)’

‘As described (Haughton & Broom 1927)...’

‘As described (Haughton et al. 1927)...’

Note: no comma separating name and year

Dagination indicated by colon, not p.

names of joint authors connected by ampersand

et al. in text for more than two joint authors, but names of all authors given in list of references.

(b) Full references at the end of the paper, arranged alphabetically by names, chronologically

within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year, e.g. Smith (1969a, 19695) and not Smith (1969, 1969a).

For books give title in italics, edition, volume number, place of publication, publisher.

For journal article give title of article, title of journal in italics (abbreviated according to the World list o,

scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses, volume number, part

number (only if independently paged) in parentheses, pagination (first and last pages of article).

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FISCHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris 88: 100-140.

FIsCHER, P.-H., DuvAL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des littorines. Archs

Zool. exp. gén. 74: 627-634.

Konn, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee region of Ceylon.

Ann. Mag. nat. Hist. (13) 2: 309-320. :

Konn, A. J. 19606. Spawning behaviour, egg masses and larval development in Conus from the Indian Ocean.

Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L. Zoologische

und anthropologische Ergebnisse einer Forschungsreise im westlichen und zentralen Siid-Afrika 4: 269-270.

Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16: 269-270. ;

(continued inside back cover)

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 94 Band

November 1983 November

Part 1 Deel

Dv HR

A NEW ANT GENUS

FROM SOUTHERN AFRICA

(HYMENOPTERA, FORMICIDAE)

By

A. J. PRINS

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town 8000

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na gelang van die

beskikbaarheid van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad 8000

OUT OF PRINT/UIT DRUK

ily DO SO), SG, A5 Oo t at), SCS, 5, Oy,

AC, bast), 1), &, C2, ), 10G-3)),

GDS. ee ICES), LUO), WL SHES), BAG), 23, 35), 250)

EDITOR/REDAKTRISE

Ione Rudner

Copyright enquiries to the South African Museum

Kopieregnavrae aan die Suid-Afrikaanse Museum

ISBN 0 86813 049 4

Printed in South Africa by In Suid-Afrika gedruk deur

PhenRustica senessa setyesleude. Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

A NEW ANT GENUS FROM SOUTHERN AFRICA

(HYMENOPTERA, FORMICIDAE)

By

A. J. PRINS

South African Museum, Cape Town

(With 4 figures)

[MS accepted 22 September 1983 |

ABSTRACT

A new genus and its two new species, Agraulomyrmex meridionalis and A. wilsoni, are

described and a key to the workers of the genera of the subfamily Formicinae present in this

subregion is provided.

CONTENTS

PAGE

its SROGHUCTION acs ope eens Rt coe Sy angen on Let ar aaa CYR eet Ete 1

PNEAMLONMDY UITLCX: © CT MO Vine oss oh te, a cncteyeh 9.2, Artois aeegra aun eolepeoee wee 2

VNCAULONMYTINEX METIGIONAUS SP. MOV. ....4-.-4.4+2-4. 08044 seos ne. 2)

PACA ONM TINOGWiSONUSDsMOVas 4.6542 2060s Sees ee 5

Key to the workers of the genera of the subfamily Formicinae in

SOMUMCHMP NU Caden Nam ga wi ate x ena cies ar acd Seo remeunayaace geese 8

PAGKMOMMECSEMENMES) ofr yn kn Gone A Sens 4 ae ayes ne oh nies ean oe 10

NCHS UCINC COMPRIS ACS TCE cate muly dee a lm so elas eee ee Mn Lae sea 11

/: SIDE VIEINOIDS atin os AER aa ee ee een ere OEL ROR Pig Cor Ti ot Iti

INTRODUCTION

In 1969 surveys of ants present in citrus orchards were conducted throughout

South Africa. During the cooler winter months and early spring one sample of

small, brownish hypogaeic ants was collected among some citrus trees in the veld

near Velddrif, a small fishing village about 200 km north-west of Cape Town. As

these ants looked very much like Technomyrmex albipes (F. Smith), which shared

the same habitat, they were noted as such in the collection records. This same

species of ant was also at the time observed in a small citrus orchard near

Citrusdal, about 60 km north-east of Velddrif. In this case they were tending the

aphid Hyadaphis coriandri (Das) on warm days, about 4 to 7 cm above ground

level on the low-growing weeds. They were, however, never observed during the

Summer and autumn months.

Three years later a sample received from the Ministry of Agriculture of

Zimbabwe (formerly Rhodesia) proved to be a different species of this new

genus, which includes the only southern African formicine ants of which the

workers have ten-segmented antennae.

Ann. S. Afr. Mus. 94 (1), 1983: 1-11, 4 figs.

D) ANNALS OF THE SOUTH AFRICAN MUSEUM

Genus Agraulomyrmex gen. nov.

Type species Agraulomyrmex meridionalis sp. nov.

by original designation herein

Diagnosis

Small ants with ten-segmented antennae, flagella incrassate towards apices

without distinct club; first and last segments of flagella longer than wide, the

remaining segments as long as or wider than long. Mandibles triangular with four

to six teeth; frontal carinae short, antennal fossae situated close to posterior

border of clypeus. Eyes large, placed in front of middle of sides of head; ocelli

absent. Maxillary palpi five- to six-segmented, labial palp three- to four-

segmented. Labrum bilobed, widely and shallowly emarginate in middle. Frontal

area clearly indicated.

Alitrunk with promesonotal and mesonotal-propodeal sutures well devel-

- oped, mesometanotal suture absent; propodeum rounded from front to back,

unarmed. Petiole with very small or rudimentary scale overhung by base of

abdomen. Legs moderately long.

Although the ten-segmented antennae point to a myrmelachistine relation-

ship, these ants share characters with the plagiolepidines, viz. the absence of a

distinct antennal club, the structure and form of the alitrunk (the absence of a

distinct metanotum as in certain members such as the genus Acropyga), as well as

the structure of the asepalous proventriculus (Fig. 4A—D). The new genus is

therefore included in the tribe Plagiolepidini.

Workers of these ants bear a close resemblance to those of the genus

Acropyga (Prins 1982); however, they are much smaller and darker in colour, and

the eyes are much larger. In members of Acropyga the eyes are very small,

consisting of only a few facets. The petiolar scale is also much more reduced than

in Acropyga.

Derivation of name

The name Agraulomyrmex (masculine gender) is derived from the Greek

word agraulos, which refers to the mode of living (in the fields).

Agraulomyrmex meridionalis sp. nov.

Description

Worker (Fig. 1A—C)

Specimens collected at Velddrif from which the type material was selected:

TL 2,10-2,20 mm; L0,90-1,0 mm; HL 0,48-0,52 mm; ED 0,26 mm; CL

0,08—0,10 mm; FL 0,40—-0,42 mm; SL 0,32-0,36 mm; WL 0,48-0,50 mm; MFL

0,28-0,30 mm; HFL 0,32-0,36 mm; PL 0,10 mm; CI 76,9-79,2; FI 61,9-65,0; SI

84,2-90,0; CLI 320,0—355,6; TI 58,0-58,3; PI 100,0; HFI 66,7—72,0.

Dark brown, tarsi and mandibles paler. Smooth and shiny all over, especially

declivity of propodeum and petiole. Finely sculptured with piliferous punctures.

A NEW ANT GENUS

Fig. 1. Agraulomyrmex meridionalis sp. nov.

A. Dorsal view of worker. B. Left lateral view of worker. C. Head of worker seen

from the front.

4 ANNALS OF THE SOUTH AFRICAN MUSEUM

Pubescence fairly long, decumbent and yellowish, dense all over except on petiole

and declivity. Pilosity yellowish, long, evident only on mandibles, clypeus,

petiole, and apical margins of abdominal segments.

Head quadrate in dorsal view, about one-fifth to one-sixth longer than wide,

and about one-third wider than pronotum, sides feebly convex, hind margin

almost straight. Clypeus convex in middle, not carinate, its anterior margin

arcuate as in the genus Acropyga. Scapes about two-thirds as long as head

(excluding mandibles) and reaching hind margin; flagellum about two-fifths

longer than scape, first segment as long as the second and third taken together,

apical one slightly longer than three preceding segments taken together; rest as

wide as, or wider than, long. Eyes large, oval, occupying nearly one-fourth of the

length of head (dorsal view—mandibles excluded), and situated in front of middle

of sides. Mandibles shiny, with few large, oval punctures, each bearing erect or

semi-erect seta, and with five to six alternate large and small teeth. Maxillary palp

- six-segmented, labial palp four-segmented.

Alitrunk similar to that of the genus Acropyga, slightly more than three-

eighths longer than wide over pronotum; seen in profile propodeum is on lower

level than promesonotum, declivity almost flat and longer than dorsum of

propodeum. Scale of petiole rudimentary, inclined forward and about as long as

wide and slightly higher than long; seen from behind-almost as wide above as

below, dorsum feebly convex.

Legs moderately long. Abdomen oval in dorsal view. Acidopore on conical

projection of hypopygium, orifice surrounded by fringe of hairs.

Specimens which do not form part of the type material and which were

collected at Citrusdal have the following representative measurements:

TL 2,0-2,20 mm; L 0,94-1,0 mm; HL 0,44—0,48 mm; ED 0,22-0,26 mm; CL

0,09-0,10 mm; FL 0,34-0,39 mm; SL 0,32-0,34 mm; WL 0,46-0,52 mm; MFL

0,26-0,28 mm; HFL 0,32-0,34 mm; PL 0,10-0,11 mm; CI 77,3-83,3; FI 64,1—

68,4; SI 85,0-94,1; CLI 320,0-355,6; TI 53,8-56,5; PI 80,0-90,0; HFI 65,3-69,61.

In most of the specimens examined the head is somewhat shorter than the

alitrunk and the scale of the petiole slightly longer than wide, otherwise similar in

all respects to the Velddrif specimens.

Holotype

1 %, Velddrif, Cape Province (32°47'S 18°10’E), collected by A. J. Prins

7 July 1959, South African Museum specimen.

Paratypes

4 9%, as above, South African Museum specimens.

Other material

Not included in the type material: 4 % 8, Citrusdal, Cape Province (32°37'S

18°58’E), collected by A. J. Prins 17 March 1960, South African Museum

Specimens.

Nn

A NBW ANT GENUS

Derivation of name

The name meridionalis is derived from the Latin word meridional-, which

refers to the southerly distribution of this species.

Agraulomyrmex wilsoni sp. nov.

Description

Worker (Fig. 2A-C)

TL 1.5-1,90 mm; L 0,66-0,7 mm; HL 0,34 mm; ED 0,20-0,21 mm: CL

0,05—0,06 mm; FL 0,28-0,29 mm; SL 0,20-0,21 mm; WL 0,34 mm; MFL 0,14

mm; HFL 0,18 mm; PL 0,07—0,08 mm; CI 88,2—94,1; FI 68,9-75,0; SI 65,6—73,3;

CLI 336,7—440,0; TI 57,8; PI 87,5-114,3; HFI 52,9.

Brown; antennae, mandibles, legs and mesonotum paler in colour, almost

brownish white in some individuals; eyes black. Fairly shiny all over. Finely

sculptured with piliferous punctures as in A. meridionalis, giving integument

an almost fine reticulate or reticulate-punctate appearance; in some specimens

head somewhat duller than in A. meridionalis. Pubescence and pilosity as in

A. meridionalis.

Head quadrate in dorsal view, about one-sixth (or slightly more) longer than

wide and three-eighths wider than pronotum, sides and hind margin almost

straight. Clypeus and mandibles as in A. meridionalis, latter with four sharp

teeth. Scapes slightly more than three-quarters as long as head (dorsal view—

mandibles excluded) and falling short of hind margin by about one-fourth of their

length; flagellum about one-half longer than scape; the first segment as long as

second to fourth taken together; apical one about as long as four preceding ones

taken together; remaining segments as wide as or wider than long. Eyes about

one-fourth the length of the head (dorsal view—excluding mandibles) and

situated in front of middle of sides. Maxillary palp five-segmented, labial palp

three-segmented.

Alitrunk as in A. meridionalis, about three-eighths longer than wide over

pronotum, sutures clearly indicated. Scale of petiole much more reduced, about

as long as wide and about as high as long or slightly higher; seen from behind

almost as in A. meridionalis. Otherwise as in latter species.

Female (Fig. 3A—-B)

Somewhat damaged, both flagella broken off.

TL 3,60 mm; L 1,48 mm; HL 0,48 mm; ED 0,32 mm; LO 0,26 mm; OD 0,13

mm; CL 0,09 mm; FL 0,39 mm; SL 0,34 mm; WL 0,94 mm; MFL 0,26 mm; HFL

Peseern-see 0512 mm; Cl 9157:FL82.1;,S177,3; CLI. 377,8; TI 53,2; Pl 133.3;

HFI 43,6.

Brown, legs, antennae and mandibles paler in colour, eyes black, moderately

shiny. Finely and superficially punctate or reticulate-punctate, especially on legs

and antennae. Pubescence and pilosity as in worker, except for some pilose hairs

on scutellum.

ANNALS OF THE SOUTH AFRICAN MUSEUM

B. Left lateral view of worker. C. Head of worker seen

Fig. 2. Agraulomyrmex wilsoni sp. nov.

A. Dorsal view of worker.

from the front.

A NEW ANT GENUS

mM

Fig. 3. Agraulomyrmex wilsoni sp. nov.

A. Dorsal view of female. B. Left lateral view of female.

8 ANNALS OF THE SOUTH AFRICAN MUSEUM

Head quadrate in dorsal view, nearly one-sixth longer than wide, and slightly

narrower than truncus; sides and hind margin almost straight. Frontal carinae,

fontal area, mandibles, and clypeus as in worker. Eyes large, occupying about

one-fourth the length of the head (dorsal view—mandibles excluded). Scapes

about three-fourths as long as head (dorsal view—mandibles excluded) and

falling short of hind margin by about the length of the second segment of

flagellum.

Alitrunk slightly less than twice as long as wide, parapsidal furrows

indicated, scutellum one-fourth the length of the scutum; seen from above

alitrunk appears oval in outline; in profile propodeum slopes down fairly steeply

towards junction of petiole. Latter with almost obsolete scale which is slightly

wider than long and about one-fourth higher than wide. Abdomen in specimen

examined nearly twice as long as head and alitrunk taken together, oval at base.

Legs moderately long, femora somewhat swollen.

These ants are much smaller than A. meridionalis, much paler in colour, and

more repletes were present.

Holotype

1 2, Gwebi, Zimbabwe (17°40’S 30°40’E), collected by K. J. Wilson

December 1972, South African Museum specimen.

Paratypes

12,7 2%, as above, South African Museum specimens.

Derivation of name

This species has been named after Dr K. J. Wilson of the Ministry of

Agriculture, Harare, Zimbabwe (formerly Salisbury, Rhodesia), who collected

the sample.

KEY TO THE WORKERS OF THE GENERA OF THE SUBFAMILY FORMICINAE

IN SOUTHERN AFRICA

Bolton’s (1973) key should be altered as follows to accommodate the genera

of the southern African subregion (south of the Zambesi River). According to his

key the acidopore is not borne on a conical projection of the hypopygium in the

genus Camponotus, and the orifice is usually not surrounded by a fringe of hairs.

However, in some of our southern African species the acidopore is raised to a

certain extent above the surface of the hypopygium on a somewhat conical area,

and is surrounded by hairs on at least the anterior edge of the orifice. This is

particularly the case in some species of the subgenera, Myrmespera, Mayria (see

Prins 1973), Orthonotomyrmex, Myrmopsamma, Myrmopiromis, and even in

some members of Tanaemyrmex. It is definitely on a conical projection of the

hypopygium in this new genus Agraulomyrmex.

A NEW ANT GENUS 1)

Fig. 4. Lateral view of the proventriculus of the workers of three species of plagiolepidine ants,

compared with that of Agraulomyrmex meridionalis sp. nov.

A. Anoplolepis steingroeveri Forel. B. Acantholepis capensis Mayr.

C. Agraulomyrmex meridionalis sp. nov. D. Acropyga arnoldi Santschi.

=

=)

| s | es fo |e

a |

10.

ANNALS OF THE SOUTH AFRICAN MUSEUM

Antennae9- or lO-sesmenteds 0.5 ieee cian areas qd hc ee ea Cae ee 2

Antennae li- ori2-sesmented tc... aos ser ae | earn eee ne ae 3

Antennae 9-seomented a Arh) vee ate to cana reeenena e Aphomomyrmex

Antennae d 0-sesmented’. <2 ee matic tec seeks ae oe eee eee Agraulomyrmex

Antennac li-seomenteds ova. eat aiapyeny ein 2a cen teal ae SeenON ci ene aes Een 4

Antennae 12-sepsmented, 5... 925. as okt ae sos clos oO ae eae cas ace 7

Maxillary palp 6-segmented, labial palp 4-segmented. Eyes distinct and fairly large;

ocellimay be present 25 2 essa tenis on nce Ue eae Coe ere alg ee rt eee 5

Maxillary palp 4- or 5-segmented, labial palp 3-segmented. Eyes minute ....... Acropyga

Propodeum bidentate or bituberculate; petiole usually bispinose or bidentate above,

occasionally only with upper border strongly emarginate.................. Acantholepis

Propodeum unarmed; petiole neither armed nor deeply emarginate above ........... 6

Metanotum small, much less than half as long as the mesonotum, the mesometanotal

suture feeble or obsolete, at least in the minor workers. Propodeum slightly to strongly

convexcandolteniicher thanthemesometanotimn en ae eee ere Anoplolepis

Metanotum at least half as long as the mesonotum. Metanotum usually fairly

prominent and separated from the mesonotum and propodeum by fairly distinct

sutures. Propodeum about as high as or sometimes higher than metanotum.... Plagiolepis

Petiole reduced to an elongate, low node, allowing the gaster to be reflexed over the

alitrunk. Mandibles elongate triangular, broad, apical tooth long. Maxillary palp

S-sesmentedslabialipalp4-sesmented ane ms sari eine eee Oecophylla

Petiole a node or scale, never as above; mandibles not as above. Maxillary palp 6-

segmented, labial palp 4-segmented, or the palp formula rarely reduced to 3,4 or 3,3,

DUEMEVER SA ek be Sate ey dias Ne MS eracis e AON ne es OND cycle ete aces 8

Antennal insertions very close to, or contiguous with, the posterior clypeal margin.

Acidopore borne on a conical projection of the hypopygium, forming a nozzle, the

onicesurrounded|by. a iringe Of hairs eo aoe a a a ee 9

Antennal insertions some distance (usually greater than basal width of scape) behind

the posterior clypeal margin. Acidopore either not borne on a conical projection, the

orifice not surrounded by a fringe of hairs, or orifice borne on a conical projection or

raised above level of hypopygium on a somewhat conical area, without a fringe of hairs

or with at least some hairs on ventral edge; or the acidopore hidden by a projection of

the pySiauMmies 5 oe eel te le Was Be ceie Sines Rieie: alte 52a Ok pens a 10

Dorsum of alitrunk with very coarse setae arranged in distinct pairs. Eyes at or in

frontokmuidlengthiokthe head) 42a secs sae ae eee Paratrechina

Dorsum of alitrunk with fine setae, not definitely paired. Eyes placed behind midlength

Ol the ead aon. SPs es Bie ees ae ae heer ee ee Prenolepis

Anterodorsal pronotal angles usually projecting as spines or teeth, at least strongly

marginate. Propodeum usually bispinose or bidentate; petiole with sharp angles, spines

Orteethtaboves Monomonpliicr. (Mah sen cia ae eae eee Polyrhachis

Anterodorsal pronotal angles rounded, unarmed. Propodeum unarmed although may

be truncate posteriorly. Petiole a node or scale, never with teeth or spines.

POlyimOrphics.\s Peak tees. BAe sata eRe ae ares Bi eee ee Se cis, <)> cers ee Camponotus

ACKNOWLEDGEMENTS

I should like to thank Dr A. J. Hesse, formerly of the South African

Museum, for his comments as well as Dr V. B. Whitehead and Miss

M. Macpherson, of the same Museum, for their advice. I should also like to thank

Prof. H. J. R. Durr, of the University of Stellenbosch, for identifying the aphid

Hyadaphis coriandri (Das).

a

A NEW ANT GENUS 11

REFERENCES

Botton, B. 1973. The ant genera of west Africa: A synonymic synopsis with keys

(Hymenoptera: Formicidae). Bull. Br. Mus. nat. Hist. 27: 319-368.

Prins, A. J. 1973. African Formicidae (Hymenoptera) in the South African Museum.

Description of four new species and notes on Tetramorium Mayr. Ann. S. Afr. Mus. 62:

1-40.

Prins, A. J. 1982. Review of Anoplolepis with reference to male genitalia, and notes on

Acropyga (Hymenoptera, Formicidae). Ann. S. Afr. Mus. 89: 215-246.

ABBREVIATIONS

CI cephalic index (head width/HL x 100)

CL clypeal length

CLI clypeal index (clypeal width/CL x 100)

ED distance between compound eyes

FI frontal index

FL frontal length

HFI hind femur index (HFL/WL x 100)

HFL hind femur length

HL head length

Ie length of anterior margin of clypeus to base of abdomen

MFL middle femur length

PI __ petiolar index (petiolar width/PL x 100)

PL _petiolar length

SI scape index (SL/head width x 100)

SL scape length

TI thoracic index (thoracic width/WL x 100)

TL total length of body

WL length of alitrunk

Ni - : = 7 7

z —

;

S ;

F \ a ME 2

. ‘ P = -

Se

I,

——s G -

‘f _ ,

ft z

= os x 4 y =

f

i a

— iy

( — 4 ae =

> -

m , a

ti,

a 1

zt i 4

2 ;

iS - c : RES 1 =

- : be : *

2 = : =~

~ - - 3

6. SYSTEMATIC papers must conform to the Jnternational code of zoological nomenclature

(particularly Articles 22 and 51).

Names of new taxa, combinations, synonyms, etc., when used for the first time, must be

followed by the appropriate Latin (not English) abbreviation, e.g. gen. nov., sp. nov., comb.

nov., syn. nov., etc.

An author’s name when cited must follow the name of the taxon without intervening

punctuation and not be abbreviated; if the year is added, a comma must separate author’s

name and year. The author’s name (and date, if cited) must be placed in parentheses if a

species or subspecies is transferred from its original genus. The name of a subsequent user of

a scientific name must be separated from the scientific name by a colon.

Synonymy arrangement should be according to chronology of names, i.e. all published

scientific names by which the species previously has been designated are listed in chronological

order, with all references to that name following in chronological order, e.g.:

Family Nuculanidae

Nuculana (Lembulus) bicuspidata (Gould, 1845)

Figs 14-15A

Nucula (Leda) bicuspidata Gould, 1845: 37.

Leda plicifera A. Adams, 1856: 50.

Laeda bicuspidata Hanley, 1859: 118, pl. 228 (fig. 73). Sowerby, 1871: pl. 2 (fig. 8a—b).

Nucula largillierti Philippi, 1861: 87.

Leda bicuspidata: Nicklés, 1950: 163, fig. 301; 1955: 110. Barnard, 1964: 234, figs 8-9.

Note punctuation in the above example:

comma separates author’s name and year

semicolon separates more than one reference by the same author

full stop separates references by different authors

figures of plates are enclosed in parentheses to distinguish them from text-figures

dash, not comma, separates consecutive numbers

Synonymy arrangement according to chronology of bibliographic references, whereby

the year is placed in front of each entry, and the synonym repeated in full for each entry, is

not acceptable.

In describing new species, One specimen must be designated as the holotype; other speci-

mens mentioned in the original description are to be designated paratypes; additional material

not regarded as paratypes should be listed separately. The complete data (registration number,

depository, description of specimen, locality, collector, date) of the holotype and paratypes

must be recorded, e.g.:

Holotype

SAM-—A13535 in the South African Museum, Cape Town. Adult female from mid-tide region, King’s Beach

Port Elizabeth (33°51’S 25°39’E), collected by A. Smith, 15 January 1973.

Note standard form of writing South African Museum registration numbers and date.

7. SPECIAL HOUSE RULES

Capital initial letters

(a) The Figures, Maps and Tables of the paper when referred to in the text _

e.g. *... the Figure depicting C. namacolus ...’; *. . . in C. namacolus (Fig. 10)...’

(b) The prefixes of prefixed surnames in all languages, when used in the text, if not preceded

by initials or full names

e.g. Du Toit but A.L.du Toit; Von Huene but F. von Huene

(c) Scientific names, but not their vernacular derivatives

e.g. Therocephalia, but therocephalian

Punctuation should be loose, omitting all not strictly necessary

Reference to the author should be expressed in the third person

Roman numerals should be converted to arabic, except when forming part of the title of a

book or article, such as

‘Revision of the Crustacea. Part VIII. The Amphipoda.’

Specific name must not stand alone, but be preceded by the generic name or its abbreviation

to initial capital letter, provided the same generic name is used consecutively.

Name of new genus or species is not to be included in the title: it should be included in the

abstract, counter to Recommendation 23 of the Code, to meet the requirements of

Biological Abstracts.

A. J. PRINS

A NEW ANT GENUS

FROM SOUTHERN AFRICA

(HYMENOPTERA, FORMICIDAE)

QH

1

eerx “Tt 2 JUNE 1984

| ISSN 0303-2515

NH

jen 2 BRE ARIES sy en

tie

ANNALS

"OF THE. SOUTH AFRICAN -

MUSEUM

CAPE TOWN

INSTRUCTIONS TO AUTHORS

|. MATERIAL should be original and not published elsewhere, in whole or in part.

2. LAYOUT should be as follows:

(a) Centred masthead to consist of

Title: informative but concise, without abbreviations and not including the names of new genera or species

Author’s(s’) name(s)

Address(es) of author(s) (institution where work was carried out)

Number of illustrations (figures, enumerated maps and tables, in this order)

(b) Abstract of not more than 200 words, intelligible to the reader without reference to the text

(c) Table of contents giving hierarchy of headings and subheadings

(d) Introduction

(e) Subject-matter of the paper, divided into sections to correspond with those given in table of contents

(f) Summary, if paper is lengthy

(g) Acknowledgements

(h) References

(i) Abbreviations, where these are numerous

3. MANUSCRIPT, to be submitted in triplicate, should be typewritten and neat, double spaced

with 2,5 cm margins all round. First lines of paragraphs should be indented. Tables and a list of

legends for illustrations should be typed separately, their positions indicated in the text. All

pages should be numbered consecutively. ;

Major headings of the paper are centred capitals; first subheadings are shouldered small

capitals; second subheadings are shouldered italics; third subheadings are indented, shouldered

italics. Further subdivisions should be avoided, as also enumeration (never roman numerals)

of headings and abbreviations.

Footnotes should be avoided unless they are short and essential.

Only generic and specific names should be underlined to indicate italics; all other marking

up should be left to editor and publisher.

4. ILLUSTRATIONS should be reducible to a size not exceeding 12 « 18 cm (19 cm including

legend); the reduction or enlargement required should be indicated; originals larger than

35 x 47 cm should not be submitted; photographs should be rectangular in shape and final

size. A metric scale should appear with all illustrations, otherwise magnification or reduction

should be given in the legend; if the latter, then the final reduction or enlargement should be

taken into consideration.

All illustrations, whether line drawings or photographs, should be termed figures (plates

are not printed; half-tones will appear in their proper place in the text) and numbered in a

single series. Items of composite figures should be designated by capital letters; lettering of

figures is not set in type and should be in lower-case letters.

The number of the figure should be lightly marked in pencil on the back of each illustration.

5. REFERENCES cited in text and synonymies should all be included in the list at the end of

the paper, using the Harvard System (ibid., idem, loc. cit., op. cit. are not acceptable):

(a) Author’s name and year of publication given in text, e.g.:

‘Smith (1969) describes .. .’

‘Smith (1969: 36, fig. 16) describes...’

‘As described (Smith 1969a, 19695; Jones 1971)’

‘As described (Haughton & Broom 1927)...’

‘As described (Haughton et al. 1927)...’

Note: no comma separating name and year

Dagination indicated by colon, not p.

names of joint authors connected by ampersand

et al. in text for more than two joint authors, but names of all authors given in list of references.

(b) Full references at the end of the paper, arranged alphabetically by names, chronologically

within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year, e.g. Smith (1969a, 19695) and not Smith (1969, 1969a).

For books give title in italics, edition, volume number, place of publication, publisher.

For journal article give title of article, title of journal in italics (abbreviated according to the World list o,

scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses, volume number, part

number (only if independently paged) in parentheses, pagination (first and last pages of article).

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FISCHER, P.—H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris 88: 100-140.

FiscHer, P.-H., DuvAL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des littorines. Archs

Zool. exp. gén. 74: 627-634.

Koun, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee region of Ceylon.

Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 19606. Spawning behaviour, egg masses and larval development in Conus from the Indian Ocean.

Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L. Zoologische

und anthropologische Ergebnisse einer Forschungsreise im westlichen und zentralen Siid-Afrika 4: 269-270.

Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16: 269-270.

(continued inside back cover)

“TS Wu 06 ‘ZSEET ISN ‘“UMPW “(OZ8T “MleUog) snivisiu9 NZ *q “O

“TS WU 8rZ “6LOLZ-NVS ‘oftuoanlorg “(OZ8T “Mlouog) smmisi4o NZ “q

“TS WU 96 “O8LLZ-WVS ‘aTtuoanlorg “(L681 ‘AqitsO) wniousdjod puapowusaq “VW

‘T ‘Sy

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 94 Band

June 1984 Junie

Part 2 Deel

nf

a>

THE FAMILIES TRACHIPTERIDAE

AND RADIICEPHALIDAE

(PISCES, LAMPRIFORMES)

AND A NEW SPECIES OF ZU FROM

SOUTH AFRICA

By

P. C. HEEMSTRA

&

S. X. KANNEMEYER

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town 8000

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na gelang van die

beskikbaarheid van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad 8000

OUT OF PRINT/UIT DRUK

fh, DOS, 5-8), UD, 4-5, 8 tai), SUS, 5, 7-9),

GCL isso), HOD), 8, NAD, 7), IOGES),

1n\(1=2; 5,7, tpn), 144-2), 15425), 240) 27, 31-3) 32.) son seas)

Copyright enquiries to the South African Museum

Kopieregnavrae aan die Suid-Afrikaanse Museum

ISBN 0 86813 057 5

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

THE FAMILIES TRACHIPTERIDAE AND RADIICEPHALIDAE

(PISCES, LAMPRIFORMES)

AND A NEW SPECIES OF ZU FROM SOUTH AFRICA

By

P. C. HEEMSTRA

J. L. B. Smith Institute of Ichthyology, Grahamstown

&

S. X. KANNEMEYER

South African Museum, Cape Town

(With 11 figures and 2 tables)

[MS accepted 26 October 1983]

ABSTRACT

The trachipterid fishes of South Africa comprise five species: Trachipterus trachypterus

(Gmelin, 1789), T. jacksonensis (Ramsay, 1881), Desmodema polystictum (Ogilby, 1897), Zu

cristatus (Bonelli, 1820), and a new species Zu elongatus. T. nigrifrons Smith, 1956, is considered

a synonym of 7. jacksonensis, and T. woodi Smith, 1953, appears to be a synonym of

D. polystictum. Zu elongatus sp. nov. is described from four specimens (including two adults)

from off the western Cape Province. It differs from Zu cristatus in body depth, size of eye,

number of lateral-line scales and vertebral counts. Zu elongatus probably also occurs in New

Zealand. The rare adult stage of Zu cristatus is described from a specimen captured off Kenya.

Observations on a juvenile specimen of Zu swimming in the ocean are reported for the first time.

A juvenile specimen of Radiicephalus elongatus Osorio, 1917 (the first South African record for

this species) is also described.

CONTENTS

PAGE

NERO GUCHON Aas Areiae ss yniin: 5A ae Mes einem on a cer eee 13

INES TNO CSW ore coc Se Tape Res tre neecae Mee rier 14

Ramilyanachiptendaes.)o% «14 sb esas selesnas one Saeeee ree 14

Key to genera of family Trachiptendae ...=...:--.5--5--- 15

(GenuSsD eSInOdcCiNGn et ts... a has ne tee eee 15

GENUS HU ACHIPICTUS ie stewie caren ee Be On he ee eee 17

GCM SZ ee recent. < MES FER eS SS 23

amMilyaRaGncepWalida: «jai ss Savatage ae nak aeae aes 35

PNCKHOWIEUS CMICTES ors tel ots cys Beweat Piae Ay ere ee aes 37

FIRE TCREIMCES ert iarae nl ie ere sehen deter eye Sen ce Ree eee Si)

INTRODUCTION

As part of the research being done for a forthcoming revision of J. L. B.

Smith’s (1953) book The sea fishes of southern Africa, the trachipterid fishes of

the J. L. B. Smith Institute were examined. In comparing specimens at the South

iS

Ann. S. Afr. Mus. 94 (2), 1984: 13-39, 11 figs, 2 tables.

14 ANNALS OF THE SOUTH AFRICAN MUSEUM

African Museum with the meagre data from the material in the Smith Institute,

the second author (S. X. K.) discovered a new species of the genus Zu. The first

South African record of the related family Radiicephalidae is based on a juvenile

specimen recently collected in an Engels trawl 70 miles south-west of Cape Point.

METHODS

All measurements are from point to point, and those from the front of the

snout were taken from the premaxillary symphysis with the upper jaw retracted.

Standard length (SL) is measured from the front of the snout to the base of the

upper caudal fin rays. Snout to vent distance (SV) is from the front of the snout to

the anus. Head length is from the front of the snout to the posterior margin of the

gill cavity. Snout length is from the front of the snout to the anterior margin of the

orbit. Lower jaw length is from the anterior tip of that jaw to the posterior end of

the articular bone. Maxilla width is the greatest width of the maxilla. The

anteriormost vertebra with a haemal spine is considered the first caudal vertebra.

Body depth is the greatest depth of the body. Pre-anal vertebrae are those

anterior to a vertical line through the anus. Vertebral counts were made from

radiographs or from cleared and stained material. Lateral-line counts begin with

the scale just above the opercle and stop at the end of the vertebral column.

In the lists of material examined, abbreviations are as follows:

BMNH British Museum (Natural History)

NMNZ National Museum of New Zealand

RUSI J. L. B. Smith Institute of Ichthyology

SAM South African Museum

SIO Scripps Institute of Oceanography

USNM US National Museum of Natural History

Family Trachipteridae

Diagnosis

Body elongate, more or less compressed. Dorsal fin origin above or slightly

behind eye; lateral edges of dorsal fin rays with minute spines or prickles (con-

spicuous in prejuveniles but less obvious in adults). Ventral rays of caudal fin

reduced (rudimentary or absent in some species); pelvic fins rudimentary or

absent in adults; in prejuveniles the pelvic fin rays are elongate and the first ray

bears a row of prickles; anal fin absent. Upper jaw highly protrusile. Posterior

vertebrae of adults elongate, two to four times longer than tenth vertebra; ribs

absent. Ink sac absent; swim-bladder rudimentary. Skin of adults with bony or

cartilaginous tubercles and pierced by numerous pores; lateral line with cartilagi-

nous or bony plates.

Remarks

The monotypic family Radiicephalidae was redefined by Harrisson & Palmer

(1968); the remaining families and genera of the suborder Trachipteroidei have

FAMILIES TRACHIPTERIDAE AND RADIICEPHALIDAE 15

been reviewed by Walters & Fitch (1960). The trachipterid species of the north-

east Atlantic and Mediterranean were revised by Palmer (1961), and those of the

eastern Pacific were treated by Fitch (1964).

Like the related oar-fish (family Regalecidae), the trachipterids are members

of the off-shore epipelagic community (Parin 1968). They have been observed

swimming in a head-up position by means of undulations of the dorsal fin

(Nishimura & Hirosaka 1964). The peculiar skin tubercles and pore system are

thought to reduce drag when the fish is swimming (Walters 1963). Three genera

are recognized. Although more than thirty species have been described, there are

probably fewer than ten valid species in this family.

KEY TO GENERA OF FAMILY TRACHIPTERIDAE

la Caudal fin comprises 4-10 dorsal rays approximately parallel to the long axis of the body;

ventral caudal rays absent; no elongate spiny plates or bony tubercles along ventral edge of

cbt PIE 8 PSS k tks GN. aay 8 es Mee lieg Ia Ps oi hie oop eels cuaiae Agia Desmodema

1b Caudal fin of two parts: 6—12 dorsal rays usually set at a steep angle to the long axis of the

body and 1-5 ventral rays (rudimentary spiny nubbins in Trachipterus and adult specimens of

Zu); elongate spiny plates or bony tubercles along ventral edge of tail.................. 4

2a Lateral line runs along ventral edge of tail as a zigzag series of elongate spiny plates; subadults

(less than 700 mm SL) with ventral edge of body distinctly scalloped; body depth 1,5—2,9 in

PON MIE ie eR et dealin ns AAG wale Sidwed Vek oe hes PSSA ME ame yeu ee. eee Aen Zu

2b Lateral line straight, well above ventral edge of tail in adults (specimens more than 1 500 mm

SL); ventral edge of tail in adults with bony tubercles but no elongate spiny plates; ventral

cancomnoay straight: body depth 3;7—4,lin SV ..-.. 2.52 52220: +s8- ens ee: Trachipterus

Genus Desmodema Walters & Fitch, 1960

Diagnosis

Seven dorsal pterygiophores before first neural spine; no predorsal bones;

caudal fin on same horizontal axis as caudal peduncle, comprising 4—10 dorsal

rays; ventral caudal rays absent; tail greatly elongated in juveniles and adults;

ventral body profile of prejuveniles (specimens less than 100 mm SV) not scal-

loped. Prejuveniles silvery, with numerous round dark spots.

Remarks

This genus was recently revised by Rosenblatt & Butler (1977). They recog-

nize two species. D. Jorum Rosenblatt & Butler, 1977, is known only from the

North Pacific.

Desmodema polystictum (Ogilby, 1897)

Fig. 1A

Trachipterus jacksoniensis polystictus Ogilby, 1897: 649.

Trachipterus sp.: Smith, 1949: 504, fig. 264b.

Trachipterus woodi Smith, 1953: 504, fig. 264b; 1966: 83, pl. 15 (fig. A).

Trachypterus woodi: King & Ikehara, 1956: 22, fig. 5.

Desmodema polystictum: Rosenblatt & Butler, 1977: 848, figs 1-4.

(See Rosenblatt & Butler (1977) for additional synonyms and references.)

16 ANNALS OF THE SOUTH AFRICAN MUSEUM

Material

Holotype of 7. woodi Smith, RUSI 133, SV 89 mm (rear end of tail missing);

Xora River mouth, Transkei.

SIO 75-139, SL 100 mm, SV 72 mm; 10°00’N 119°00’W; midwater trawl;

0-50 m.

SAM-—27786, SL 96 mm, SV 68 mm; found in a tidal pool at Simon’s Town,

False Bay, Cape Peninsula.

USNM 164171, SL 113 mm, SV 78 mm; 04°41’'N 159°53’W (radiograph only

seen).

Diagnosis

This diagnosis is according to Rosenblatt & Butler 1977, plus data from

specimens mentioned above.

Dorsal fin rays 120—128; pectoral fin rays 12-14; caudal fin rays 7-10; gill-

rakers 2-3+9-10. Vertebrae: 71-74 total, 18-20 precaudal, 37—42 pre-anal.

Eye diameter greater than snout length. Tail greatly elongated in adults; SV

about two-thirds SL in prejuveniles and one-third SL in adults.

Remarks

Smith (1966) gave an amended, amplified description of the holotype of

T. woodi. The following data not supplied by Smith are added here. SV 89 mm,

body depth at pelvic fin origin 33 mm, head length 24mm, eye diameter

10,1 mm, snout length 7,0 mm, maxilla width 6,0mm, lower jaw length

10,6 mm. All the pelvic fin rays have been broken off at the base, but there seem

to be 7 nubbins on the left side and 8 on the right. Because of the poorly ossified

bones of this small specimen (perhaps made worse by preservation in formalin), it

is difficult to get a clear radiograph. The anterior dorsal fin pterygiophores in

relation to the first neural spine and the transition from precaudal to caudal

vertebrae cannot be discerned, but the number of pre-anal vertebrae is clearly 41.

Data from the SAM specimen are as follows: SL 96 mm; SV 68 mm; dorsal

fin rays 128, first five rays elongate, the longest is 37 % SL; pectoral rays 13; pelvic

rays 9, elongate, reaching caudal fin; caudal rays 8; gill-rakers 3+ 9; lateral-line

scales 158, each plate with a single spine except fifteenth, which has a double

spine; anus asymmetrical, on left side 1,5 mm above ventral edge of body.

Vertebrae: precaudal 20, pre-anal 37, total 74.

Comparison of the holotype of T. woodi with one of the prejuvenile speci-

mens of D. polystictum examined by Rosenblatt & Butler (1977) revealed no

significant differences between these two specimens. 7. woodi is therefore con-

sidered a synonym of D. polystictum. The possibility of T. woodi being a synonym

of D. polystictum and the record of T. woodi from the Pacific (King & Ikehara

1956) were apparently overlooked by Rosenblatt & Butler (1977).

Distribution

South Africa, Australia, Japan, eastern Pacific and western Atlantic.

FAMILIES TRACHIPTERIDAE AND RADIICEPHALIDAE 17

Genus Trachipterus Gotan, 1770

Diagnosis

Body depth of adults 3,7—4,4 in SV, not greatly constricted behind vent.

Ventral edge of body straight, with enlarged pointed tubercles. Lateral-line plates

of adults more or less fused to form an almost continuous tube with rudimentary

blunt spines; prejuveniles and juveniles with separate, oval lateral-line plates,

each with a central, well-defined spine. In adults the lateral line runs well above

the ventral edge of the tail, but in smaller specimens the lateral line runs fairly

close to the ventral edge of the tail along the rear half of the tail. Dorsal and

pelvic fin rays of prejuvenile stage without series of bulbous or leaf-like appen-

dages; first five or six dorsal rays, pelvic and caudal rays elongate in prejuveniles

and juveniles. Dorsal caudal rays well developed, set at steep angle to horizontal

axis of peduncle. One pterygiophore before first neural spine, and 8 or 9 between

first and second neural spines. Scales absent.

KEY TO SOUTH AFRICAN SPECIES OF TRACHIPTERUS

la Pectoral fin rays 8-11; eye diameter 1,4-1,6 in lower jaw length; tubercles small and

indistinct except along ventral edge of body and tail, not arranged in rows along each side of

SsiiSRPECRYUSIOPHORES 5 2. Gs. eeu gids Sean ake ote ere eee meee Welee T. trachypterus

15 Pectoral fin rays 13 or 14; eye 1,9-—2,2 in lower jaw length; tubercles distinct, those along

ventral edge of body and tail larger, pointed and slightly curved; tubercles arranged in rows

alonmeaen side of dorsal fin pterypiophores ..........2..26.2..5-2---=- T. jacksonensis

Trachipterus jacksonensis (Ramsay, 1881)

Figs 2-3

Regalaecus jacksonensis Ramsay, 1881: 631, pl. 20.

Trachipterus jacksonensis: McCulloch, in Whitley, 1927: 296, pl. 25 (fig. 2).

Trachipterus nigrifrons Smith, 1956: 449, text fig. and pl. 12; 1965: 18.

Material

Holotype of Regalaecus jacksonensis Ramsay, Australian Museum A9114

(radiograph only).

Holotype of T. nigrifrons Smith, RUSI 52, 1 210 mm total length (fide Smith

1956), East London.

RUSI 13128, 1 320+ mm total length; East London.

SAM-—25093, 2 110 mm SL, 1 050 mm SV; trawled west of Cape Town.

SAM-—26731, 1 220+ mm SL, 600 mm SV; trawled west of Cape Peninsula.

SAM-—28380, 1 490 mm SL, 700 mm SV; trawled 34°31’S 17°59’E.

Diagnosis

Dorsal fin rays 166-173, the first 6 rays short and hairlike; pectoral fin rays 13

or 14; gill-rakers 3-5+8. Vertebrae: 81-83 total, 31—34 precaudal, 44—48 pre-

anal. Eye diameter 1,9—2,2 in lower jaw length, 6,0—7,3 % SV. Dermal tubercles

on body distinct, those along ventral margin of body and tail enlarged, pointed

and slightly curved; tubercles in rows along each side of dorsal fin pterygiophores

18

ANNALS OF THE SOUTH AFRICAN MUSEUM

Drawn from holotype of T. nigrifrons Smith, RUSI 52, 1 210 mm TL. Pectoral

f anus; inset shows detail of skin tubercles from middle of tail

fin reconstructed. Arrow indicates position o

Fig. 2. Trachipterus jacksonensis (Ramsay, 1881).

FAMILIES TRACHIPTERIDAE AND RADITCEPHALIDAE 19

Fig. 3. Trachipterus jacksonensis (Ramsay, 1881). Arrangement of tubercles in

vertical rows along dorsal fin pterygiophores.

(Fig. 3). Head and body silvery; front of head, interorbital area, dorsal part of

upper jaw and tip of lower jaw black.

Remarks

Smith (1956) compared his new species 7. nigrifrons with two species

(T. ishikawae and T. ijimae) described by Jordan & Snyder (1901) from Japan.

T. ijimae was synonymized with Zu cristatus by Fitch (1964: 231). In comparing

T. nigrifrons and T. ishikawae, Smith (1956) used data from a later paper by

Jordan & Snyder (1904), which is based on two non-type specimens that may or

may not be conspecific with the holotype of T. ishikawae. In the original descrip-

tion of T. ishikawae the numbers of pectoral fin rays and gill-rakers are not given,

and without having recourse to the holotype the identity of this species must

remain in coubt.

Smith (1956) apparently overlooked the detailed description of T. jack-

sonensis (Ramsay, 1881) given by McCulloch (in Whitley 1927). McCulloch’s

20 ANNALS OF THE SOUTH AFRICAN MUSEUM

description agrees with Smith’s (1956) account of T. nigrifrons, and McCulloch,

after comparing his specimen with the holotype of T. jacksonensis, stated that

(p. 297) ‘it appears a little deeper than the holotype, to which it is exactly similar

in every other detail’.

Distribution

South Africa and Australia.

Trachipterus trachypterus (Gmelin, 1789)

Fig. 4

Cepola trachyptera Gmelin, 1789: 1187.

Trachypterus arawatae Clarke, 1881: 195, text-fig.

Trachipterus jacksonensis (non Ramsay): Hamilton, 1916: 372, fig. 1.

Trachipterus arcticus (non Briinnich): Barnard 1925: 353,; pl. 14 (fig. 3); 1948: 359, fig. 17.

Trachipterus iris: Smith, 1949: 141, fig. 264.

Trachipterus trachypterus: Hamilton, 1916: 374, figs 3-6. Palmer, 1961: 342, pl. 62 (fig. 2).

(See Palmer (1973) for additional synonyms and references.)

Material

BMNH 1891.8.31.27-35, 7 specimens, North Atlantic (radiograph only

seen).

RUSI 13106, 1 620+ mm SL, 710 mm SV; South Africa, no data.

SAM-=11894, 540+ mm SL, 236 mm SV; South Africa, no data.

SAM-22078, 1 860 mm SL, 900 mm SV; trawled off Dassen Island (33°25'S

18°E).

SAM-—23677, 113 mm SL, 82 mm SV; trawled off Slangkop, Cape Peninsula.

SAM-—24024, c. 520 mm SL (damaged), 250 mm SV; trawled west of Cape

Town.

SAM-—26708, 55 mm SL, 40 mm SV; 33°06’S 16°08'E; 150-0 m; N100B

plankton net.

Diagnosis

Dorsal fin rays 145-184, first 5 or 6 rays elongate in juveniles and prejuve-

niles, forming a dorsal pennant; pectoral rays 9-11; gill-rakers 3-5 + 10. Verte-

brae: 84-96 total, 35-39 precaudal, 49-57 pre-anal. (Vertebral counts for the

New Zealand specimen NMNZ 2056 are c. 97 total and about 42 precaudal.) Eye

diameter (of adults) 1,4—1,6 in lower jaw length, 8,4-8,9 % SV. Dermal tubercles

on body small, except along ventral margin of body and tail, uniformly distributed

over dorsal fin pterygiophores (Fig. 4). Head and body silvery; juveniles with 3-5

dark blotches on body.

Remarks

The present specimens generally agree with Palmer’s (1961) account of

T. trachypterus (Gmelin, 1789), except that they do not have the body axis curved

dorsally in the posterior caudal region, and the 540mm SL specimen

FAMILIES TRACHIPTERIDAE AND RADIICEPHALIDAE 21

Fig. 4. Trachipterus trachypterus (Gmelin, 1789). Uniform arrangement of tuber-

cles along dorsal fin pterygiophores.

(SAM-11894) seems to have the greatest body depth nearer the anus than the

head, instead of just behind the head. According to Palmer (1961) T. trachypterus

differs from 7. arcticus in the dorsally curved tail and in having the greatest body

depth immediately behind the head; in T. arcticus, the tail is straight and the

greatest body depth is ‘3 to 3 of the way along its length, except in specimens of

less than 300 mm’.

The original description of Cepola trachyptera Gmelin, 1789, is inadequate to

distinguish which species Gmelin was trying to describe: ‘C. capite declivi, maxilla

utraque fornicata, pinnis aculeatis serratis scabris. Habitat in mari adriatico, linea

laterali recta, squamarum serie simplici; an secundum Forsterum proprium genus?’

(Cepola with downward-sloping head [i.e. steep forehead], maxilla arched on either

side, fins rough [with] saw-like prickles. Habitat in Adriatic Sea, lateral line straight,

scales in a single row; according to Forster it may belong in a genus of its own.)

C. trachyptera was the third species of Cepola listed by Gmelin (1789: 1186-

1187); the first two (C. taenia and C. rubescens) are considered synonyms of

pip ANNALS OF THE SOUTH AFRICAN MUSEUM

C. macrophthalma (Linnaeus, 1758). By his placement of trachyptera in Lin-

naeus’s genus Cepola, Gmelin probably recognized at least a superficial similarity

between his new species and C. macrophthalma. Hence it seems reasonable to

suppose that C. trachyptera was an elongate fish with a markedly compressed

body and a dorsal fin that ran the length of the body. The steep forehead rules out

trichiurids.

Except for trachipterids, the only Adriatic fishes with an elongate com-

pressed body, steep forehead, and long dorsal fin are the oarfish Regalecus glesne

Ascanius, 1772, and Lophotus lacepedei Giorna, 1809. The statement about the

fins ‘pinnis aculeatis serratis scabris’ apparently refers to the row of minute spines

or prickles along the lateral edges of the dorsal fin rays of trachipterid fishes. The

species name trachyptera means ‘rough fin’, being derived from the Greek words

trachys (=rough) and pteron (= wing or fin). These fin-ray prickles are lacking in

the oarfishes but very conspicuous in juvenile trachipterids.

We are left, then, with Trachipterus trachypterus (sensu Palmer 1961) and Zu

cristatus as the Adriatic species that best fit the description given by Gmelin.

Because of its greater body depth (which would render the ‘Cepola’ designation

less likely) and the distinctive features of the prejuveniles (which are unlikely not

to be noticed and mentioned in a description), Zu cristatus seems the less

plausible of the two species. Furthermore, 7. trachypterus is fairly common in the

Adriatic Sea (Jardas 1980).

By a process of elimination together with a few assumptions, we have come

to the conclusion that Cepola trachyptera Gmelin is the species currently recog-

nized as Trachipterus trachypterus.

Hamilton (1916) published descriptions of seven specimens of trachipterids

from New Zealand. The specimen from Nelson, South Island, that he illustrated

(fig. 2) from his original sketch was doubtfully assigned to T. jacksonensis. This

specimen, later illustrated by Whitley (1968: 46, fig. 1) under the name ‘Des-

modema arawatae’, appears to be Zu elongatus sp. nov. Judging from the count

of eleven pectoral fin rays given for five of the specimens, Hamilton’s other

specimens appear to be TJ. trachypterus. Hamilton identified his largest,

‘approximately 6 ft’ specimen with T. jacksonensis because it had smooth dorsal

fin rays, instead of ‘all rays spinous, and having a spiny tubercle at the base’ for

his juvenile specimens of T. trachypterus. Hamilton assumed that the minute

spinules or granulations on the dorsal fin rays of juveniles were a good specific

character, and were not lost or greatly reduced with growth. However, this is

precisely what happens with all trachipterids. In our adult specimens of

Trachipterus and Zu the spinules are absent or very much reduced, compared

with those of juveniles.

Distribution

Mediterranean Sea, eastern Atlantic to South Africa, Japan and New Zea-

land.

FAMILIES TRACHIPTERIDAE AND RADITCEPHALIDAE 23

Genus Zu Walters & Fitch, 1960

Diagnosis

Young with pelvic and anterior 6 dorsal fin rays greatly elongated, bearing

serially arranged, membranous flaps (Figs 1D, 8). One pterygiophore before first

neural spine, and 8 or 9 between first and second neural spines. Scales present,

though very deciduous, and not apparent in prejuveniles. Body depth 1,7—2,8 in

SV, more or less constricted behind anus. Ventral edge of body distinctly scal-

loped in specimens less than 700 mm SL; belly of specimens more than 800 mm

SL nearly straight, but with a low fleshy keel from pelvic fin rudiments to anus.

Lateral line drops to ventral edge of tail shortly behind anus, from where it runs

aS a zigzag series of elongate spiny plates to base of caudal fin and is then

continued on to the ventral caudal filament (of prejuveniles) as a straight tube.

Total vertebrae 63-69 or 84-87.

KEYolO;SPECIESIOE ZU

la Body depth 12-16% SL; prejuveniles with body not abruptly constricted behind vent; eye

phere lateral-linesplates 126-130 2... ce. ne a le ee Z. elongatus sp. nov.

1b Body depth 20-26% SL; prejuveniles with body abruptly constricted behind vent; eye

easier orsy « lateral-line plates 99-106... a... ccs ae sees aa eee Z. cristatus

Zu cristatus (Bonelli, 1820)

Figs 1B—D, 5-8

Trachypterus cristatus Bonelli, 1820: 487. Smith, 1949a: 99; 1949b: 142, fig. 264a. Dieuzeide et

al., 1954: 151, figs. Tortonese, 1958: 1. Parin et al., 1978: 176.

Gymnetrus repandus Metaxa, 1833: 53, fig. 7.

Trachypterus bonelli Valenciennes, in Cuvier & Valenciennes, 1835: 331.

Gymnetrus miillerianus Risso, 1840: 13, pl. 125.

Trachypterus repandus: Costa, 1842: 11, pl.

Trachypterus semiophorus Bleeker, 1868: 279, pl. XII. Weber & de Beaufort, 1929: 89.

Trachypterus iris (non Walbaum): de Buen, 1917: 23, fig.

Trachypterus gavardi Bounhiol, in Bounhiol & Gavard, 1923: 1.

Trachypterus ijimae Jordan & Snyder, 1901: 311, pl. 17 (fig. 9). Tanaka, 1915: 358, pl. 96 (fig.

301). Bolin, 1933: 35. Okada & Suzuki, 1956: 168, fig. 1. Lindberg & Legeza, 1965: 298.

Zu cristatus: Walters & Fitch, 1960: 445. Palmer, 1961: 346. Fitch, 1964: 229, fig. 1. Backus er al.,

1965: 139, fig. 2. Ibanez & Gallego, 1974: 523, fig. 1. Karrer, 1975: 72. Fitch & Schultz, 1978:

89, fig. 5. Parin et al., 1978: 176. Oliver, 1982: 60, figs 1-2.

(See Palmer (1973) for additional references.)

Material

RUSI 7185, 2 specimens: 64 mm SL, 31 mm SV; 204 mm SL, 90 mm SV;

South Africa, no data.

RUSI 7187, 232 mm SL, 106 mm SV; Algoa Bay, eastern Cape Province.

RUSI 7188, 173 mm SL, 77 mm SV; Kariega River mouth, 33°36'S 26°42’E.

RUSI 13352, 950 mm SL, 450 mm SV; off Mombasa, Kenya.

SAM-—24522, 415 mm SL, 170 mm SV; off Slangkop lighthouse, Cape Penin-

sula, Cape Province.

SAM-—27679, 248 mm SL, 107 mm SV; 28°23,5’S 32°41,9’E.

ANNALS OF THE SOUTH AFRICAN MUSEUM

24

[He] UO saTkdg “6LOLZ-NVS ‘a[uoanloig “(OZ8T “H[eUog) snqvrsi4o NZ

‘¢ “314

Lv a)

—N

FAMILIES TRACHIPTERIDAE AND RADITCEPHALIDAE

‘sajejd oul] [e19je] JO saseq BULIDAOD AqjeIIed sayeog “6L9LZ-WVS ‘oytuoanloig

‘(OZ8I

6

JoUOg) SnIvISMI NZ “9 “Bly

ANNALS OF THE SOUTH AFRICAN MUSEUM

26

‘snue

puryaq ysnf poyenqis ‘poyoryje juowrsid YIM sojeos uTYy) AIOA Jo sayojed Aulys “6L9LZ-WVS ‘oTtueanfoig “(OZ8T ‘I[ouog) sniwisiua NZ */ ‘BLA

FAMILIES TRACHIPTERIDAE AND RADITCEPHALIDAE a

Diagnosis

D 120-150; P 10-12; V 5-7 (absent in adults); dorsal caudal fin rays 6-12,

set at a steep angle to horizontal axis of caudal peduncle; ventral caudal rays

represented by spiny nubbins in adult; prejuveniles with 1—5 ventral caudal rays,

with 1 or 2 produced into long filaments. Gill-rakers 2-3+8-9. Lateral-line

plates 99-106; lateral line begins above middle of eye and continues beyond the

vertebral column on to the caudal fin between the second and third ventral caudal

rays (about 16 pored scales on the caudal membrane of SAM-—27679). Vertebrae:

63— 69 total, 22—24 precaudal, 32 pre-anal. SV 41-48 % SL; body depth 20-26 %

SL; eye diameter 1,3—1,7 in length of lower jaw, 13-16% SV. Young silvery,

with about 6 wavy dark vertical bars on dorsal part of body and 4 on ventral part;

tail with about 6 complete black bars; caudal fin mostly blackish, proximal part

pale. Adults silvery grey, paler ventrally; dorsal fin scarlet; caudal fin reddish

black, darker distally.

Description of adult (RUSI 13352, 950 mm SL)

D 145; P 11; dorsal caudal fin rays 9; gill-rakers 3+ 8. Body robust (not

nearly as compressed as adults of Trachipterus or Zu elongatus sp. nov.), with

fleshy keel along belly; body covered with small, very deciduous, thin, cycloid

scales. (All scales on body are missing, but the scale pockets are distinct; a few

scales were left on the tail near the lateral line.) Lateral line curving gradually

from above opercle to ventral edge of tail 100 mm behind anus; lateral-line scales

spiny, represented by two rows of staggered spiny scutes along ventral part of tail.

Pelvic fins represented by a few nubbins. Single nostril on each side of snout.

Pseudo-branch well developed. Pyloric caeca very numerous. Swim-bladder not

apparent.

Measurements for six specimens are given in Table 1.

TABLE 1

Measurements (in mm) of specimens of Zu cristatus.

RUSI RUSI RUSI RUSI SAM SAM

7185 7187 7188 13352 24522 27679

Standard length (SL) 205 232 173 950 415 248

Snout to vent (SV) 91 108 76 451 170 107

Greatest body depth 49 56 43 223 84 64

Body depth at anus 19 21 18 123 4] 23

Head length 31 — 7a 161 60 SH

Eye diameter 12 14 10 69 Di 14

Maxilla width 6,3 4.3 5,8 By 18) 9

Lower jaw length 19 22 16 92 36 8,

Remarks

Adults of Zu cristatus are apparently rarely captured; only four specimens

longer than 800 mm SL have been reported (Bolin 1933; Tortonese 1958; Ibanez

& Gallego 1974; Fitch & Schultz 1978). Judging from the illustration of a 610 mm

ANNALS OF THE SOUTH AFRICAN MUSEUM

28

‘(souor ua Aq Yyo}ays B UO pase) ULID0 UT SUTUIUIMS

ajiuaanf aaty e Jo SuIMLIG “YISUIT UI WU QQT NoGY “(OZ8T ‘]aUOg) snwISU9 NZ °g “314

aca

| | i 4

— — ' «1200IEN S EE OO OOOO.—OOOOanawnaawn nn I.

FAMILIES TRACHIPTERIDAE AND RADIICEPHALIDAE 29

prejuvenile by Tortonese (1970, fig. 190b), metamorphosis takes place in the

600-800 mm size range. The small prejuveniles that we have examined

(64-248 mm SL) are much more compressed than the adults. The pelvic fins and

elongated anterior dorsal fin rays are lost during the transition from prejuvenile to

juvenile stage.

Despite previous reports of deciduous cycloid scales on specimens of Zu

cristatus (Bolin 1933; Tortonese 1958; Walters & Fitch 1960; Palmer 1961; Fitch

1964), Rosenblatt & Butler (1977) discounted these observations, because they

could not find scales on their eight specimens (27,5—-811 mm SL). As noted

above, adults of Zu do have scales but the epidermis is very fragile, and there are

only a few scales remaining on the tail of the 950 mm SL specimen. On the

248 mm SL prejuvenile (SAM-—27679), the scales are most apparent along the tail

(Fig. 5), which they appear to cover completely, and continue on to the caudal fin

rays. Some of the bases of the lateral-line plates are partially covered by scales

(Fig. 6). The lateral surface of the tail just behind the anus has scales that seem to

have pigment attached to them. These scales are apparent on the fish as large

shiny patches (Fig. 7). The scales forming these shiny patches are exceedingly

thin, but will separate on being lifted with a fine needle. There is evidence of

Squamation over the entire body area, particularly ventrolaterally where many

scaled areas remain intact. The head does not appear to have scales.

Except for the report of Nishimura & Hirosaka (1964) of a large Trachipterus

specimen in a relatively small aquarium, there have been no published observa-

tions on live trachipterids. In view of the dearth of such reports, the following

observations of L. Jones of Fynnlands, Natal are of considerable interest. Judging

from Jones’s sketch, the specimen that he observed was probably Zu cristatus.

‘I was swimming just beyond the surf-line in 20 feet of water when I noticed

what at first appeared to be a small fish caught by a stinging plankton or jellyfish

some 6-10 inches below the surface. On closer examination I found it to be a fish

with long thread-like fins: . . . a small silvery fish with big head and eye, with red-

brown markings, about 33 inches long. The six [elongated] anterior dorsal rays

were at least 15 inches long, with what appeared to be pairs of tiny brown leaf-like

appendages set 13 to 2 inches apart down the length of each dorsal ray. Its pelvics

were also very long, about the same length as the body, but I can’t remember if

the “‘leaves”’ were present on these or not. Its tail was all but invisible, but the

dorsal fin seemed to be part of its propulsive mechanism (the lower or rear part

that is) for it swam in a head-up-tail-down manner which I have drawn for you’

(Fig. 8).

Zu elongatus sp. nov.

Figs 9A-11B

? Trachipterus jacksonensis (non Ramsay): Hamilton, 1916: 374, fig. 2.

Trachipterus jacksoniensis (non Ramsay): Phillipps, 1944: 120, pl. 52. (Unjustified emendation. )

Trachypterus arcticus (non Briinnich): McCann, 1953: 5 (figs 5—6 only).

? Desmodema arawatae (non Clarke): Whitley, 1968: 45, fig. 1.

ANNALS OF THE SOUTH AFRICAN MUSEUM

30

‘skei [epned JO aseq puke snue jo UOT}ISOd a}edIpUI SMOIIY “TS WUT CET |

‘T6E6T-NVS ‘VWNpe ‘odAjeieg ‘aou ‘ds snjvsuoja nZ -g

wool

WIOl

“TS WU 99T I “POLPZ-NVS ‘MNpe ‘edAjojox] ‘aou “ds snwsuoja NZ “VW “6 “SIA

q

FAMILIES TRACHIPTERIDAE AND RADITCEPHALIDAE 31

Holotype

SAM-—24704, 1 166 mm SL, 404 mm SV; adult, ?¢. Trawled south-west by

west of Cape Columbine, north-western Cape Province; bottom depth 411 m;

donated by Captain Jacoby, S. T. Rochea; collected 22 November 1965.

Paratypes

SAM-—29392, 1 135 mm SL, 421 mm SV; adult, ?¢. Trawled west of Cape

Town by Irvin & Johnson Ltd trawler; depth and date of collection unknown.

SAM-—19870, 630+ mm SL (part of tail missing), 228 mm SV; juvenile.

Trawled north-west of Table Bay by trawler of Irvin & Johnson Ltd; bottom

depth 580 m; collected in 1946, exact date unknown.

SAM-—24414, 480 mm SL (upper jaw fully protracted), 166 mm SV; prejuve-

nile. Trawled off Cape Columbine, north-western Cape Province; bottom depth

548 m. Collected and donated by P. Henry, Midwest Fisheries Ltd, Cape Town, 1

February 1965.

Other material

SAM—29396, 1 123 mm SL, 421 mm SV; adult, badly damaged; no data.

NMNZ P.834, c. 425 mm SL, 138 mm SV; French Pass, New Zealand.

NMNZ P.2248, 315 mm SL, 108 mm SV; Awanui, New Zealand.

Diagnosis

D 142-147; P 11-12; V 7, elongate in young, absent in adults; C 17 (12

dorsal rays and 5 ventral rays of which 2 may be represented by tiny spines). Gill-

rakers 2—3+ 7-9. Lateral-line plates 127-130. Vertebrae: 84—87 total, 29-31

precaudal, 38—40 pre-anal. SV 34-37 % SL; body depth 12-16 % SL; eye 9-10 %

SV. Teeth: 9-21 in upper jaw, 6—9 in lower law; adults with 2 vomerine teeth;

palatine teeth present or absent. Dermal tubercles and pore system present at all

stages; scale pockets well developed in adults (Fig. 10).

Description (measurements in Table 2)

Adult (holotype, 1 166 mm SL; Fig. 9A)

Body laterally compressed, tapering gradually to the elongate tail; dorsal

body profile convex, the ventral profile irregular; fleshy keel along ventral midline

of belly. Body, tail, and cheeks covered with tubercles and conspicuous pores;

scale pockets distinct; tubercles on cheek enlarged.

D 143, the posterior rays long, filamentous; P 11; pelvic fins absent, repre-

sented by bony ridge within a slit-like pocket; 2 of the 5 ventral caudal fin rays are

small rudimentary spines; the other caudal rays elongate, the longest 160 mm.

Gill-rakers 3 + 8; lateral-line plates c. 128. Teeth: upper jaw with 10 on right side

of symphysis, 11 on left side; lower jaw with 6 teeth on right side of symphysis and

3 on left side; 2 teeth on vomer, 3 teeth on right palatine, 2 teeth on left palatine.

Colour in alcohol: body tan, interspersed with paler tubercles; front of head

with triangular dark area, extending on to lips and gular region; posterior dorsal

fin rays black; caudal fin black except for a small, pale area proximally.

32 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 10. Zu elongatus sp. nov. Paratype, adult, SAM—29392, 1 135 mm SL. Scale

pockets.

TABLE 2

Measurements (in mm) of specimens of Zu elongatus sp. nov. Specimen SAM-24414 was fixed

with the upper jaw completely protracted and the head is somewhat damaged, consequently the

measurements SL, SV and head length are approximate estimates of what they would be with the

upper jaw retracted. Both eyes are also missing on this specimen.

Holotype Paratypes

SAM Catalogue No. 24704 19870 24414 29392 29396

Standard length (SL) 1166 627+ 460 1168s 1123

Snout to vent (SV) 404 228 154 421 421

Greatest body depth Wa 99 63 177 176

Body depth at anus 1138) 66 34 121 13%

Head length 137, 81 66 134 135

Eye diameter 4] 7h — 42 42

Maxilla width 38) 16 13 33 36

Lower jaw length 73 42 34 74 80

FAMILIES TRACHIPTERIDAE AND RADITCEPHALIDAE 33)

Prejuvenile (480 mm SL; Fig. 11A)

Body elongate, laterally compressed, gradually tapering to caudal fin, not

abruptly constricted behind vent; dorsal profile of body convex; ventral profile

anterior to anus strongly scalloped, with three well-defined ventral lobes. The

strong upward curve of the tail, characteristic of prejuveniles in Zu cristatus, is

not apparent in this specimen. Lateral line curving gradually from above opercle

to ventral margin of tail 30 mm behind anus. Body surface abraded; no scales or

scale pockets apparent.

D 147, the first 6 rays elongate; P 11; pelvic fins elongate; caudal directed

upward, the rays elongate (longest 170 mm). Gill-rakers 3+ 8-9; lateral-line

plates c. 127. Teeth: upper jaw with 8 on right side of symphysis, 10 on left side; 4

on each side of lower jaw symphysis; no teeth on palatines or vomer.

Colour in alcohol: three incomplete dark cross-bars on body; five dark bars

on tail. Triangular dark area above and extending across symphysis of upper jaw;

dusky area across lower jaw symphysis, extending on to gular region. Distal part

of caudal fin dark; some dorsal fin rays with traces of dark pigment.

Juvenile (630+ mm SL; Fig. 11B)

Body shape as for prejuvenile; ventral lobes smaller and less distinct. Caudal

fin missing; anterior dorsal rays and pelvic fins elongate. Dermal tubercles on

cheek enlarged; scale pockets visible at 40 x magnification.

D 137; P 12; gill-rakers 2-3 +7; lateral-line plates c. 130. Teeth: upper jaw

with 7 on right side of symphysis, 4 on left side; lower jaw with 2 teeth on right

side and 4 on left; 2 teeth on vomer and 1 on right palatine.

Colour in alcohol: six dark cross-bars on tail; symphysis of lower jaw dark.

Body abraded, with no other pigmented areas apparent.

Paratype SAM-—29392, 1 135 mm SL (Fig. 9B), differs from the holotype

SAM-—24704, 1 166 mm SL and from the other adult, SAM—29396, 1 123 mm SL,

in that it lacks palatine teeth. SAM-—29396 has only 9 teeth in the upper jaw.

However, the teeth in Zu elongatus are easily detached and since SAM-29396 is a

badly damaged specimen some teeth might have been lost.

Remarks

As denoted by the name, Zu elongatus is more elongate than Zu cristatus

(body depth 12-16% SL, versus 20-26 % SL). Other differences between these

two specimens are the size of the eye (9-10 % SV in Zu elongatus, 13-16 % SV in

Zu cristatus), number of the lateral-line plates (126-130 in Zu elongatus, 99-106

in South African specimens of Zu cristatus, but 114-118 reported by Fitch (1964)

for eastern Pacific material), and vertebral counts (84-87 total, 29-31 precaudal,

38—40 pre-anal for Zu elongatus; 63-69 total, 22-24 precaudal, 32 pre-anal for Zu

cristatus). Palmer (1961) and Fitch (1964) give counts of 64-65 and 63 for total

vertebrae of Zu cristatus from the eastern Atlantic and eastern Pacific. The ranges

of vertebral counts for South African specimens given above are from 6 speci-

mens of Zu cristatus and 5 of Zu elongatus.

ANNALS OF THE SOUTH AFRICAN MUSEUM

34

‘skei Jepned Jo aseq pue snue jo UOT}IsOd o}eoIpUI SMOIIY “TS WW +0¢9

‘OL86I-NVS ‘ortuaanl ‘adAjereg ‘aou ‘ds snjp8uoja nz ‘gq “TS WU O8P ‘pIpPZ-NVS ‘oTtueantaid ‘odAjereg ‘aou ‘ds snvsuojanZ Wy ‘[] S14

ws G

q

FAMILIES TRACHIPTERIDAE AND RADITCEPHALIDAE 35

Fitch (1964) recognized Trachipterus semiophorus Bleeker (described from

Indonesia) and T. ijimae Jordan & Snyder (from Japan) as synonyms of Zu

cristatus. All the other synonyms of Zu cristatus are based on specimens from the

Mediterranean Sea, and all the recognizable published descriptions and illustra-

tions of Zu from the Mediterranean are of Zu cristatus. It seems, therefore, that

Zu elongatus does not occur in the Mediterranean Sea.

The specimen described and figured by Whitley (1968: 45, fig. 1, as ‘Des-

modema arawatae’) is the same one reported by Hamilton (1916) from Nelson,

South Island, New Zealand as ‘Trachipterus jacksonensis Ramsay?’. This fish

appears to be Zu elongatus. The measurements given by Whitley (1968: 45), as

well as his figure, show the body depth to be 16 per cent of the standard length

(which is within the range for Zu elongatus, but not for Zu cristatus). We have

examined the specimen from French Pass, Cook Strait, reported (as Trachipterus

jacksoniensis) by Phillipps (1944) and (as Trachypterus arcticus) by McCann

(1953). Data from this specimen (NMNZ P.834) plus one other from New

Zealand (NMNZ P.2248) are not included in the diagnosis or description of Zu

elongatus given above. Counts and measurements from this New Zealand ma-

terial are as follows (P.834 given first): SL c. 425 (tip of tail separated from rest of

specimen), 315; SV 138, 108; greatest body depth 66, 48; body depth at anus

34, 22; head length 53, 43; eye diameter 15, 14; lower jaw length 31, 25. Lateral-

line scales 125, 131; pectoral-fin rays 12, 12; dorsal-fin rays, damaged, 6+132;

total vertebrae c. 88, 88; pre-anal vertebrae 39, 37; and precaudal vertebrae 32, 30.

Family Radiicephalidae

Diagnosis

Body elongate, strongly compressed, the depth decreasing gradually from

just behind head to last vertebra. Dorsal fin origin over eye. Anus well before

midpoint of body. Anal fin rudimentary, located about half-way between anus

and last vertebra. Pelvic fins rudimentary in adults, with 6—-9(?) rays in juveniles.

Lateral line continued beyond last vertebra where the canals of either side are

joined to form a large, thin-walled tube carried on the greatly elongated ‘rays’

that emanate from the ventral surface of the last three vertebrae. Length of

posterior vertebrae less than twice that of tenth vertebra. Upper jaw highly

protrusile. Skin with scattered soft papillae and numerous pores. Swim-bladder

well developed; ribs and ink-sac present.

Radticephalus elongatus Osorio, 1917

Radiicephalus elongatus Osorio, 1917: 114, figs 2-4. Walters, 1963: 456. Harrisson & Palmer,

1968: 185, figs 1-6. Palmer, 1973: 333. Parin et al., 1974: 116, fig. 13. Karrer, 1976: 179, pl. 1,

text-fig. 1. Parin, 1978: 163. Parin et al., 1978: 176.

Material

SAM-29416, c. 310 mm SL (upper jaw protracted), 420 mm TL; 70 miles

south-west of Cape Point; 600 m; Engels midwater trawl; collected by the South

African Sea Fisheries Institute, 24 May 1982.

36 ANNALS OF THE SOUTH AFRICAN MUSEUM

Description (measurements in mm)

Body and tail elongate, strongly compressed. SL c. 310; body depth at

pectoral fin base 32, at anus 24,3, and at anal fin origin 11,0; body width at

pectoral base 7,5; SV c. 124; snout to anal fin c. 188; pelvic fins to vertical at rear

end of pectoral base 20,5; from pelvic fins to anus 44, from anus to anal fin 67;

head length c. 51; orbit diameter c. 13; interorbital width 8; lower jaw length 26;

maxilla width 10,3.

Dorsal fin with 160 rays, the origin above the eyes, the last ray over the

twenty-sixth preural centrum; pectoral fin rays 10 (including the short, spine-like

first ray), the longest ray 10 mm; pelvic fins damaged, most of the rays broken off

at their bases, which number 6 or 7; one 47 mm ray is still attached to the pelvic

fin base; anal fin rays 6, the longest ray 3 mm; length of anal fin base 3 mm. Gill-

rakers 3+ 8, the upper rudimentary, the anteriormost very short; branchiostegal

rays 4+3 (left side with 4+2 rays, but the membrane is torn and one ray has

probably been lost).

Lateral line begins over eye and drops to 6 mm above the anus, thence

running close to the ventral edge of the tail (1 mm above anal fin base). No

individual lateral-line scales apparent; the canal comprises a large, thin-walled

(almost transparent) tube; at about the anal fin the tube runs along the ventral

edge of the vertebral column on a shelf formed by the elongate slender haemal

spines (haemal spine of eighteenth preural vertebra 12,4 mm long; centrum

length 2,8 mm); posterior to the end of the vertebral column, the lateral-line

tubes are carried between the dorsal and ventral caudal fin ‘rays’, which are

108 mm long; width of lateral-line tube at penultimate vertebra 1,0 mm and the

vertical height of this centrum is 0,6 mm.

Total vertebrae 121, precaudal vertebrae 38, pre-anal 45; ribs are present on

all precaudal vertebrae except the first three. Length of tenth centrum (2,2 mm)

equals height of centrum at its posterior end; length of tenth preural centrum

(3,2 mm) equals 4,5 times height of centrum at its posterior end.

Upper jaw with a patch of small sharp teeth plus one much larger retrorse

tooth at each side of premaxillary symphysis; a series of 7—9 slender sharp teeth

on each side of lower jaw symphysis; 3 slender sharp teeth on vomer and 2 or 3

similar teeth on each palatine.

Body and tail silvery in life; dorsal fin base blackish.

Remarks

The specimen described above agrees well with the 597 mm SL neotype

described by Harrisson & Palmer (1968) and the 520 mm SL specimen described

by Karrer (1976). The lack of countable lateral-line scales in the SAM fish may be

due to the abrasion that the specimen has suffered and/or its smaller size. The

difference in the number of branchiostegal rays reported by the previous authors

(6 versus 7 on the right side of the SAM fish) may reflect damage to their

specimens, or perhaps the fact that the anterior two rays on the epihyal are close

together and might have been counted as a single ray.

FAMILIES TRACHIPTERIDAE AND RADIICEPHALIDAE 37

The exact nature of the caudal fin ‘rays’ is not clear; the ventral group of two

or three ‘rays’ seems to comprise the greatly elongated haemal spines of the last

two or three preural centra. The rays above and below the lateral-line tubes are

longitudinally striate, but none of these rays is segmented.

The specimen described above represents the first record of this species for

the south-eastern Atlantic Ocean. Radiicephalus elongatus is known from the

Atlantic and Pacific oceans and may also occur in the Indian Ocean.

ACKNOWLEDGEMENTS

We are indebted to Mr Alwyne Wheeler and the British Museum (Natural

History) for supplying radiographs of Trachipterus trachypterus and copies of rare

publications. Mr Wheeler also reviewed the manuscript and offered advice on the

validity of Trachipterus trachypterus (Gmelin, 1789). Dr Richard H. Rosenblatt

of the Scripps Institute of Oceanography kindly lent a specimen of Desmodema

polystictum. A radiograph of D. polystictum was supplied by the Fish Division of

the US National Museum of Natural History. We thank Dr Douglas Hoesse and

the Australian Museum for data and a radiograph of the holotype of T. jack-

sonensis. We are grateful to Dr Graham Hardy of the National Museum of New

Zealand for the loan of specimens from New Zealand. Mr Jean-Michel Vinson

did the drawings (Figs 2 and 8). Mr Len Jones contributed his observations of a

live specimen of Zu. Prof. M. N. Bruton, Miss A. E. Louw and Dr P. A. Hulley

offered constructive comment on the manuscript. Mr R. E. Stobbs produced

several excellent radiographs.

REFERENCES

ASCANIUS, P. 1772. Icones rerum naturalium, ou figures enluminées d’histoire naturelle du Nord.

Part 2: 1-8. Copenhagen.

Backus, R. H., Meap, G. W., HAEprRicH, R. L. & EBELING, A. W. 1965. The mesopelagic

fishes collected during Cruise 17 of the R/V Chain, with a method of analysing faunal

transects. Bull. Mus. comp. Zool. Harv. 134: 139-158.

BARNARD, K. H. 1925. A monograph of the marine fishes of South Africa. Ann. S. Afr. Mus. 21:

1-418.

BARNARD, K. H. 1948. Further notes on South African marine fishes. Ann. S. Afr. Mus. 36:

341-406.

BLEEKER, P. 1868. Description et figure d’une nouvelle espéce de Trachypterus d V'Ile d’Am-

boine. Archs néerl. Sci. 3: 279-280.

Boutin, R. L. 1933. New fish records from southern California. Copeia 1933: 35-36.

BonELLI, F. A. 1820. Description d’une nouvelle espéce de poisson de la Méditerranée apparte-

nant au genre Trachypteére (T. cristatus) avec des observations sur les caractéres de ce méme

genre. Memorie Accad. Aci. Torino 24: 485-494.

BouNHIOL, J. P. & Gavarp,—. 1923. Une espéce nouvelle de Trachypterus Goiian: le

Trachypterus gavardi Bounhiol. Bull. Inst. océanogr. Monaco 432: \|-4.

Brunnicu, M. T. 1771. Collectio nova scriptorum Societatis Scientiarum Hafnensis 3: 418.

BUEN, F. DE. 1917. Pesces poco comunes de nuestras costas. Boln Oceanogr. Pesc. 2: 23-27.

CiarKE, F. E. 1881. Description of a new species of Trachypterus (T. arawatae). Trans. N.Z.

Inst. 13: 195-199.

Costa, O. G. 1842. Fauna del regno di Napoli ossia enumerazione di tutti gli animali, Pesci. Part

2. Naples.

38 ANNALS OF THE SOUTH AFRICAN MUSEUM

CuviER, G. & VALENCIENNES, A. 1835. Histoire naturelle des Poissons 10. Paris: Levrauit.

DIEUZEIDE, R., NoveLLa, M. & RoLanp, J. 1954. Catalogue des poissons des Cétes algérien-

nes. Bull. Stn Aquic. Péch. Castiglione 3 (5): 1-258.

Fitcu, J. E. 1964. The ribbonfishes (family Trachipteridae) of the eastern Pacific Ocean, with a

description of a new species. Calif. Fish Game 50: 228-240.

FitcH, J. E. & ScHuttz, S. A. 1978. Some rare and unusual occurrences of fishes off California

and Baja California. Calif. Fish Game 64: 74-92.

GiornA, M. E. 1809. Mémoire sur des poissons d’espéces nouvelles et de genres nouveaux.

Memorie Accad. Sci. Torino 16: 177-180.

GMELIN, J. F. 1789. Pisces. In: Carott A LINNE. Systema naturae per regna tria naturae

secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis.

13th ed.

Gowan, A. 1770. Historia piscium sistens ipsorum anatomen externam, internam, atque genera in

classes et ordines redacta. Histoire des poissons, contenant la description anatomique de

leurs parties externes et internes et le caractére des divers genres rangés par classes et par

ordres. Strasbourg.

HamIiLtTon, H. 1916. Notes on the occurrence of the genus Trachipterus in New Zealand. Trans.

N.Z. Inst. 48: 370-381.

Harrisson, C. M. H. & PALMER, G. 1968. On the neotype of Radiicephalus elongatus Os6rio

with remarks on its biology. Bull. Br. Mus. nat. Hist. (Zool.) 16: 187-211.

IBANEZ, M. & GALLEGO, L. 1974. A new record of Zu cristatus (Trachipteridae Pisces) off the

coast of Blanes (Spain). Vie Milieu 24: 523-526.

JarDAS, I. 1980. Contribution a la connaissance des trachiptéres dans la mer Adriatique. 1.

Trachipterus trachypterus (Gmelin, 1789). Acta adriat. 21: 3-17.

JORDAN, D. S. & Snyper, J. O. 1901. Description of nine new species of fishes contained in

museums of Japan. J. Coll. Sci. imp. Univ. Tokyo 15: 301-311.

JORDAN, D. S. & SNYDER, J. O. 1904. On a collection of fishes made by Mr Alan Owston in the

deep waters of Japan. Smithson. misc. Collns 45: 230-240.

Karrer, C. 1975. Uber Fische aus dem Siidostatlantik (Teil 2). Mitt. zool. Mus. Berl. 51: 63-82.

Karrer, C. 1976. Uber drei mesopelagische Fischarten aus dem Golf von Guinea. Mitt. zool.

Mus. Berl. 52: 177-182.

Kina, J. E. & IkEHARA, I. I. 1956. Some unusual fishes from the central Pacific. Pacif. Sci. 10:

17-24.

LINDBERG, G. U. & LaGeza, M. I. 1965. Ryby Yaponskogo morya i sopredel-’n ykh chastei

Okhotskogo 1 Zheltogo morei. Pt. 2. Teleostomii: XII. Acipenseriformes—XXVIII. Poly-

nemiformes. Opred. Faune SSSR 84: 1-391.

LINNAEuS, C. 1758. Systema Naturae. 10th ed. Holmiae: Laurentii Salvii.

McCann, C. 1953. Ichthyological notes, with species reference to sexual dimorphism in some

New Zealand fishes. Rec. Dom. Mus. Wellington 2 (1): 1-17.

MetaxA, T. 1833. Descrizione di una specie de Gymnetrus. In: Memorie zoologico-mediche,

Roma: 53-57, pl. (fig. 7).

NisHiIMuRA, S. & Hirosaki, Y. 1964. Observations on the swimming behaviour of some

taeniosomous fishes in aquaria and in nature. Publs. Seto mar. biol. Lab. 12: 165-171.

OaiLBy, J. D. 1897. On a Trachypterus from New South Wales. Proc. Linn. Soc. N.S.W. 3:

646-659.

Oxapba, Y. & SuzuKI, K. 1956. Additional note on the rare fish, Trachypterus ijimai Jordan and

Snyder. Annotnes zool. jap. 29: 168-170.

OLIVER, P. 1982. Sobre la aparicidn de algunos peces raros en las islas Baleares. Boln Inst. esp.

Oceanogr. 6: 59-64.

Osorio, B. 1917. Nota sobre algumas especies de peixes que vivem no Atlantico occidental.

Archos Univ. Lisb. 4: 103-131.

PALMER, G. 1961. The dealfishes (Trachipteridae) of the Mediterranean and north-east Atlantic.

Bull. Br. Mus. nat. Hist. (Zool.) 7: 335-351.

PALMER, G. 1973. Trachipteridae. In: HurEAu, J. C. & Monon, Tu., eds. Checklist of the fishes

of the north-eastern Atlantic and of the Mediterranean (CLOFNAM) 1: 330-332. Paris:

UNESCO.

Parin, N. V. 1968. [khtiofauna okeanskoi epipelagiali. Moskva: Akademiya Nauk SSSR. (1970.

Ichthyofauna of the epipelagic zone (English translation). Jerusalem: IPST.)

FAMILIES TRACHIPTERIDAE AND RADIICEPHALIDAE 39

ParIN, N. V. 1978. New records of midwater fishes from off New Guinea and Tonga Islands with

description of two new species of Eustomias (family Melanostomiatidae) and-Benthodesmus

(family Trichiuridae). Trudy Inst. Okeanol. 111: 156-168. (In Russian.)

ParIN, N. V., ANDRIASHEV, A. P. BoropULINA, O. D. & TcHuUvaAsov, V. M. 1974. Midwater

fishes of the south-western Atlantic ocean. Trudy Inst. Okeanol. 98: 70-140.(In Russian.)

Parin, N. V. Sazonov, Y. I. & MIKHAILIN, S. V. 1978. Deep-sea pelagic fishes in the collection

of R/V Fiolent in the Gulf of Guinea and adjacent areas. Trudy Inst. Okeanol. 111: 169-184.

(In Russian. )

Puituipes, W. J. 1944. An immature Trachipterus from French Pass. Rec. Dom. Mus. Wellington

1: 120-122.

Ramsay, E. P. 1881. On a new species of Regalaecus, from Port Jackson. Proc. Linn. Soc.

N.S.W. 5: 631-633.

Risso, A. 1840. Note sur une nouvelle espéce du genre Gymnetrus. Arch. Naturgesch. 6: 13-16.

RosEnBLaTT, R. H. & Butter, J. L. 1977. The ribbonfish genus Desmodema, with the descrip-

tion of a new species (Pisces, Trachipteridae). Fish. Bull. Calif. 75: 843-855.

SmiTH, J. L. B. 1949a. Forty-two fishes new to S. Africa, with notes on others. Ann. Mag. nat.

Hist. (12) 2: 97-111.

SmiTH, J. L. B. 1949b. The sea fishes of southern Africa. Cape Town: Central News Agency.

SmitH, J. L. B. 1953. The sea fishes of southern Africa. 2nd ed. Cape Town: Central News

Agency.

SmiTH, J. L. B. 1956. A new dealfish from South Africa. Ann. Mag. nat. Hist. (12) 9: 449-452.

SmiTH, J. L. B. 1965. New records and descriptions of fishes from South West Africa. Occ. Pap.

Dep. Ichthyol. Rhodes Univ. 3: 13-23.

SmiTH, J. L. B. 1966. Interesting fishes from South Africa. Occ. Pap. Dep. Ichthyol. Rhodes

Univ. 8: 83-94.

TANAKA, S. 1915. Figures and descriptions of the fishes of Japan, including Riukiu Islands, Bonin

Islands, Formosa, Kurile Islands, Korea and Southern Sakhalin. 2nd ed. Tokyo.

TorToNESE, E. 1958. Cattura di Trachypterus cristatus Bon. e note sui Trachypteridae del mare

Ligure. Doriana 2: 1-5.

TorTONESE, E. 1970. Osteichthyes (Pesci ossei), Parte Prima. Fauna Ital. 10: 1-565. Bologna:

Calderini.

WaLtTERS, V. 1963. The trachipterid integument and an hypothesis on its hydrodynamic function.

Copeia 1963: 260-270.

Watters, V. & Fitcu, J. E. 1960. The families and genera of the lampridiform (Allotriognath)

suborder Trachipteroidei. Calif. Fish Game 46: 441-451.

WEBER, M. & BEAuFortT, L. F. DE. 1929. The fishes of the Indo-Australian archipelago.

Anacanthini . . . Heterosomata, Berycomorphi, Percomorphi (Kuhliidae . . . Centropomi-

dae). Leiden: E. J. Brill.

WuitLey, G. 1927. Studies in Ichthyology. No. 1. Rec. Aust. Mus. 15: 289-304.

WuitLey, G. 1968. A check-list of the fishes recorded from the New Zealand region. Aust. Zool.

15: 1-102.

ue

~

|

cb

Cray,

6. SYSTEMATIC papers must conform to the Jnternational code of zoological nomenclature

(particularly Articles 22 and 51).

Names of new taxa, combinations, synonyms, etc., when used for the first time, must be

followed by the appropriate Latin (not English) abbreviation, e.g. gen. nov., sp. nov., comb.

nov., syn. nov., etc.

An author’s name when cited must follow the name of the taxon without intervening

punctuation and not be abbreviated; if the year is added, a comma must separate author’s

name and year. The author’s name (and date, if cited) must be placed in parentheses if a

species or subspecies is transferred from its original genus. The name of a subsequent user of

a scientific name must be separated from the scientific name by a colon.

Synonymy arrangement should be according to chronology of names, i.e. all published

scientific names by which the species previously has been designated are listed in chronological

order, with all references to that name following in chronological order, e.g.:

Family Nuculanidae

Nuculana (Lembulus) bicuspidata (Gould, 1845)

Figs 14-15SA

Nucula (Leda) bicuspidata Gould, 1845: 37.

Leda plicifera A. Adams, 1856: 50.

Laeda bicuspidata Hanley, 1859: 118, pl. 228 (fig. 73). Sowerby, 1871: pl. 2 (fig. 8a—b).

Nucula largillierti Philippi, 1861: 87.

Leda bicuspidata: Nicklés, 1950: 163, fig. 301; 1955: 110. Barnard, 1964: 234, figs 8-9.

Note punctuation in the above example:

comma separates author’s name and year

semicolon separates more than one reference by the same author

full stop separates references by different authors

figures of plates are enclosed in parentheses to distinguish them from text-figures

dash, not comma, separates consecutive numbers

Synonymy arrangement according to chronology of bibliographic references, whereby

the year is placed in front of each entry, and the synonym repeated in full for each entry, is

not acceptable.

In describing new’species, One specimen must be designated as the holotype; other speci-

mens mentioned in the original description are to be designated paratypes; additional material

not regarded as paratypes should be listed separately. The complete data (registration number,

depository, description of specimen, locality, collector, date) of the holotype and paratypes

must be recorded, e.g.:

Holotype

SAM-—A13535 in the South African Museum, Cape Town. Adult female from mid-tide region, King’s Beach

Port Elizabeth (33°51’S 25°39’E), collected by A. ‘Smith, 15 January 1973.

Note standard form of writing South African Museum registration numbers and date.

7. SPECIAL HOUSE RULES

Capital initial letters

(a) The Figures, Maps and Tables of the paper when referred to in the text

’

e.g. *... the Figure depicting C. namacolus ...’; °*. . . in C. namacolus (Fig. 10)...’

(b) The prefixes of prefixed surnames in all languages, when used in the text, if not preceded

by initials or full names

e.g. Du Toit but A.L.du Toit; Von Huene but’ F. von Huene

(c) Scientific names, but not their vernacular derivatives

e.g. Therocephalia, but therocephalian

Punctuation should be loose, omitting all not strictly necessary

Reference to the author should be expressed in the third person

Roman numerals should be converted to arabic, except when forming part of the title of a

book or article, such as

‘Revision of the Crustacea. Part VIII. The Amphipoda.’

Specific name must not stand alone, but be preceded by the generic name or its abbreviation

to initial capital letter, provided the same generic name is used consecutively.

Name of new genus or species is not to be included in the title: it should be included in the

abstract, counter to Recommendation 23 of the Code, to meet the requirements of

Biological Abstracts.

P. C. HEEMSTRA & S. X. KANNEMEYER

THE FAMILIES TRACHIPTERIDAE AND

RADITCEPHALIDAE (PISCES, LAMPRIFORMES)

AND A NEW SPECIES OF ZU

FROM SOUTH AFRICA

VOLUME 94 PART 3 JULY 1984 ISSN. 0303-2515

CAPE TOWN

INSTRUCTIONS TO AUTHORS

|. MATERIAL should be original and not published elsewhere, in whole or in part.

2. LAYOUT should be as follows:

(a) Centred masthead to consist of

Title: informative but concise, without abbreviations and not including the names of new genera or species

Author’s(s’) name(s)

Address(es) of author(s) (institution where work was carried out)

Number of illustrations (figures, enumerated maps and tables, in this order)

(b) Abstract of not more than 200 words, intelligible to the reader without reference to the text

(c) Table of contents giving hierarchy of headings and subheadings

(d) Introduction

(e) Subject-matter of the paper, divided into sections to correspond with those given in table of contents

(f) Summary, if paper is lengthy

(g) Acknowledgements

(h) References

(i) Abbreviations, where these are numerous

3. MANUSCRIPT, to be submitted in triplicate, should be typewritten and neat, double spaced

with 2,5 cm margins all round. First lines of paragraphs should be indented. Tables and a list of

legends for illustrations should be typed separately, their positions indicated in the text. All

pages should be numbered consecutively.

Major headings of the paper are centred capitals; first subheadings are shouldered small

capitals; second subheadings are shouldered italics; third subheadings are indented, shouldered

italics. Further subdivisions should be avoided, as also enumeration (never roman numerals)

of headings and abbreviations.

Footnotes should be avoided unless they are short and essential.

Only generic and specific names should be underlined to indicate italics; all other marking

up should be left to editor and publisher.

4. ILLUSTRATIONS should be reducible to a size not exceeding 12 x 18 cm (19 cm including

legend); the reduction or enlargement required should be indicated; originals larger than

35 x 47 cm should not be submitted; photographs should be rectangular in shape and final

size. A metric scale should appear with all illustrations, otherwise magnification or reduction

should be given in the legend; if the latter, then the final reduction or enlargement should be

taken into consideration.

All illustrations, whether line drawings or photographs, should be termed figures (plates

are not printed; half-tones will appear in their proper place in the text) and numbered in a

single series. Items of composite figures should be designated by capital letters; lettering of

figures is not set in type and should be in lower-case letters.

The number of the figure should be lightly marked in pencil on the back of each illustration.

5. REFERENCES cited in text and synonymies should all be included in the list at the end of

the paper, using the Harvard System (ibid., idem, loc. cit., op. cit. are not acceptable):

(a) Author’s name and year of publication given in text, e.g.:

‘Smith (1969) describes .. .’

‘Smith (1969: 36, fig. 16) describes... .’

‘As described (Smith 1969a, 19696; Jones 1971)’

‘As described (Haughton & Broom 1927)...’

‘As described (Haughton et al. 1927)...’

Note: no comma separating name and year

Dagination indicated by colon, not p.

names of joint authors connected by ampersand

et al. in text for more than two joint authors, but names of all authors given in list of references.

(b) Full references at the end of the paper, arranged alphabetically by names, chronologically

within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year, e.g. Smith (1969a, 19695) and not Smith (1969, 1969a).

For books give title in italics, edition, volume number, place of publication, publisher.

For journal article give title of article, title of journal in italics (abbreviated according to the World list o,

scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses, volume number, part

number (only if independently paged) in parentheses, pagination (first and last pages of article).

Examples (note capitalization and punctuation)

BULLOUGH, W..S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FIscHER, P.—H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris 88: 100-140.

FiscHer, P.-H., DuvAL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des littorines. Archs

Zool. exp. gén. 74: 627-634.

Konn, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee region of Ceylon.

Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 19606. Spawning behaviour, egg masses and larval development in Conus from the Indian Ocean.

Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L. Zoologische

und anthropologische Ergebnisse einer Forschungsreise im westlichen und zentralen Siid-Afrika 4: 269-270.

Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16: 269-270.

(continued inside back cover)

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 94 Band

July 1984 Julie

Part 3 Deel

LOBEDU MATERIAL CULTURE

A COMPARATIVE STUDY

OF THE 1930s AND THE 1970s

By

PATRICIA DAVISON

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town 8000

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na gelang van die

beskikbaarheid van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad 8000

OUT OF PRINT/UIT DRUK

1, AS), 58), SD, 45, 8 teat, SUES, 5, 79).

AG, toa). CED), Be SOD, 7), CCS).

11(@-2;-5, 7, t=p.i.), 15(4=5), 24), 27, 3113), 326), 33, 362). 45@)

Copyright enquiries to the South African Museum

Kopieregnavrae aan die Suid-Afrikaanse Museum

ISBN 0 86813 059 1

Printed in South Africa by In Suid-Afrika gedruk deur

ithe Rusticas Press) Etycjltds: Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

LOBEDU MATERIAL CULTURE

A COMPARATIVE STUDY OF THE 1930s AND THE 1970s

By

PATRICIA DAVISON

South African Museum, Cape Town

(With 119 figures)

[MS accepted 1 November 1983]

ABSTRACT

In this museum and field study a collection made by E. J. Krige and J. D. Krige between

1936 and 1938 is described and compared with data recorded 40 years later by the author.

Within the descriptive framework, material is ordered according to three categories of re-

sources—natural, domestic, and imported. Dependence on the indigenous environment is

shown to have been important in the traditional technology and in understanding the value with

which certain materials were invested. Effects on the material culture of the transition from a

subsistence to a money economy are discussed. It is noted that women have proved both con-

servative and adaptable in changing economic circumstances and that the practice of certain

crafts has increased to supply both domestic needs and an informal market. Conclusions are

drawn regarding the interrelated nature of material and social change.

CONTENTS

PAGE

MEE HO GU CLIO Me hoo ties sek a asd ae Sie eae ee 42

RIS CORIC AOU EMC = oe al aia. euskal eee ea ey age Nea 44

ESV OMIMETID eo hi Se a Sane en are ee aD Pea ee 47

| DYSSgaVOVEd C2) 0) 0 het ee eR san cia tric. cea cinnics cigs ois 49

Socialicontexts ca ay ieta sala Oh UE ee ee Sik

MC Slt OLO ON e eee. aee vce acacyes ony ie eaets ee aha serene eae 53

Wyse ofmaturalsreSOUrCES:., S45. Sl. See ne ee eo es 54

Uy ire ee Ne costed «cute hens tea eee Os a 54

Grasses andiother plant fibies 7. sees nee eee 68

WOO Gran Gime eG ce sone cose eee ee eee A pine Saar 100

SSO TUE eae ane a os A Se ee rt te le 143

Summianysand commento! ss 4 ee se oe eer eee 147

Wsetofidomestic RESOUICES).... =. asi eee ee ee eee 150

GOUndSS Ake Reta oe Jal ee eto he ae et ee oe ane en 150

Skansiandehrdesy 2.200 0. a0. aac See oe ne oe ne See 156

AOR AN GD OMNEH. Ae Skok ee ee a AT Orne Huet ra oor ae 164

IB) RIA Sey 8 ee 8 iit ees cil tea eS eh a ates ea 169

Summanysand comment (ec. yes ee ese oe ee 170

WseotimpontedimatertalS a4. shee acess Soe ee 7A

INCRE AGAR ar oa ef ere rot OO OME tence RACERS. | EN 171

Glassiamidislee Ula ek ee) e n a rR ea 178

COTS Yd ne See aa eH ee a a eT Pee eee ee YA 183

Sa ere a me A ee Be 2 als CORO ate TEE 185

Summany-andicomment 4.25 on shaae eas soem ee 187

41

Ann. S. Afr. Mus. 94 (3), 1984: 41-201, 119 figs.

42 ANNALS OF THE SOUTH AFRICAN MUSEUM

PAGE

Material’cultuneandisocialichangen s.70 oa eee 189

Historical, environmental and social change c. 1880-1980 ... 189

Change imimatenaliculturec, 1940=S80 9 ee ee 192

Conclusions regarding the nature of change ................ U7)

ACKNOWIEGSEMIEIIS Was an yu 5 Oe, Gatch ae ancy EI Cue ae ae 199

IRCIEREHCES nto eS cee Moe ee ne HAE Cee Che ee ares 199

INTRODUCTION

The Lobedu of the northern Transvaal have been the subject of anthropolo-

gical study from the 1930s to the present. Their social structure has been de-

scribed in detail (Krige 1931, 1932, 1938, 1964, 1974, 1975a, 1975b, 1975c, 1981;

Krige & Krige 1943, 1954; Monnig 1961, 1963) and the ethnography has pro-

vided the basis for comparative analytical studies (Leach 1961; Sansom 1974;

Kuper 1975, 1982). Thus far little has been published on their material culture.

The aim of this paper is to describe traditional Lobedu material culture as rep-

resented in a collection (hereafter referred to as the Krige Collection) made be-

tween 1936 and 1938 by social anthropologists E. J. and J. D. Krige, and to

record changes observed 40 years later during fieldwork undertaken by the

author. .

The Krige Collection was made with the purpose of bringing together ob-

jects that represented the full range of the material culture of a single society at

a particular time. It was built up during the course of anthropological research

involving participant observation among the Lobedu people over a continuous

period of nearly 18 months. In 1970 the collection, together with short notes on

each item, was presented to the South African Museum, Cape Town, where it is

now housed. It is primarily on the Krige Collection and notes that the descrip-

tion of Lobedu technology in the 1930s is based.

The collection comprises 248 specimens (SAM-—9671 to 9919) and represents

almost all aspects of Lobedu material culture with the exception of very large

objects and immovable structures. The number of examples of each type of ob-

ject, however, is not large enough to assess the range of variability. The area in

which the collection was made covered most districts of the original Modjadji

Location demarcated in 1892, as well as farms and Crown Land to the north and

the area of the old Medingen mission station. For the purpose of this study, it is

assumed that the artefacts in the collection are representative of the material

culture in the 1930s unless the collector’s notes state otherwise. The impression

gained from the collection is essentially qualitative but the large body of related

data allows the material culture to be viewed within wider social, economic, and

environmental perspectives. The anthropological literature on the Lobedu pro-

vides valuable contextual data for the collection. In addition, E. J. Krige has

made available unpublished photographs taken during the 1930s, and relevant

unpublished data relating to her fieldwork in the 1930s and during the period

1962-1980, and has further assisted the study by providing additional informa-

tion on the situation in the 1930s (Krige 1982: 1-45), which allowed a more

i) es

LOBEDU MATERIAL CULTURE 43

accurate comparison to be made with the 1970s data. Unacknowledged photo-

graphs in the text were taken by the author.

Field research undertaken in September 1973, April 1975, June 1976, and

April 1978 was planned with the aims of recording in detail the technological

processes used by the Lobedu and of comparing their recent material culture

with that represented in the Krige Collection. A total of 3: months was spent in

the research area. It was considered important to collect data at different times

of year in order to observe the productive activities associated with the seasonal

cycle. As far as possible this was done, with the exception of the summer wet

season. While doing fieldwork it was necessary to commute daily between the

Lobedu reserve and Duiwelskloof, some 20 km away, as permission to camp in

the reserve was not granted by the local magistrate. With the guidance of

Simeon Modjadji, a member of the royal family living at the capital and one of

the assistants of Krige & Krige, and by using the Krige Collection as a baseline

record of the material culture, more was achieved in a relatively short period of

fieldwork than would otherwise have been possible. Charts with the vernacular

terms and photographs of all the items in the Krige Collection were used to

record whether traditional artefacts were still currently in use. Frequency of

occurrence was noted and absence of any item was followed up with inquiry

regarding replacement by alternatives and reasons for disuse. The focus of this

investigation was Modyjadji’s village, the capital, which has been continuously

occupied for the past 70 years, as well as smaller and newer villages on the sur-

rounding hills and in the Molototsi valley. Areas beyond the boundaries of the

old Modjadji Location, the Medingen mission area and Gakgapane were not in-

cluded.

Once it had been established that many of the traditional crafts, notably

pottery, basketry and wood-working, were still active, the processes involved

were observed and recorded during the normal work routine of the craftsmen.

This was supplemented by demonstrations of particular techniques no longer in

regular use but still remembered. Where possible all stages in the processes were

photographed in situ and examples of partly made artefacts collected for further

study. Specimens of raw materials were collected, and the source and availabil-

ity noted; preparation of materials, tools used, and techniques employed were

observed in action and confirmed in interviews with the craftsmen. Inquiry was

made into specialization and the transmission of skills from generation to gener-

ation.

In the case of processes that were no longer practised, information was

sought from older people who may have recalled techniques formerly used. No

satisfactory information on iron-smelting was obtained as the complex tech-

nology involved seemed to have been forgotten even by the older generation.

Accounts of skin-working technology were obtained mainly from people who

had had past experience but no longer practised it.

Specimens from Modjadji’s area in the private collections of M. Klapwijk

(now part of the Transvaal Provincial Museum Service collections housed at

44 ANNALS OF THE SOUTH AFRICAN MUSEUM

Eiland in the northern Transvaal) and J. Witt, both of Tzaneen, and in the col-

lection of the National Open Air and Cultural History Museum in Pretoria were

recorded and photographed to supplement the Krige Collection.

The descriptive framework adopted orders the data according to the ma-

terials utilized in the technology. The concept underlying this approach is that

there is an interaction between the resources of the environment, the technical

skills used in exploiting these resources, and the social context in which the tech-

nology operates. Following this framework, objects made of different materials

and the related processes are described in separate sections. In reality, however,

everyday activities integrate artefacts made of clay, wood, grass, gourds, and a

variety of other materials. Cross-references in the text indicate where objects

are used concurrently and where similar processes are applied to different ma-

terials. In conclusion the separate sections are drawn together in a discussion of

change in the material culture.

Simeon Modyjadji acted as interpreter throughout the study. There is no

standard orthography for the Lobedu dialect, which has characteristics of both

Venda and North Sotho. North Sotho is the language taught in schools. In the

ethnographic literature spelling is inconsistent. For example, the bi-labial v in

Lobedu is inconsistently rendered as b or v. The spelling ‘Lobedw’ has been used

as this is the standard rendering in North Sotho (E. O. J. Westphal 1982 pers.

comm.). The vernacular terms have been given in khelovhedu dialect, spelt by

using Venda for the consonants and Sotho for vowels following the advice of

N. J. van Warmelo (1977 pers. comm.).

HISTORICAL OUTLINE

Oral traditions recorded by Krige & Krige (1943: 1-12) trace Lobedu his-

tory to the disintegration of the Karanga state north of the Limpopo River dur-

ing the sixteenth century. The Lobedu royal line is thought to descend from one

of the three great dynasties, Mwene Mutapa, Torwa, and Changamire. In the

late sixteenth or early seventeenth century a break-away group migrated south-

ward across the Limpopo (Alpers 1970: 217) taking with them the knowledge of

rain-making and the sacred glass beads that were among the insignia of

chieftainship. The early Lobedu moved into the Lowveld and established them-

selves as rulers in their present environment by subduing the previous inhabi-

tants, thought to have been scattered groups of Sotho-speaking people. A

period of internecine strife was followed in about 1800 by the accession of the

first woman chief, Modjadji I, who restored political stability and achieved wide-

spread fame and power as a rain-maker.

Lobedu oral traditions have parallels among those of the Venda who

crossed the Limpopo later in the seventeenth century and also trace their ances-

try to the Karanga (Stayt 1931: 12). During the post-migratory period bonds be-

tween Lobedu and Venda were strengthened through trade, marriage alliances

and the Venda practice of coming to Modjadji to supplicate for rain.

LOBEDU MATERIAL CULTURE 45

Notable similarities in language, custom and material culture confirm their past

association.

During the nineteenth century Modjadji’s influence spread to other Sotho-

speaking people living in the Lowveld, including the Letswalo, Kgaga, Thabina,

and Narene. Culturally the Lobedu have much in common with these neigh-

bouring people, as do they with the Phalaborwa people further east (Du Toit

1968; Hammond-Tooke 1981; J. D. Krige 1937). By comparison, they have little

in common with Sotho-speaking people living in the different environment of

the Highveld (Krige & Krige 1943: 306; Hammond-Tooke 1981: 154).

From about 1840 onward successive waves of Tsonga-speaking refugees en-

tered the Lowveld as a result of the social disruption caused between 1835 and

1840 by the invasions of the Ndwandwe warrior Shoshangane in what is now

southern Mozambique, and after 1856 by the wars of succession followed by at-

tenuated raiding, drought, and related famine and disease. Prior to this time a

few Tsonga-speaking groups had been assimilated as Modjadji’s subjects but the

large number of nineteenth-century immigrants who had to settle in the un-

healthy low-lying areas were not absorbed through intermarriage and were re-

garded as outsiders. This exclusion was reflected in the refusal of Modjadji to

allow implements of Tsonga manufacture or origin, such as wooden mortars for

stamping maize and the tambourines used in the maldb6 cult, to be adopted at

the capital (Krige 1931: 233; Mantwa Modjadji 1981 pers. comm.; Krige 1982:

21).

During the second half of the nineteenth century expansion of the Zuid-

Afrikaansche Republiek into Modjadji’s territory initiated a period of conflict

and fundamental change. From as early as 1855 the Volksraad imposed on chiefs

in the Soutpansberg area an annual tax of five head of cattle (or five elephant

tusks, or twenty-five copper rods, or twenty leopard skins) and appointed offi-

cials to exact this tax (Grimsehl 1955: 209). In 1861 Albasini, the official respon-

sible for the Lobedu area, had cause to complain that Modjadji refused to pay

the annual tax. Later that year a commando force took action against her and

confiscated a large number of cattle and small livestock as penalty. In the years

that followed Modjadji remained intransigent and prepared to resist further in-

vasions of her territory (Grimsehl 1955: 205).

In 1866 the Volksraad changed the law regarding taxation so that Africans

could be taxed individually and in 1876 there was a further revision according to

which every male over 20 years of age was required to pay 10 shillings per year,

and, in addition, a hut tax of 10 shillings per year was payable. By the same rul-

ing it was decided that ‘locations’ would be assigned to the more important

chiefs who would be paid a salary and be required to keep law and order, and

supervise the collection of taxes (Grimsehl 1955: 206). Modjadji, however, was

not prepared to co-operate and her indignation was further aroused when in

1886, before the ‘location’ boundaries had been fixed, the Volksraad placed ‘oc-

cupation farmers’ on land already occupied by her subjects. The Lobedu resisted

with arms and by burning down farmers’ houses and raiding cattle. In 1890 a

46 ANNALS OF THE SOUTH AFRICAN MUSEUM

commando force was sent to Modjadji to exact taxes, confiscate arms, and settle

the unrest in the area (for details see Grimsehl 1955: 211-221).

In 1892 the boundaries of the Modjadji Location were fixed despite strong

opposition from Modjadji’s councillors and headmen. The area was less than

one tenth of the area that Modjadji had previously controlled. Discontent over

this dispossession of land continued and eventually led to armed confrontation

in 1894 when General Joubert finally crushed Lobedu resistance and confiscated

about 10 000 head of cattle in retribution (Grimsehl 1955: 236-250).

During these last unsettled decades of the nineteenth century the Berlin

Missionary Society established a station in Modjadji’s area (Reuter 1905: 249).

In 1881 Reuter founded the mission settlement at Medingen and started prosely-

tizing among the Lobedu but with little early success. The drought of 1881-2

was attributed to the presence of the mission and Khashane, a Lobedu evangel-

ist who was treated with great suspicion, was eventually murdered in 1884.

Reuter, however, persevered in his cause, established a school, trained young

men in technical skills, and interceded on behalf of the Lobedu in negotiations

with the Volksraad over their territorial boundaries.

In 1895 Modjadji II followed the tradition of her ancestors in committing

ritual suicide. By the end of her reign her territory had been greatly reduced and

the value system that sanctioned her authority was under pressure.

In the years that followed white encroachment continued. The appropria-

tion of resources undermined local production, as did the importation of mer-

cantile goods. Higher taxes payable in cash and an increasing need for money to

buy commodities forced men to become migrant labourers. By 1920 a railway

line linked the Lowveld to the industrial centres of the Transvaal and facilitated

mobility. Pass Laws, however, which had been introduced as early as 1901, re-

stricted the movement of migrants.

One of the results of the South African War was that many Africans moved

back on to the land that had previously been appropriated. None the less, press-

ure on arable land was high. In 1910 it was reported that ‘In the vicinity of Dui-

welskloof and around Modjadji’s location some farms are so densely populated

with squatters that it is difficult for them to find sufficient land to cultivate with-

out interfering with white settlers’ (Blue Book on Native Affairs 1910: 269). The

introduction of the plough, which allowed maize to be more effectively culti-

vated, only partly offset the pressure on arable land as in many cases, owing to

shortage of draught animals, the plough was not set deep enough to break more

than the top-soil. Furthermore, use of the plough is dependent on good rainfall

and suitable topography, with the result that in dry seasons and in steep areas

hoe-cultivation continued.

Regulations for the dipping of cattle were received with suspicion by the

Lobedu, wary of interference. Public-health measures were introduced, quinine

was freely distributed but malaria remained prevalent in the low-lying areas, es-

pecially in the wet season. The annual recurrence of malaria in the Lowveld was

one of the factors that discouraged white economic development of this area

LOBEDU MATERIAL CULTURE 47

until much later in the 1950s after the large-scale use of pesticides had lowered

the risk of disease. During the 1930s when Krige & Krige carried out their field-

work, the Lobedu area was still relatively isolated and despite external con-

straints a subsistence economy was still viable.

During the years between the mid-1930s and the mid-1970s, which are the

main focus of this study, the Lowveld was opened up to far more intensive white

economic exploitation as a result of the eradication of malaria and improved ac-

cess with the construction of new roads. New employment opportunities were

provided for blacks living in the area and more money came into the local econ-

omies. At the same time population of the rural areas increased following the

implementation of influx control in urban centres. The resultant pressure on

arable land made subsistence agriculture an impossibility for many living in the

Lobedu area and the economy became increasingly monetized.

Also during this period political constraints increased. The Bantu Author-

ities Act of 1951, which provided for the establishment of tribal, regional and

territorial authorities, laid the foundation for the later ‘homeland’ policy. Within

this system the authority of the traditional chiefs, councillors and headmen is

greatly circumscribed by the imposed political structure and appointed officials.

The Lobedu district was included in the North Sotho homeland, Lebowa, which

in 1972 was declared a self-governing area within the Republic of South Africa.

The challenge to traditional values initiated during the late nineteenth cen-

tury gained in intensity during the twentieth century as the authority of the

chiefs was undermined, as education facilities improved and mobility increased.

In 1959 Modjadji III died naturally, having rejected the rite of suicide, and was

succeeded by her daughter, Modjadji IV, who died in 1980 and was in turn suc-

ceeded by her daughter. Modjadji V is represented on the Lebowa Legislative

Assembly by her brother, Legugela.

ENVIRONMENT

Modjadji’s territory is situated in the foothills of the northern extension of

the Drakensberg in the eastern Transvaal, at less than 1 000 m above sea level.

The topography varies from relatively high mountainous areas to low-lying river

valleys and plains. Climatic conditions differ considerably with topographic

variation and, despite mountains and rivers, there are stretches of land with

little or no water. In general the lowlands are less well watered and hotter than

the highlands where the temperatures are seldom extreme. The predominant

type of soil is a fertile reddish loam suitable for raising a variety of grain crops.

Much of the highland area is too steep for ploughing (Hattingh 1975: 60) but

well suited to terraced hoe-cultivation. Clays of different types are found in

many places, mainly in river-banks, and are used extensively in the technology.

The area falls within the summer-rainfall region. The rains normally start in

October and continue to March. On average January and February are the wet-

test months, having falls of about 150 mm. Rain is usually plentiful on the

mountains but less predictable on the plains; in one season there may be good

48 ANNALS OF THE SOUTH AFRICAN MUSEUM

rains in one area and drought in another. This unpredictability of rainfall was a

major source of insecurity in the agricultural economy. By having fields in differ-

ent areas and by planting drought-resistant sorghum as well as maize the risk of

total crop failure was reduced. In times of drought wild foods formed an import-

ant component of the diet.

The natural vegetation is bushveld savannah rich in trees, shrubs, creepers,

and grasses. Most of the area is of the veld type described by Acocks (1975:

27—28) as ‘Lowveld sour bushveld’, which provides poor-quality grazing. The

hillslopes are thickly wooded including, in one place, a unique forest of cycads.

A nineteenth-century traveller (Nelson 1973: 221) described Modjadji’s area as

‘some of the richest, wildest and most romantic scenery to be met with .. .’.

The valleys have open woodland vegetation with an abundance of large trees,

among which the marula and a number of Acacia, Combretum and Ficus species

are prominent. Traditionally the Lobedu made extensive use of these plant re-

-sources both for food and in the manufacture of artefacts. The particular value

of the marula tree is clearly reflected in the material culture and in the diet.

Oral traditions also reflect the importance of natural resources. Kruger

(1936: 91) translated a vernacular account as follows: ‘Now the Lobedu were

looking for a nice country with fertile soil and with ant-hills. They sent the

young men in advance to look for such a country. And indeed . . . they found

the right country having marula-trees and ant-hills. . . .. Thus recorded are the

essential elements of Lobedu subsistence, a natural vegetation rich in useful

trees and wild fruits, edible insects, and fertile arable land.

By the 1940s natural timber resources had been seriously depleted, mainly

through the clearing of land for fields and settlement, and legislation to protect

certain indigenous trees had been introduced. Many of the trees, formerly used

by the Lobedu, were affected and could no longer be cut down without payment

of a fee. Dead-wood, however, continued to be used without restriction both for

fire-wood and for making utensils, and illegal use of wood was not uncommon.

Formerly, when the indigenous forests and woodlands were thickly in-

habited with game, hunting was an important source of meat as well as skins and

horns for domestic use. Ivory, horns and skins were also of considerable econ-

omic importance in external trade through Delagoa Bay. Modjadji’s territory lay

just south of the rich Soutpansberg hunting-grounds known to have been ex-

ploited from the early eighteenth century by Venda and Hlengwe hunters (Wag-

ner 1980: 324). By the mid-nineteenth century the introduction of fire-arms by

white hunters had changed both the nature of hunting and the scale of exploita-

tion.

Modjadji’s territory was undoubtedly rich in game but, according to Krige

(1983 pers. comm.), although the Lobedu hunted for meat, they did not play an

active role in hunting for the ivory trade as did their Venda and Tsonga neigh-

bours. In the mid-nineteenth century ivory was a common medium of exchange

and Modjadji received ivory in tribute from hunters in her territory and from

neighbouring chiefs such as Mawewe (Das Neves 1879: 104). Tribute brought to

LOBEDU MATERIAL CULTURE 49

Modjadji from supplicants for rain ensured that she benefited from the game

trade while not actively participating. In the Lobedu context, tribute in goods

had rain as its reciprocal counterpart. By the 1930s not only had the game popu-

lation been drastically reduced but both hunting and the possession of fire-arms

were restricted by law.

The Lowveld was subject to infestation by a number of insects that threat-

ened both human and animal life. Malaria was prevalent in the low-lying areas,

especially during the wet seasons. Tsetse fly (Glossina sp.) was widespread and a

serious threat to cattle, which were also subject to epidemics such as east-coast

fever and rinderpest. Swarms of locusts threatened crops and wild plant foods;

wood-boring termites attacked structures and utensils, and grain stores had to be

protected against weevils. Not all insects, however, were regarded as pests.

Edible termites (Macrotermes sp.) formed an important part of the diet, which

was also supplemented by grasshoppers.

There are some indications that prior to the alienation of grazing-land and

the rinderpest, herds were relatively large, despite the presence of tsetse fly.

Reuter, reporting to the Holloway Economic Commission in 1930, commented

that when he came to the area in the 1880s ‘the lands were lying idle but the

country was full of cattle’. Not only were cattle more numerous but they were in

better condition because grazing-land was plentiful. It was the rinderpest of

1896, together with lung-sickness and east-coast fever, that destroyed almost all

the cattle in the area. Recovery thereafter was slow and restricted by limitations

on grazing-land and the poor quality of the sourveld pasture. The small number

of approximately 3 000 cattle recorded in the Lobedu reserve in 1936 (Krige

1941: 2) seems to be less related to the effects of tsetse than to the disruption of

grazing-patterns brought about by the fixing of territorial boundaries.

DEMOGRAPHY

The Lobedu are a heterogeneous group of originally unrelated people who

over the centuries have become linked together through a network of social

relations and who now tend to conform to a common cultural pattern. The

Lobedu who trace descent from the original founding group comprise less than

10 per cent of the Lobedu as a whole (Krige & Krige 1943: 85).

Before the territorial boundaries were drawn up in 1892, Modyjadji was ac-

knowledged as chief over the vast area bounded by the Great and Little Letaba

rivers (Krige & Krige 1943: 13). The demarcation of the Modjadji Location

(Fig. 1) in 1892 limited Modjadji’s territory to approximately 17 500 morgen

(179 km?) despite complaints by Lobedu headmen to the Location Commission

that this was insufficient land to support their people (Grimsehl 1955: 228-231).

During the 1930s the Lobedu extended their land to approximately 389 km? by

the purchase of farms to the north and east (Krige & Krige 1956: 12), and in the

1960s trust lands to the north were included under Lobedu jurisdiction.

In 1904—5 (Report of the Transvaal Native Administration) the population

of the original Modjadji Location was 14 434, giving a mean density of

50 ANNALS OF THE SOUTH AFRICAN MUSEUM

MODJADII

fo OES

a

Duiwelskloof

30°15’ 30° 30'

Fig. 1. Original Modjadji Location demarcated in 1892 and BoLobedu magisterial district,

1979:

81 persons/km?. In 1936 the population of the extended territory was estimated

at 33 000 (Krige & Krige 1943: 13) giving a mean density of 85 persons/km?. By

1980, according to the official census for the BoLobedu district, the population

had risen to 116 643, and the average density was well over 200 persons/km?. In

areas of most concentrated settlement the density is considerably higher. The

resultant shortage of arable land has led to an inevitable move from agriculture

to dependence on wage labour.

In the late 1970s over 65 per cent of males were migrant labourers and ab-

sent from the reserve at any one time, compared to 37 per cent in the mid-1930s

(Krige 1981: 151). The 1970 census figures indicate that women outnumber men

LOBEDU MATERIAL CULTURE Sit

in the ratio 3:2, a proportion effectively increased by the absence of men work-

ing outside the district. Relatively few Lobedu women are employed as migrant

labourers but there is a trend for women to work as daily labourers on farms,

tea plantations, and in light industry (Krige 1981: 155).

The traditional settlement pattern was one of dispersed villages situated

mainly on the highlands and hillslopes. A village was composed primarily of an

extended family, the size of which varied considerably with the number of wives,

children, and other dependants. The largest settlement was, and still is, the capi-

tal, mosatha (see Fig. 73A), which in 1975 had a population of over 500 people

living in a concentrated clustering of dwelling-units around the central court-

yard. Most settlements are much smaller, consisting of between two and ten

groups of huts. In the 1930s there were a few larger villages of up to eighty huts

(Krige & Krige 1943: 17). New homesteads tend to be very small, sometimes

having only two huts, and they are sited wherever land is available. Resettle-

ment schemes, planned for the whole area, have been implemented gradually,

causing the division of extended families. Increased mobility is reflected in the

tendency for new settlements to be sited near roads leading out of the area (Hat-

tingh 1975: 64).

SOCIAL CONTEXT

Lobedu social structure has been well documented and analysed by Krige

(Krige & Krige 1943; E. J. Krige 1964, 1975a, 1975b, 1975c, 1981) and re-

analysed by Leach (1961) as well as Kuper (1975, 1982) who suggests that the

Lobedu social order can be understood as an extreme variant of the Sotho-

Tswana system. While sharing many cultural characteristics with other Bantu-

speaking people in southern Africa, the Lobedu have certain distinctive features

or emphases that find their closest parallels among neighbouring Lowveld Sotho

people and the Venda.

Among the Lobedu it is not unusual for women to hold high public office

and, indeed, the last five rulers have been women. Women play significant roles

in the social structure both as sisters (in which capacity they are frequently

leaders in ritual) and as wives (who are important in the agricultural economy).

This importance of women in the social structure is a factor that has encouraged

conservatism and adherence to tradition.

Although descent is patrilineal and residence after marriage viri-local, there

are no corporate, property-controlling patrilineal groups. In practice matrilateral

kin are as important as patrilineal kin. This is best understood with reference to

the preferred form of marriage which is between a man and his mother’s

brother’s daughter, his matrilateral cross-cousin. After a number of generations

this marriage pattern results in a merging of kin groups and an emphasis on bi-

lateral kinship.

An intricate network of royal marriage alliances has operated through time

to extend and consolidate Modjadji’s control over a wide area. From throughout

her area of influence men sent daughters to Modjadji as a form of tribute or to

Sy ANNALS OF THE SOUTH AFRICAN MUSEUM

solicit aid. These girls were accepted as ‘wives’, vhatanoni, and were expected to

remain chaste and to serve Modjadji as they would a husband. Those who did

not remain chaste were given in marriage to headmen or to royal kinsmen, or

they could be ‘placed’ to rule over a district. The network of affinal links so

formed, and implying both rights and obligations, has been an important inte-

grating force within the social and political structure (Krige 1975b: 63-64).

When Modjadji ‘placed’ a wife, no bride-wealth was exchanged but it was ex-

pected that in time a daughter would be sent to the Queen, so ensuring continu-

ity of the affinal link between ruler and subject.

It has been noted that the role of a sister was central to the Lobedu social

system (Krige & Krige 1943: 70). The importance of the sister—brother bond

was related to the pattern of marriage. Before money was used as bride-wealth,

one of the few ways in which a young man could acquire the cattle to contract a

‘marriage was to use his sister’s bride-wealth. As most cattle were tied up in mar-

riage transactions, ideally every brother had a sister whose bride-wealth he

could use in order to marry and establish a ‘house’. This ‘cattle-linked’ sister had

certain rights in the ‘house’ which she had helped to establish. In particular she

had a claim to a daughter of that ‘house’ as a wife for her son, 1.e. as a daughter-in-

law who would cook for her and help her with other domestic tasks, especially

in her old age. This would, in fact, be the preferred cross-cousin marriage,

and the girl would become the chief wife and bear the heir. Should a ‘cattle-

linked’ sister not have a son, she still had a right to a daughter of the ‘house’

which she had helped to found. In this case the sister might ‘marry’ the girl her-

self, using the bride-wealth of one of her own daughters, and appoint a man to

father children in the name of the son that she lacked.

The influence of a sister was often further reinforced by her position as

ritual leader in her brother’s household. In this capacity she would officiate at all

important ceremonies and would be the most appropriate person to intercede

with the ancestors on their behalf.

Young wives (with the exception of vhatanoni) had lower status than sisters.

On marriage the rights to a woman’s children and to her labour were transferred

from her own family to that of her husband. At first a young wife was subordi-

nate to her husband’s mother, cooking at her hearth and using her utensils.

Later, often after the birth of a child or when a younger wife joined the house-

hold, she was given her own mosha or ‘house’. In a polygynous family each wife

and her children formed an independent economic unit.

The subsistence economy was based on agriculture. Three seasons were

recognized: khelemo the time of hoeing (September—January); /etavula, sum-

mer, when the foods were plentiful (February—May) and maria, winter, when

the fields were bare (June-August). Grain provided the staple diet, supple-

mented by vegetable produce and veld foods. Neither milk nor meat was a

major component in the diet.

Division of labour was not rigidly defined. Both men and women worked in

the fields to produce the essential crops of maize, sorghum, and millet. Women

LOBEDU MATERIAL CULTURE 53

also cultivated fields of ground-nuts, beans, gourds, and other vegetable pro-

duce. Children helped with weeding and in protecting the crops from predators.

During the dry winter months both men and women worked on renewing the

homestead and its material contents. As migrant labour increased women took

over many of the tasks formerly done by men thus ensuring some degree of

stability in the productive economy.

Religion centred around belief in the ancestors. A creator, Khuzwane, was

acknowledged but was thought to be remote from everyday life. The ancestors,

on the contrary, were thought to be intimately involved with the daily lives and

fortunes of their descendants. Their displeasure could cause misfortune, ill-

health and, in the extreme, drought. To ensure health, fertility, and prosperity,

it was necessary to approach the ancestors by making offerings of beer or grain

at their shrines or by invoking them through sacred thugula objects, such as heir-

loom beads or hoe-heads. The presence of ancestor shrines, which often took

the form of a small mound of clay in which thugula objects were embedded, is

an aspect of Lobedu religion also found among other people of the Lowveld and

is distinctive to this area (Hammond-Tooke 1981: 153).

Cultural parallels with other Lowveld people are also reflected in the im-

portance of drums, which are sounded on almost all ceremonial occasions, and

in the vyali and vuhwera rituals which are essentially fertility rites, as well as

complex initiation ceremonies (Krige & Krige 1943: 126-140). Four sacred

drums are believed to be mystically associated with the Queen and the welfare

of the people. Formerly they were beaten at the g6mana initiation to appeal for

rain and in thanksgiving for the harvest, practices that have long since fallen into

disuse.

Modjadji was in the powerful position of being able to intercede with the

royal ancestors on behalf of all her subjects. Her ritual power as a rain-maker

was a vital factor in validating the chieftainship. Praised as “Transformer of

Clouds’, she was believed able to make or withhold rain, the life-giving force on

which fertility was dependent. The whole seasonal cycle was regulated by Mod-

jadji, not only through rain-making but in the tasting of the first fruits and the

ritual ‘biting’ of the first termites.

Rain was symbolic of order in nature. By association with rain, cool or

damp places and the materials found there were imbued with positive forces. A

river stone buried at the entrance to a village was thought to ‘cool’ or counteract

the forces of evil. Coolness denoted a state of harmony, conversely ‘heat’ was

perceived as a force leading to disorder (Krige & Krige 1954: 68). These two

concepts are important in understanding the symbolic content of the material

culture and the Lobedu perception of their environment.

TECHNOLOGY

Technology has been simply defined as the complex of learnt behaviour that

gives rise to material culture (Spier 1973: 1). It is behaviour concerned with

54 ANNALS OF THE SOUTH AFRICAN MUSEUM

making things—the materials, tools, skills and processes employed to fulfil ma-

terial needs. It is also adaptive behaviour, the means by which man interacts

with and modifies his environment.

Adaptive strategies, however, do not necessarily correlate directly with

available resources. The environment provides a range of resources but cultural

responses account for selectivity, for methods of exploitation, and for the value

with which certain resources are invested. Ideological as well as economic fac-

tors influence the manufacture and use of artefacts.

In the framework for description adopted here, objects are grouped accord-

ing to the materials of which they are made and three categories of resources are

reflected: natural resources indigenous to the environment, domestic resources

that are cultivated or the product of animal husbandry, and imported materials.

Within this arrangement it is unavoidable that artefacts used concurrently but

made of different materials are described separately. Context and related use,

however, are indicated and illustrated in the photographs.

USE OF NATURAL RESOURCES

CLAY

Sources of clay are well distributed in the Lobedu area. The weathering of

granitized rocks, which predominate in the area, yields clay minerals and quartz.

Through selective transport of these materials by water or wind erosion the

lighter clay minerals may be concentrated locally, as in the case of clay beds in

river-banks. This is the source most frequently utilized by potters. If decomposi-

tion of the rocks takes place on site with no transport, the clay will be mixed

with quartz particles as the latter are virtually inert and resistant to decay. In

this case the clay requires sorting and sifting before use in making pottery.

Muds of rougher texture found generally throughout the area are used for

plastering walls and floors of huts, for building courtyard walls and ledges and

for making sun-dried bricks (see p. 138). Natural earth pigments such as ochres

and chalks occurring in pockets in certain places, are collected, ground into a

fine powder and used for colouring walls (Fig. 2), clay utensils and, in certain

circumstances, the body. Graphite, which occurs less generally in the area, is

also used as a colourant.

Pottery techniques

The manufacture of pottery, described in an earlier paper (Davison &

Hosford 1978), is summarized below.

Among the Lobedu, pottery is always made by women, some of whom are

specialists and make pots for sale as well as for their own use. Most potters learn

how to make pots by watching and following the technique of an experienced

potter. The craft is not necessarily passed on from mother to daughter and the

present generation of young women show little interest in learning the craft.

In general, pottery manufacture can be separated into at least five stages:

digging the clay, preparing it, forming the pot, drying it, and firing. Decorating

LOBEDU MATERIAL CULTURE 55

Fig. 2. Ochre decorations being applied to wall of courtyard, Modjadji’s village, 1973.

>

a 22 y

erty

3 ee

is ieee

Fig. 3. Collecting clay from river-bank, Molototsi valley, 1975.

56 ANNALS OF THE SOUTH AFRICAN MUSEUM

or burnishing often occurs during the drying stage and may be retouched after

firing.

Lobedu potters in the vicinity of the royal village collect clay, ledzoba, from

two deposits, one about a kilometre from the village, the other about three kilo-

metres away. Both clays, mainly kaolinite, are dug from the banks of dry river-

beds (Fig. 3), but the clay from the further source is darker and finer than the

yellowish, coarse clay of the nearer site and needs less preparation for use. The

finer clay requires only kneading and the removal of extraneous matter; the

coarser clay is dried, crushed, and ground to a fine powder between grinding-

stones, covered with water and allowed to stand for a few days before being

worked into a homogeneous mass by being pounded with a wooden pestle on a

concave stone (Fig. 4A). No tempering material or filler is added to alter the

plasticity of the clay.

All Lobedu pots are formed in a similar way. The start is made by placing a

rough ring of clay on a movable support (a potsherd, wooden dish, enamel basin

or plate) (Fig. 4B). From the ring the lower part of the pot is formed by pinch-

ing, scraping and smoothing by hand, adding pieces of clay (Fig. 4C—D) and,

when it is roughly the required shape, by smoothing it with the shell of a seed-

pod from the mothema creeper (Bauhinia kirkii). The pod, thema, is left in

water to become pliable and is then used in a variety of ways, the convex side

for smoothing, -shevola, the concave side for scraping, and the edge for shaping.

At first the thema is used to smooth the outer wall in an upward, diagonal move-

ment while the other hand supports the wall on the inside and presses lightly to

heighten and shape the sides. Small rolls of clay are added to the sides and

worked into the body of the pot. Excess clay from the inner wall is scraped off

and any irregularities in the clay are removed by hand or with an awl. To make

the rim the upper edge is flattened evenly, using the forefinger and the thumb,

and a thin roll of clay is carefully added and smoothed into position. To give a

very smooth finish a piece of wet calf-skin, khekhupuld, is folded over and

moved deftly round the rim. The rim profile is defined by modelling with a

thema.

The pot is covered with a cloth and taken indoors for a day or two until it is

firm enough to be inverted for the filling in of the base. If the pot is to be dec-

orated, designs are incised with an awl before the pot is taken off its support

(Fig. 5A). Red ochre, /edzogu, powdered and mixed with water, is applied to

chosen areas with a finger-tip (Fig. 5B) and burnished with a smooth river stone,

thidél6. On the following day when the pot is drier, graphite, phomd, is applied

in the same way and burnished to a metallic shine. When the pot is firm but still

pliable, it is inverted and removed from the support. The thick edges of the

open base are pinched inward and excess clay is scraped off; small pieces of clay

are added until the opening is completely closed (Fig. 6A). The base is then

smoothed with a thema (Fig. 6B) and patted into shape with a light wooden

beater, khevhad6é. At this stage unevenness may be corrected and finishing-

touches applied to the decoration. Thereafter the pot is set aside in a hut to dry

LOBEDU MATERIAL CULTURE

Fig. 4. First stages in making a pot, Modjadji’s village, 1975. A. Preparing the clay.

_ B. Positioning a rough ring of clay on the base. __C. Starting to build the wall.

D. Adding small pieces of clay.

Sy //

58 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 5. Decorating a pot for drinking beer, Modjadji’s village, 1975.

A. Incising the design. B. Applying ochre to the design.

for at least a week, and often for much longer, until the firing, -tshuba dibidzha,

takes place.

A number of pots are fired together except in the case of very large pots,

which are fired singly. The firing-place is a shallow depression in the ground

sited in a place sheltered from the prevailing wind. Towards sunset the pots are

positioned and supported on small stones. Slow-burning bark is stacked inside

and round the pots; kindling and old thatching-grass is piled on top and ignited.

As the grass burns, more is added until the flames are smothered and the burnt

grass is insulating the coals within. The fire is left to burn overnight. The follow-

ing morning, when the pots are cool, they are taken home. The fired ware is

usually a reddish terra cotta with some blackened patches due to reducing con-

LOBEDU MATERIAL CULTURE 59

Fig. 6. Completing the base of a pot, Modjadji’s village, 1975.

A. Filling in the opening. B. Smoothing the closed base with a

seed-pod.

ditions in parts of the fire. Burnishing may be retouched after firing (Fig. 7).

Breakages in the firing are explained in terms of poor-quality clay or the excess-

ive heat of the fire.

From the available evidence there do not seem to have been any major

changes in pottery technology over the past 40 years. The process used today

accords with Krige’s description for the 1930s (South African Museum corre-

Spondence 3.3.66) and the tools are similar to those collected by Krige in the

1930s and recorded by Lawton (1967: 172) in the 1960s. Lawton’s (1967: 176)

account of a single potter using the method of continuous coiling seems to

represent an exceptional case.

60 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 7. Retouching the burnish after firing, Molototsi valley, 1978.

Types of pottery and uses

Pottery is differentiated primarily according to function which is, in general,

related to the shape, size, and decoration of the vessel.

1. Nkho (Fig. 8)

The nkho is a very large, spherical beer-pot, usually decorated on the

shoulder and round the mouth. In the 1930s these pots were used on social oc-

casions when many people were drinking beer. If a family did not have its own

nkhoé, one was borrowed from kinsmen or neighbours. By the 1970s in many

places the half 44-gallon drum had replaced, or was_used concurrently with, the

nkho for brewing beer on a large scale. In 1975 the beer-hut of the Molokwane

family contained six very large beer-pots partly embedded in the ground, two

metal drums in which the beer had been brewed, and two smaller beer-pots

from which the beer was served. By 1983 the mud-brick hut had been replaced

by a wooden shed but the large beer-pots were still in use.

2. Modzhéha (Fig. 9)

This is a large spherical pot usually decorated with bands of cross-hatching

on the shoulder and incised geometric designs round the mouth; it is used for

carrying and storing water. These pots are still much in use today as water keeps

fresh and cool in the porous earthenware. A lighter tin container is widely used

to carry water, but there is a modzhéha at every home for storing water. The

shapes of the pots from the 1930s and 1970s show great similarity, as does the

decoration.

LOBEDU MATERIAL CULTURE

Fig. 8. Beer-pot, nkhé, Modjadji’s village, 1975.

Fig. 9. Water-pot, modzhéha, Modjadji’s village, 1975.

61

62 ANNALS OF THE SOUTH AFRICAN MUSEUM

3. Thukhwana (Fig. 10)

A fairly large, spherical or elliptical pot from which beer is served into

drinking-vessels. A thukhwana is always decorated, traditionally with incised de-

signs coloured with graphite and red ochre but nowadays also with enamel

paints. Size, shape, and design vary greatly. These and pots for drinking beer,

which are the most ornately decorated pots, are used for entertaining. In the

1930s an elliptical-shaped pot, thukhwana ya lebile (SAM-—9793), said to have

been inspired by the tread and shape of a motor-car tyre, was adopted from

Kgale potters on the highveld and became very popular among the Lobedu

(E. Krige 1977 pers. comm.). Highveld Sotho influence is also seen in the arc

design and the use of stamped lines on some of these pots (e.g. SAM-—9797,

UCT 38/83).

Fig. 10. Pots in washing-area. Beer-pot, thukhwana, left; water-pot, centre; cooking-

pot, right. Modjadji’s village, 1975.

4. Ledzhémeéla (Fig. 11A-C)

This small decorated drinking-vessel is used when offering beer individually

to an honoured guest. The smallest ones are also used to store baby food. Their

shapes vary but they are always decorated with incised, coloured designs. Many

modern examples are flat-based. ;

The ledzhéméla was introduced by Tsonga immigrants, as was the term

used to describe it (Lawton 1967: 177), and it has largely replaced the traditional

calabash beer-vessel, phafha.

LOBEDU MATERIAL CULTURE

PAE : aa

LIK DDO AS

Cc

Fig. 11. Drinking-vessels, madzhdémeéla.

A. SAM-9800, Krige Collection, 1936-8. Tradi-

tional pot for drinking beer. B. SAM-—9801, Krige

Collection, 1936-8. Flat-based pot inspired by the

shape of a jam jar. C. SAM-—10603, collected

1976. Pot of traditional shape and decoration but

with a flat base.

63

64 ANNALS OF THE SOUTH AFRICAN MUSEUM

5. Bidzha ya khelalélé (Fig. 12)

This spherical cooking-pot is made in various sizes. The large ones are used

mainly for cooking porridge and the smaller ones for preparing vegetable relish

or meat. Decoration on these pots is minimal, usually a single line or a band of

cross-hatching on the shoulder and graphite colouring round the rim. They

blacken rapidly on the hearth and become seasoned and non-porous through

use. The simplicity of decoration and lack of red ochre distinguishes cooking-

pots from pots of similar shape not used for cooking.

a

HO we

Fig. 12. Cooking-pots waiting to be washed before use, Modjadji’s village, 1975.

6. Morifhi (Fig. 13A)

This is an open bowl used for cooking certain relishes such as termites and

locusts that do not need to be covered. It is also used as a lid for a cooking-pot

and has many other household uses. A morifhi is seldom decorated with incised

designs. Occasionally graphite is applied to the rim as a finishing-touch. |

7. Kheritswana (Fig. 13C)

|

This fairly small open bowl is used for drinking marula-cider, mokhobe, and |

for serving relish. A kheritswana used for marula-cider is always burnished with

graphite both inside and out. This distinguishes it from a morifhi of similar

Fig. 13. Bowls.

LOBEDU MATERIAL CULTURE

A. Morifhi, SAM-9706. B. Lebéda, SAM-9787

C. Kheritswana, SAM-9788. D. Lesabél6, SAM-9804. (All Krige

Collection, 1936-8.)

65

66 ANNALS OF THE SOUTH AFRICAN MUSEUM

shape. It is often, but not always, decorated with an incised band of hatching or

cross-hatching near the rim. Modern examples are sometimes made with flat

bases.

8. Lebéda (Fig. 13B)

This large open-mouthed pot, roughly half-spherical in shape, was tradition-

ally used for steeping maize prior to stamping and during the beer-making pro-

cess. When woven beer-strainers were used, beer was always strained into a

lebéda (Krige & Krige 1943: pl. 11b) but this is no longer done, as beer is now

brewed on a much larger scale in metal drums. The /ebéda is still made and used

for soaking grain, particularly during and just after the reaping season but its use

has declined since the 1930s as many people no longer produce their own grain.

A lebéda is usually undecorated except for an incised band below the mouth.

9. Lesabélé (Fig. 13D)

This open bowl used:as a wash-basin is always decorated with a broad, in-

cised band of hatching round the rim. White chalk is rubbed into the incisions

and the bowl is burnished with graphite inside and outside. It was customary for

a wife to bring her husband a /esabél6 of warm water for washing in the morn-

ing. In the 1930s Krige recorded that these bowls were rare and had been re-

placed almost entirely by enamel basins. In the 1970s the traditional bowl was

said to be used for ritual purposes especially by mal6bo dancers.

10. Khetsikhiy6 (Fig. 14)

A deep bowl with a heavy, flat base and striated inner walls is used for

grinding tobacco, -sila leféla, into snuff (see Fig. 54). Occasionally there is an in-

cised band near the rim but most examples are undecorated.

Fig. 14. Pot for grinding snuff, khetsikhiyo,

SAM-9802, Krige Collection, 1936-8.

LOBEDU MATERIAL CULTURE 67

Snuff-making is still widely practised and involves drying tobacco leaves and

grinding them to a powder in the khetsikhiy6 with a wooden pestle, kherégélé.

The ground tobacco is passed through a wire gauze sieve or winnowed to re-

move the chaff and is then reground to a fine powder. This is blended with the

ground ash of a special plant, mokhag6 (Amarantus spinosus) to give the snuff

the desired pungency. The mixture is dampened with water to ensure a good

blend and the pot is placed on the hearth to dry the snuff.

11. Roughly made and broken pots

The Queen’s rain medicines are kept in rough earthenware pots, mehago

(Krige & Krige 1943: 273).

Roughly made pots, sometimes unfired, are used for storing seeds, dried to-

bacco or ground-nuts. These pots may represent the efforts of a novice still

learning to make pots or a poorly made pot that the maker does not want to dis-

card.

Pots cracked near the rim are often repaired with beeswax or resin

(Fig. 15). Fine wire or fibre may be bound round the rim to hold the join in pos-

ition.

Potsherds are used in a number of ways. The bowls of large broken pots are

used as basins for washing a baby (Krige & Krige 1943, pl. 5b) or when cleaning

utensils before use. Smaller sherds are useful as containers for mixing ochre,

graphite, or herbs. New pots are built up on flat potsherds. Certain medicines,

especially those burnt to give ashes, are kept on sherds. Small pieces of broken

pottery are embedded in the floor of the courtyard to act as door-mats and to

prevent erosion in drainage areas. Sherds were also embedded in the bases of

threshing-pits in the fields.

Fig. 15. Pot repaired with resin, Modjadji’s village, 1975.

68 ANNALS OF THE SOUTH AFRICAN MUSEUM

12. Pots used in termite traps

In the early summer months traps, involving the use of small pots, are made

to catch dinntwa termites (a very important food source). A hole, mokhobd, is

dug in a termite heap, kheulu, and a pot placed inside. A framework of sticks,

madzwari, is placed over the hole and covered with leafy twigs. A small opening

is left to allow light in near the pot. When the termites swarm they fly toward

the light and get trapped in the pot.

Other uses of clay

Toys

Clay is used by children to model toys. Herd-boys make clay oxen to resem-

ble the long-horned cattle they admire. Small girls make miniature clay imita-

tions of their mothers’ utensils and use them in play-housekeeping which is an

informal training in home-making.

Tuyeres

When iron-working was an active industry, clay was used to form the

nozzles for the bellows and for constructing the walls of the furnace.

Whorl

A ceramic whorl was used to weight the spindle used in spinning wild cotton

(see p. 91):

GRASSES AND OTHER PLANT FIBRES

Formerly the indigenous vegetation provided a very rich source of grasses,

sedges, creepers, barks, and other plant fibres. By the 1980s much of the natural

vegetation had been cleared for cultivation with the result that the plant re-

sources had been reduced and greater distances had to be travelled to find par-

ticular plants. Krige & Krige (1943: 47) commented on the impressive body of

botanical knowledge that the Lobedu put into practice in their craftwork. Crafts-

men had specialized, practical knowledge of the many different plant fibres suit-

able for making string, baskets, and mats. The quality of the material was

carefully selected for the object in mind. Pliability was necessary for strainers,

strength for baskets, and resilience for sleeping-mats. The long-established use-

fulness of certain plants is reflected in the vernacular terminology. The species

of acacia most frequently used for weaving winnowing-baskets is known as mo-

loha, from -loha meaning to weave. Many objects have the same name as the

material used to make them, for example the sleeping-mat, pate, made from the

sedge of the same name. Symbolic associations also influenced the choice of ma-

terials especially in the case of initiation costumes.

Seasonal availability of grasses and sedges restricts their use to certain times

of year. Thatching-grasses may not be cut before Modjadji has given her con-

sent. This restriction reflects the chief’s control over an essential resource but it

LOBEDU MATERIAL CULTURE 69

is also a practical measure ensuring that the grass is not cut until ripe (Krige &

Krige 1943: 47). Ease of collection influences the choice of material. If the ap-

propriate material is inaccessible because of drought or distance from the

source, a suitable substitute may be used. Substitute materials, however, do not

usually have the particular qualities of the original material and this is evident in

the product. The accessibility of cultivated sisal as well as the strength of the

sisal fibre has made it an acceptable substitute for many of the indigenous fibres

formerly used for the making of rope and string.

Basketry techniques

The two main technical categories of basketry, coiled or sewn and woven,

and the variations found in Lobedu basketry are defined briefly below.

In coiled and sewn basketry a coiled foundation is sewn together with one of

a variety of stitches. A needle may be used but is not necessary if the sewing-

fibre is fairly rigid and is used in conjunction with an awl. In simple oversewing

each stitch passes over the foundation coil and pierces the row below. If the

stitch splits the stitch in the previous row a forked effect is produced, this is

known as furcate sewing. Another variation, called bee-skep, has widely spaced

stitching connecting the coils at intervals, so that each stitch is just above the

one below. The foundation material is variable and may consist of a single rod

known as a simple foundation or a cluster of strands known as a multiple founda-

tion. If the material of the coil is too stiff to allow a stitch to pass through it, the

coil is joined by passing a strand through the fabric of the basket to link adjacent

rows of the coil. A sewn technique is used for certain mats and beer-strainers in

which case the work is flat or cylindrical and the sewing-strand passes through

the other element.

In woven basketwork two elements, the warp and weft, cross over and

under each other. The manner of crossing or weaving creates different patterns.

A twill, which is widely used for making winnowing-baskets, is characterized by

the weft passing over and under two or more warps with each successive row

starting one warp ahead of the previous row. Other woven techniques used by

the Lobedu include wicker-work in which the warps are rigid and the wefts flex-

ible (as in a reed-door), and twining in which two wefts pass alternately in front

of and behind each warp (as in sleeping-mats).

In practice both woven and sewn techniques may be used in a single basket,

for example the rim of a woven winnowing-basket is oversewn. When used with-

out qualification, the terms apply to the main fabric of the basket.

Uses of basketry

Baskets

1. Coiled, sewn baskets

The mothagha or kherodo (Fig. 16A) is a large basket for carrying produce

from the fields and for general household use. A coiled technique is used in its

70 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 16. Sewn baskets. A. Mothatha, SAM-9738. B. Khethathana, SAM-9741. (Both

Krige Collection, 1936-8.)

manufacture, and the material most often used for both the coil and the sewing-

strand is the stem of the creeper lebibye (Cocculus hirsutus). To prepare it for

use, the creeper is boiled in water until the outer bark is loose and can be pulled

off easily. While still damp and flexible from soaking in warm water, the stem is

cut through the middle and stripped of the inner pith (Fig. 17). The stems can

then be scraped down to the required thickness for use. For the foundation it

may be left whole or split in half; the wrapping-fibre is pared down into thinner

strands. Lebibye may be prepared well in advance and stored until required.

The mothatha and similar smaller khethathana baskets (Fig. 16B) are made

LOBEDU MATERIAL CULTURE 71

Fig. 17. Women preparing strips of /ebibye. Note pot of water for soften-

ing the fibre and the use of the toes for holding the fibre. Photo: E. J.

Krige, 1936-8.

by women (Fig. 18). The foundation is joined by passing the thinned creeper

over two coils to interlock with the previous row (Fig. 19A). An awl, morogola

(see Fig. 110B), is used to make a space for the creeper to pass through. Dry

lebibye is inflexible and must be dampened with warm water before use, as is

generally the case in much basket-work. The rim of the basket is thickened with

any pliable stick and oversewn in a herring-bone pattern (Fig. 19B—C). A piece

of hide may be sewn onto the base to protect it from wear. The final product is a

very strong durable basket.

a2 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 18. Woman working on a mothatha. She is using an awl to make a

hole for the sewing-strand to pass through. Photo: E. J. Krige, 1936-8.

The smaller domestic basket, khethathana, is generally made in the same

way but not always of the same materials. Bark fibre may be used for the sew-

ing-strand but this is not as durable as /ebibye as it frays when cleaned in the

customary way of scouring with sandstone. If grass is used for the foundation, as

in one example (Fig. 16B), the coil is many stranded and is sewn together, in

this case with bark fibre, using a furcate (Fig. 20) stitch. This basket lacks the

firmness and strength of those made with /ebibye.

LOBEDU MATERIAL CULTURE

Y. BRANCO

Cc

Fig. 19. A. Interlock oversewing-technique. B. Herring-bone

binding. C. Rim attachment to basket.

= 2V2\ oe [NINN

Be

=v: VEO VR

VY. BRANCO

Fig. 20. Furcate sewing-technique.

He:

74 ANNALS OF THE SOUTH AFRICAN MUSEUM

Formerly, when there was a surplus of grain to be stored, men made large,

coiled grain baskets with lids that could be sealed with dung. A kheséh6é was

made with a foundation of lefhe grass sewn together with bark fibre using a bee-

skep stitch (Fig. 21). The rim of the mouth was oversewn with bark fibre in her-

ring-bone stitch.

Fig. 21. Beeskep sewing-technique.

2. Woven baskets

The woven winnowing-fan, lesél6 (Fig. 22A), the smaller but similarly

formed basket, tsana, and the serving-basket, kherodwana (Fig. 22B) were

made, and in the case of the Jesél6 and tsana still are made, by men from thin

strips of wood cut from a branch of a suitable tree, most commonly moloha

(Acacia ataxacantha). Their manufacture is described here and not in a later sec-

tion on the use of wood as, apart from the cutting of the wood, basketry tech-

niques are employed.

In addition to the wood slivers used for both warp and weft elements,

lengths of firm but flexible wood and creeper stem are required for the rim. The

materials are collected and prepared in advance. The strips of wood are pared

with a knife to similar lengths and thickness; the wood for the rim is cut to the

length that will allow it to be bent into a hoop, legdk6, the size of the circumfer-

ence of the rim, and the overlapping ends are secured with a small dowel. The

creeper lebibye is prepared as described above for the coiled baskets.

The craftsman first dampens the wood slivers in warm water, takes four

strips to start and incorporates additional strips by weaving in a straight twill

over two (Fig. 23A) until a square is complete. While weaving the craftsman

supports the work on his lap or on a piece of wood on the ground. The hoop is

tied to the flat woven square in four places, the corners of the square are cut off

and the remaining circular piece is then pushed through so that the hoop is on

the outside. A thin plant stem, khuzwane, is tied on the inside of the rim, which

|

'

|

|

.

/

LOBEDU MATERIAL CULTURE 75

Fig. 22. Woven baskets. A. Winnowing-basket, Jlesélé,

SAM-9679. B. Serving-basket, kherodwana, SAM-9673.

(Both Krige Collection, 1936-8.)

is then oversewn with the creeper binding (Fig. 23B—C). An awl, morogola, is

used to make holes through which the creeper is threaded (Fig. 24A). The ends

are worked into the binding and the finished product is very strong and neat.

Winnowing-baskets collected in the 1970s do not differ from those collected

by Krige in the 1930s and they are used whenever it is necessary to separate

coarse from finer matter. Winnowing, -fhefhéra, is essentially an up-and-down

fanning movement during which lighter matter is separated from heavier. It is a

— ee

——_—_——7;

—_— =

TU

LOBEDU MATERIAL CULTURE

Fig. 24. A. Basket-worker binding the rim of a /esél6. Note use of awl. Mohale’s village, 1973.

B. Relish drying on an old /esél6, Modjadji’s village, 1975.

78 ANNALS OF THE SOUTH AFRICAN MUSEUM

process much used by Lobedu women both in the fields to free the chaff from

the heavier grain (Figs 25-27) and at home when stamping maize into meal or

when they wish to separate dry ingredients.

When maize is stamped by hand the J/esél6 is a necessary complement to the

pestle and mortar in the successive stages of the process that produces meal of

different texture and so adds variety to the staple porridge diet. After the first

stamping, -thovola, the grain is winnowed in an up-down action, -fhefhéra, to

separate the outer husks, which are made to fall to the ground. Thereafter the

basket is moved in a rotating action, -séléla, so that the larger husks come to the

Fig. 25. Woman winnowing sorghum with an up-down movement,

-fhefhéra, to separate the grain from the husks. Photo: E. J. Krige,

1936-8.

LOBEDU MATERIAL CULTURE 79

Fig. 26. Woman holding the winnowing-fan to the wind to separate the

heavier grain from the chaff. Photo: E. J. Krige, 1936-8.

top in the centre. These are removed by hand and kept as food for the pigs; the

de-husked grain is returned to the stamping-block and the process is repeated at

least twice. The husks of the next winnowing are kept for making mabudu, a

light beer; those of the following are used for making maase, a fairly granular

porridge. At this stage the stamped grain, thovole, is placed in a pot, covered

with water and left for a number of hours before being drained on a /esél6. The

water in which it was steeped is kept for making soft porridge, vhothithi, which

has a slightly sour flavour. The grain is then ready for the second stamping-

process, -seda, which results in the fine, white flour used for making the staple

porridge, vhoswa.

80 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 27. Woman separating the coarser grain in the centre of the basket, after winnowing with a

rotating action. Photo: E. J. Krige, 1936-8.

Apart from its use in winnowing, the Jesé/6 can be used as a tray for serving

guests at beer-drinks and for many other general household purposes, including

the drying of relish (Fig. 24B) for use in the dry winter months when fresh

greens are scarce. The smaller tsana has many domestic uses and is especially

useful for holding maize-meal or relish during the preparation of food, and as a

lid for other utensils.

The kherodwana is a waisted basket with a lid, that was used traditionally

for keeping food warm, for serving an honoured person, and also for carrying

goods. The lid, made in the same way as a tsana, could be used on its own as a

small basket. The shape of the basket was designed to fit a wooden dish, delod,

on which porridge was served and set aside in the basket to keep warm until

eaten. The base of the kherodwana was made in the same way as a lesélé and

similar materials, strips of moloha wood and ditsidwa or lebibye creeper, were

used. For the sides of the baskets the ends of a rectangular mat of twilled wood

slivers were woven together to form a cylindrical shape that narrowed at the

lower edge to fit into the rim of the base. The join was oversewn with creeper

binding. The upper edge of the basket was fitted with a ring of wood as for the

rim of a lesél6 and was oversewn. The making of this basket required more

technical skill than other baskets and it had particular social importance.

LOBEDU MATERIAL CULTURE 81

Fig. 28. A. Woman making a twined sleeping-mat. Photo: E. J. Krige, 1936-8. B. Woman

making a twined sleeping-mat. Ramalepe’s village, 1973.

82 ANNALS OF THE SOUTH AFRICAN MUSEUM

In the 1930s every Lobedu bride took a kherodwana to her new home as a

symbol of honour to her husband. It was customary for the bridegroom’s father

to instruct the bride, through his wife, to serve food to the groom in this ‘mar-

riage’ basket (Krige notes). By the 1970s changing social conditions had made

the kherodwana obsolete. This followed changes in the pattern of marriage re-

lated to an increase in migrancy. As it became common for young men to be

away for long periods of migrant labour, many marriages were contracted in the

absence of the bridegroom. In these circumstances there was no need for the

customary use of the kherodwana in marriage ceremonial or in the domestic

sphere.

Mats

Two types of mat were used: pafe, a twined sleeping-mat, and khegdyo, a

_sewn mat for drying meal or for sitting on. Both mats were formerly made only

by men (Krige notes) but as migrancy reduced the number of men in the villages

women took over the task of making sleeping-mats. In the 1930s both men and

women wove mats (Figs 28A, 29), in the 1970s they were made almost exclu-

sively by women (Fig. 28B).

Fig. 29. Man making a twined sleeping-mat. Note use of pegs and pot of water for softening

the fibres. Photo: E. J. Krige, 1936-8.

LOBEDU MATERIAL CULTURE 83

LA EZ

V. BRANCO

Fig. 30. Techniques of finishing edges of mats. A. Whipped edge of sleeping-mat. B. Knotted

edge of twined sleeping-mat. C. Wrapped edge of sewn grain mat.

84 ANNALS OF THE SOUTH AFRICAN MUSEUM

Sleeping-mats are made from a sedge, pate (a species of Cyperaceae). Four

or five lengths of sedge are tied together to form multiple warps, and work pro-

ceeds from the one side or the centre by twining the warps together with string,

the ends of which may be secured with pegs (Fig. 29). The string is usually made

from sisal, mokhdéda, but formerly molete (Triumfetta pilosa) bark was used.

Rows of twining are evenly spaced across the width of the mat. The ends of the

sedge are normally trimmed and finished off with a simple knot (Fig. 30B). The

ends of the string are wound round the last warp a few times and knotted. Warp

ends may be whipped with single-ply twisted sedge (Fig. 30A). The completed

product is a thick and comfortable mat on which to sleep.

Making sleeping-mats on a vertical wooden frame (Fig. 31) is a recent prac-

tice which seems to have been introduced by Tsonga-speakers. Grooves are

made at regular spaces along the horizontal beam of the frame to mark the posi-

tion of the strings that are tied to the first length of sedge. Work proceeds by

_ adding bundles of sedge that are twined together with the strings until a mat of

the required length is made. During the work the strings are weighted with

stones to keep the mat in position and the tension even. This is a much quicker

way of making mats than the traditional method.

The khegdyo mat for drying meal after stamping is no longer in regular use

as most people now buy their maize-meal. It was made from lengths of thin

sedge, khegdyo (Cyperus textilis), sewn together by a strand of twisted molete

bark fibre passed through the sedge with a metal needle, /emao, in a series of

rows. The edges were trimmed and wrapped with string (Fig. 30C). According

to Krige (notes) this type of mat was adopted from Tsonga immigrants.

Fig. 31. New method of making a sleeping-mat on a wooden frame. The twining-threads are

weighted with stones. Molototsi valley, 1973.

LOBEDU MATERIAL CULTURE 85

Wicker doors

A door, khezwala vhatimana, made by men using a wicker basketry tech-

nique, was the traditional door for a cooking-hut. Translated literally, the ver-

nacular term means ‘something by which selfish people keep each other out’, i.e.

to obviate the need for hospitality.

These doors were constructed by making a lattice of thick mbaseriti reeds

and any suitable thinner poles or sticks held together by wrapping (Fig. 32) with

a strong bark fibre such as motswiriri (Bauhinia galpinit).

CCM

D

Cm

bil

( i

aa T

Ye

i

——

y.BRANCO

Fig. 32. Wicker technique formerly used for making doors.

86 ANNALS OF THE SOUTH AFRICAN MUSEUM

Reed doors were not hinged. To close one from the inside a pole was

passed through a thong loop attached to the centre of the door, and across the

door frame; from outside it was held up with a heavy object such as a wooden

pestle. These doors were smeared with cow-dung both to prevent draughts and

to protect them from infestation by weevils. When not in use as a door it could

be used as a large rigid mat on which to place things. Krige records that they

were rare in the 1930s; they were not seen at all in the 1970s.

Strainers

The traditional Lobedu beer-strainer lethéd6 (Fig. 33A) was woven by men

(Fig. 34) from a number of different plant fibres, for example ‘bobbejaanstert’,

Fig. 33. Beer-strainers. A. Woven, SAM-9750. B. Sewn, SAM-9749. (Both Krige Collection,

1936-8.)

kheruje (Vellozia retineris), tsenyane grass (Ophioglossum sp.), or strips of wood

from the molete or moloha-thédo tree (Triumfetta sp.). The same material was

used for both warps and wefts, which were woven together in a twill pattern.

The start was made by knotting a row of strands together (Fig. 35A) and weav-

ing continued in a tubular manner. To close the base the strands were gathered

tightly, the ends knotted and turned inside the strainer.

A sewn strainer lethéd6 la horokiwe (Fig. 33B), said in the 1930s to have

been recently adopted from Tsonga-speakers (Krige notes), was made from the

midrib of the leaf of modolo (Ensete ventricosum) twisted into two-ply string.

The start was made at the open end and the sewing-strand was passed through

LOBEDU MATERIAL CULTURE 87

Fig. 34. Man weaving a strainer while attending a court case in the

khoro. Photo: E. J. Krige, 1936-8.

knots in the string (Fig. 35B) to form a circle. A needle was used for the sewing

and work continued in a spiral with the sewing-strand passing through the string

(Fig. 35C). At the base the strands were sewn into a tight bunch.

Both woven and sewn strainers were used in the same manner. Unstrained

beer was poured through the /ethéd6, which was wrung in the hands with a

twisting action so that the liquid passed through the fibres and the dregs re-

mained behind. The receptacle for the strained beer was a wide-mouthed pot,

lebéda (see p. 66). By the 1970s both types of basketry strainer had been re-

placed by a large wooden sieve, sef6 (see p. 113). This replacement related to an

increase in the scale of brewing beer for sale.

SSSSSSS SIO

SSSESSOSO FGF" /LYHFLFOLE

SS 1 ¥

YSSC raf (CE

VESSELS BSS eS

D

]

LEESO ES

6

SSSeeesG

eae

4

) S

Wea SSS GREP

ie,

Sees EGOOOO

SSS S53ES65 DBS

SSS oo

SSeSoHPO]ES <SSAEGCCSES

co

. C. Sewing-strand

wn Strainer

chniques. A. Diagonal twill weave,

-strainer te

start. B. Start of a se

through two-ply foundation.

LOBEDU MATERIAL CULTURE 89

Playthings

In summer when the grasses are long, boys and girls make a variety of play-

things and ornaments, using basketry techniques. Examples in the Krige Collec-

tion include armbands, a grass peaked cap, képisi ya bosigijane, made in

imitation of a schoolboy’s cap, and a hat, gadiba, also made of bosigijane grass

(Sporobolus pyramidalis). The hats have coiled foundations of bundles of grass

sewn together with single strands of grass. Plaited and woven techniques

(Fig. 36) are used for making armbands that imitate the ornaments worn by

Fig. 36. Plaiting- and weaving-techniques used for

making armbands.

adult women and are referred to by the same names, maséga and khefhddo.

Single lengths of grass are made into toy arrows to be used with a small bow

made of a pliant stick and bark.

Other uses of grass and plant fibres

Brooms

The simplest broom, J/efiélé6 (Fig. 37A), is easily made from a bundle of

lefiél6 grass (Aristida congesta) pulled out by the roots and tied at the grip with

string. Rough brooms for sweeping outside and for cleaning the hearth are made

in the same way from motataile (Arthrixia phylicoides) or lefaladza maru (As-

paragus phimosus). When bosigijane grass is used the broom is made in a differ-

ent way (Fig. 37B) as this grass does not pull out easily by the roots. Lengths of

grass are twined together, folded over, rolled up, and bound at the grip with

90 ANNALS OF THE SOUTH AFRICAN MUSEUM

Manin

RSS

cee

B

Fig. 37. Two types of broom. A. Lefiél6 made of grass pulled out by the roots. B. Grass

twined and bent over before being bound to form grip of broom.

string. Grass brooms are still made by women during the season when the grass

is available, and they are in general household use.

Cordage and spun thread

Lenti is string or cord made from plant fibres. It is made by both men and

women but traditionally it was made mainly by men.

The choice of specific barks, creepers, fibrous leaves or roots depends on

the intended use. The quality of the fibre is taken into account and it is prepared

accordingly. Certain bark and bast fibres, for example moga (Acacia rehman-

niana), produce very strong lashings suitable for structural work; other fibres,

such as molete (Triumfetta pilosa) have pliability suitable for basketry. Strips of

palm or banana leaf make adequate temporary lashings for tying bundles but

would not be durable enough for use in hut-building. Other fibres recorded for

making ropes and string include mokhéba (Dombeya rotundifolia), mothanari

(Colophospermum mopane), motswiriri (Bauhinia galpinii), mofwara tsweni

LOBEDU MATERIAL CULTURE 91

(Grewia sp.), mothala malagani (Acacia karoo), and mokhéda (Sansevieria sp.).

Many barks are cut into strips and used as lashings without further prepara-

tion. If collected in advance, the strips are rolled into balls or bundles and stored

until required. Before use, dry brittle bark-fibre is soaked in water or may be

boiled to strengthen it.

String is made by separating the fibres of the chosen bark or plant and twist-

ing the fibres together by rolling them by hand on the thigh or calf (Fig. 38). The

Fig. 38. Woman rolling sisal string, Modjadji’s village, 1975.

deftly executed rolling action, -hofa, is in two directions, both backward and for-

ward. Two or three lengths of string may be twisted into two- or three-ply cord,

following the same method.

The introduction of cultivated sisal, Agave americana (also mokhéda),

made it more readily available than the indigenous fibres and this, together with

the ease of working with the fibres, has resulted in its replacing most other ma-

terials for making rope and string.

Formerly wild cotton, /egudu (Gossypium herbaceum var. africanum)

(Fig. 39A), was spun by men into thread and plaited into girdles or necklets by

women. Krige records that in the 1930s cotton was spun on spindles

(Fig. 39B—C) weighted with ceramic whorls, kheritswana, and that there was

considerable demand for handspun thread, Jedede or letshida, which was used

for the ties on skin skirts and baby-slings as well as for necklets. By the 1970s

bought thread had replaced handspun cotton, spinning was remembered only by

a few old men, and pieces of handspun thread were treasured as heirlooms.

92 ANNALS OF THE SOUTH AFRICAN MUSEUM

B

Fig. 39. A. Wild cotton growing near Modjadji’s

Nek, 1973. B. Lobedu man spinning cotton.

Photo: E. J. Krige, 1936-8. C. Spindle with

ceramic whorl, SAM-9772, Krige Collection,

1936-8.

LOBEDU MATERIAL CULTURE 93

Headpads

Bundles of grass shaped into the form of a ring and bound roughly with any

plant fibre such as grass, bark, or banana leaf are used to support heavy loads

carried on the head. A headpad, khare, may also be made of a rolled-up piece

of cloth serving the same purpose.

Nests for hens

A nest, khetutu, for hens is made by coiling and sewing bundles of moha-

lana grass (Rhynchelytrum repens) to form a sphere and cutting out an opening

for the entrance. Nests would be placed on a high platform to keep the hens out

of reach of predators at night. The keeping of hens is a Tsonga practice that has

been gradually adopted by the Lobedu but not without some resistance. This

was strongest in the royal village and even today, although many people else-

where in the Lobedu area keep hens, they may not be kept at the capital.

Another kind of hen-house, recorded in the 1930s, was a small, conical

grass structure bound with grass at the apex. A stone was placed over the open-

ing to close the entrance. This was not seen at all in the 1970s.

Thatching

Thatching is a seasonal activity that takes place after June when the grasses

are ready to be picked and before the rains start in late October. Women gather

the grasses in large bundles (Fig. 40A) and men do the thatching. The old way

of doing this was simply to lay the grass on the roof framework of closely packed

rafters, to tie down the first layer of thatch at the lower edge so that the roots

pointed downwards, and thereafter to place the other layers of grass on top with

the roots facing in the opposite direction so that they interlocked with the lowest

layer. Letshumelo grass was used for this kind of thatching (Krige 1982 pers.

comm.). Withies or grass rope may be placed on top to hold down the thatch

more firmly (Fig. 40B).

The modern thatching-technique is to sew the grass on to the rafters with a

long needle, /emao, made of a hard wood such as moréjé (Dicrostachys cinerea).

Work starts at the apex of the roof. Thazi grass (Hyparrhenia filipendula) is con-

sidered the best grass for the first layer as it has a smooth stem that does not

break easily. Thereafter other grasses are placed on top and sewn down. By sus-

pending a small ladder from the roof pinnacle the thatcher can move around on

the roof quite easily (Fig. 41). The sewn thatching-technique is said to have been

adopted from the method used by white farmers.

Vuhwera and vyali costumes

Young men in the vuhwera (North Sotho: bogwéra) initiation, which is held

concurrently with the girls’ vyali (North Sotho: baale) (for full description see

Krige & Krige 1943: 126-140), are taught by older initiated men how to make

94 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 40. A. Woman carrying thatching-grass, Modjadji’s village, 1973. B. Old type of thatched

roof, Modjadji’s village, 1975.

eee ee

nav

als

.

ie Maen

-

ue) vm wa i vy, 2) : :

= aa s & MO a =

7) — 1 fel bb =

) Cc = a)

On ma ~% 6SeSs ©

O ee or |

= ' ’ = ” i _

5 ea Ua Fs _—_——

fas | f, g i =P = =)

f e

~ “A @, | ei o) sh

: LW) uP) : ; | {'

) < & | ~~ aw 8 —|

t cs ry oy | ( of i 4)

5) S gspd oF»

; ~ A, = tent t = ey 8)

af r @ 2 och oO < eI

ep " oom 8 se ©

| 3 Soin #

F | rm fe Ww 4] vy oj =

= ee | Ri ee Gay

~ f fi ' ’

iA be ~ ‘ld (2 wey

y ‘3 S686 6&3 wv

" trl sme sé qj 0

| fl) aN =) | —

' [ ra ( a= © =

~| = ( dent 4 ‘ fa

4 - |) An rey | = ra)

= 4 5s wre a &

; : 5 2 wm ©

iS) a mw Oo. O'T 5

mn = —_ = © iv

| one i ' ‘ DO w=

S Sl = «ew

> Li t =

i BP 7) 4 2 oog 8

7 s bn) = 2 Go

| \ j | WwW 7) * S \ S

i 4 FE "7 © © a o Pa

tae | ( ~ 1s

z | ref ae - ys rm 0 ry

s! a t= = Ee Ss

. - =] wy ras =

iS A q — 5

| Fa) ;

=) ee ee ees =

(1 Bh hey Ni a & nv

B = 8 ERE OS

( e 2 . be as

“ Sek as :

b = ~ = -) ‘fom (J

ey RK) 80 @,

als: em fj - 7s

a. pt 98 °K

A ral — a a

oO 2 oO ed Ja | = Ah) =

- ~ + —

5 a oo) = Da =

c o & a — 6 wd Pa

= © w © -

bee eg S 2s Rk aE

Oot mi. A O mS

=| 9 ¥ g Oe D4 GS

Pe avo Pesos §

& Bog = o> & P<)

ae eae sgue sg

—— '

ta. & oy 3 S

se Bae SE S24S5

oo —_ ‘BI

=

—

a \

-

re =|

oy 2

-=

B's

A 4

YW ©

SaR=

SS

af

c

q

Bee

5 =|

oh

y

to

So,

6.8

96 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 42. Mohwera dancing in courtyard. Photo: E. J. Krige, 1938.

indigenous fibres, there is an awareness among conservative Lobedu of the sym-

bolic importance of the traditional materials.

The following information was given by Simeon Modjadji in 1975. There

are two types of vuhwera costume—a light one, mamegwane, worn when out in

the bush hunting, and a heavier dancing-costume, thégwa. Both have loose

fringed skirts, shoulder-bands and a head-dress but the dancing-costume is far

more elaborate and has row upon row of grass cross-bands forming an immense

chest and shoulder covering on which rests a crested head-dress adorned with

feathers or fur (Fig. 43). Armbands, anklets and, in some cases, underskirts are

made of lengths of grass twined tozether to form flat bands. Thazi grass (Hypar-

rhenia filipendula) which has a smooth glossy stem, is the basic material used for

all parts of the costume except the skirt, which is made from strips of molala

palm leaf. Two-ply fibre string is used to join the pieces of grass together in a

special twining technique referred to as khavhe (Fig. 44B). This kind of twining

is used for the shoulder-bands, mofhaga, the arm- and legbands, zwifhddo, the

head-dress, thohd, and for all other parts of the costume except the skirt, mod-

javhe. This is made by knotting long strips of palm leaf on to a rope girdle that,

when worn, is wrapped round the waist in layers.

Towards the end of the vuhwera the initiates spend most of their time going

about to dance at different villages on the invitation of the headmen, or at the

Se fe eS ea

LOBEDU MATERIAL CULTURE 97

Fig. 43. Elaborate costumes worn by the magdégébya during the 1938 vyali-vuhwera initiation.

Photo> E~J> Krige, 1938:

capital to pay respect to Modjadji. The costumed dancers perform in pairs or

singly, accompanied by drumming and singing of the vyali girls. During the vu-

hwera dancing the skirt flaps rhythmically, but the weight of the entire costume

restricts the dance to slow graceful movements. The wearing of the heavy cos-

tume abrasive to the skin of the wearer is part of the ordeal that initiates must

endure. The overall effect is an imposing spectacle much enjoyed by the on-

lookers who reward their favourite initiates with gifts of ornaments or snuff.

In the 1938 initiation the most spectacular costumes were worn by the vah-

wera of Rabothada whose performance at the capital traditionally closed the ini-

tation. Their dancers, the magdgébya (Fig. 43), had magnificent costumes with

head-dresses surmounted by animal figures, and had underskirts trailing the

ground with dramatic effect. According to Krige (1982: 8) these costumes rep-

resented the finest weaving-skills.

The costumes worn in the vuhwera of 1974, which was the next school held

on a national scale after 1938, displayed a deterioration of weaving-skills. The

elaborately costumed magégdbya did not appear at all as it was said that no one

knew how to make the costumes (Krige 1982: 8). The vuhwera costumes worn in

schools run independently by headmen, but with the permission of Modjadji,

also revealed a decline in craftsmanship.

ANNALS OF THE SOUTH AFRICAN MUSEUM

B

Fig. 44. A. Mohwera dancing to accompaniment of vyali girls, Molototsi valley, 1973.

B. Detail of technique used in mohwera costume.

LOBEDU MATERIAL CULTURE 99

Vyali girls do not wear elaborate costumes at any time during the initiation.

During the early stages they wear only short wraps round the hips. Formerly

these were made of hard, unworked skins but in the 1938 initiation cotton cloth

was substituted (Krige & Krige 1943: 133) and has continued to be used instead

of skin. At a later stage the girls wear bandoliers (Fig. 44A) plaited from Jetate

grass and, according to Krige & Krige (1943: 136), dancing skirts, dali, were

made of a special kind of reed strung together in pieces about 2,5 cm long; they

were used at both the 1938 and 1974 vyali (Krige 1982: 9).

An essential aspect of the vyali and vuhwera is the revealing of digdma,

mysterious figures or objects unknown before initiation. These have not been

seen by the writer but, according to Krige & Krige (1943: 135), in 1938 many of

the digOma were masked figures in costumes made of grass, reeds or leaves.

These masked figures were connected with mummery or verbal formulae, each

with an esoteric meaning.

Khiudogane, the Bird of Muhale, is the greatest géma and ruling spirit of

the vyali. It appears in the moonlight wearing a strange costume consisting of a

conical wooden framework similar to the structure of a hut roof. From this

framework, which rests on the shoulders of the crouching wearer, hang long

strings of blackened fibre. The fibre of the moga tree (Acacia rehmanniana),

which becomes black when soaked in water, was traditionally used but more re-

cently this has been replaced by sisal fibre blackened by rubbing with charcoal.

The knot by which the strings are attached to the framework is the same as that

used to attach the fringes of the initiation skirt to the girdle (Fig. 45). At the

pinnacle, thoho, there is a small sphere of wood coated with black resin in which

red seeds and a feather are embedded. The Bird is believed to come from a pool

in the river and the vyali girls sing its praises even when it does not appear in the

moonlit courtyard. It may communicate its commands at a distance by whistling

through a special instrument, mphenyi, made of a reed pipe stopped at one end

with a membrane from a spider’s web (Krige 1982: 27). The Bird arrives only

when food is plentiful and may be seen as the embodiment of fertility (Krige &

Fig. 45. Technique used for attaching skirt fringe to mohwera costume.

100 ANNALS OF THE SOUTH AFRICAN MUSEUM

Krige 1943: 135, 139-140). The entire vyali, which spans a full agricultural

cycle, is at one level an initiation for girls and at another level a fertility and rain

rite.

For comparative data on the Kgaga bogwéra and baale initiation schools,

which have many elements in common with the Lobedu, see the recent analysis

by Hammond-Tooke (1981: 55-82).

WOOD AND REED

The indigenous environment is rich in trees and traditionally provided a

wide range of woods for carving, for constructing huts, granaries and livestock

enclosures, as well as for fire-wood.

Integrated with a practical knowledge of the natural vegetation, was the Lo-

bedu concept of order in nature as something controlled through the power of

medicine and magic, through the ancestors, through the divinity of the Queen,

and through the mysteries associated with fertility (Krige & Krige 1954: 68).

Certain woods used mainly for medical or magical purposes, for example woods

used for the protective wand buried across the entrance to a village or for the

medicated pegs around the perimeter, could not be brought into the village for

secular use as this would reduce their supernatural potency. The same taboo ap-

plied and still applies today to the trees associated with shrines and the wood

from which charms, dithugula, are made (Krige & Krige 1943: 167). The movil6

(wild medlar, Vangueria infausta), the moludu (white stinkwood, Celtis africana)

and kherale (Gardenia jovistonantis) are widely believed, not only among the

Lobedu, to have magical properties (Palmer & Pitman 1972: 203). Kherale was

one of the woods used for the medicated pegs (National Open Air and Cultural

History Museum, ET 62/7, collected near Duiwelskloof) which were used, to-

gether with medicines, to protect villages from evil spirits. Twigs from the khadi

creeper (Adenia gumnifera), which contains much moisture in its stems, were

used for making brushes for sprinkling ‘cooling’ medicines in rituals of purifica-

tion (Krige & Krige 1943: 274).

Symbolic associations of certain trees are important in understanding their

uses. The vitality, luxuriant growth and abundant fruits of the Celtis africana are

positive qualities that give the tree protective powers. The lush fruits of the mar-

ula (Sclerocarya caffra) and the wild-fig (Ficus sansibarica) are the epitome of

fertility, apart from being edible and rich in essential proteins. An infusion of

Cussonia spicata, a species with swollen underground roots, is used by Lobedu

mothers to make their babies fat and strong. The great wild-fig trees shading

shrines have ‘cooling’ properties, as do the succulent bulbs used in the rite for

removing the impurity or ‘dirt’ (khetshila) of death (Krige & Krige 1954: 69).

The Daja forest in which royal chiefs are buried is held sacred and it is a

serious offence to cut wood there (Krige 1931: 210). This taboo applies also to

Moholwe where the sacred drums are kept and, in general, to all burial sites.

The felling of trees here would cause the anger of the ancestors. Certain mem-

bers of the community who have died unnaturally are buried in wet soil under

LOBEDU MATERIAL CULTURE 101

overhanging trees near rivers, in order to ‘cool’ their ‘heat’, which contaminates

the earth and causes drought. If the trees around these graves are desecrated,

even by accident, a ritual of purification must be performed. These beliefs are

still held even among converts to Christianity as shown by a purification ritual,

phuphutshéla madiba, witnessed in 1973. A schoolteacher, who had accidentally

started a fire that spread to a burial site and burnt many trees, had to sacrifice a

sheep and be sprinkled ritually with ‘cooling’ medicines containing, among other

things, rain-water and the undigested stomach contents of the sheep. A bunch of

khadi twigs was used as a brush. The burnt graves were also treated with medi-

cines. This was necessary to appease the ancestor spirits and prevent them from

sending wind that would dispel the rain clouds.

Although important in understanding the Lobedu concept of their natural

environment, taboos on the use of specific woods did not seriously restrict the

technology as they applied only to a small fraction of the many available woods.

Wood-working techniques

‘For each of the crafts—hut-building, fencing, making stamping-blocks,

sledges, hoe-handles, milk-pails, spoons, platters, winnowing-baskets, drums

.. . different qualities in the wood are regarded as important’ (Krige & Krige

1943: 47). Morididi (leadwood, Combretum imberbe), for example, was known

to yield charcoal of great enough heat for smelting iron, and mbidibidi (Ekeber-

gia capensis) was valued for its soft light wood that was easy to carve. Hard, ter-

mite-resistant woods such as kiaat and mopane were selected for structural work

and, although difficult to carve, for pestles and spoons that would wear down

too quickly if made of softer wood. Even the everyday task of collecting fire-

wood for the hearth required a knowledge of which woods burnt quickly or

slowly, which smoked a lot, and which would produce long-lasting coals.

The choosing of a suitable wood was, in fact, one of the most skilled aspects

of wood-working. The preparation of wood for carving and the actual carving

techniques were, by comparison, less demanding. Wood was seldom seasoned be-

fore use and most objects were carved from the solid without need for joinery.

The traditional tool-kit consisted of an adze, bédwana, a poker, mor6é, and

a number of gouges, mah6r6, of varying size each handled in a particular way

(Fig. 46). Before contact with white traders, iron was obtained through trade

with neighbouring people, notably the Phalaborwa and the Venda, and tools

were forged locally. For many years now scrap metal, motor-car springs and files

have been heated and hammered into blades for tools and hafted in the old way.

More recently modern tools have come into use and axes, saws, chisels, pen-

knives, and drills are used by those who can afford to buy them. The traditional

tools, however, were well designed for their many uses and they have not been

replaced. It is common to find hand-made tools used concurrently with bought

tools (Fig. 47). Modern tools have made the felling of large trees much easier

than in the past when it was necessary to make a fire at the base of the trunk and

gradually cut through the burnt layers of wood.

102 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 46. Woodcarver making a dish, de/6, using a long-handled gouge

with a bent blade. Note the characteristic position in which he is sitting

while supporting the dish against a log of wood and holding it with his

feet. Photo: E. J. Krige, 1936-8.

Traditionally there was some degree of specialization in the making of ob-

jects that required particular skill, but in general most men could make domestic

utensils and handles for tools (Krige & Krige 1943: 50). Even skilled craftsmen

were not full-time specialists and they seldom specialized in making only one

type of object. It was not uncommon, however, for a man or a family to win a

reputation for excelling in making particular objects.

Drums and mortars that required the felling of a tree were usually only

made to order by craftsmen who possessed the necessary tools and expertise.

LOBEDU MATERIAL CULTURE 103

Fig. 47. Traditional and modern tools used by Calvin Ramatswalela, 1973. Left to right:

purchased axe, traditional adze and three gouges, modern chisels, files, and knives.

Smaller objects such as dishes and spoons were more generally made. If a man

had time to carve in excess of his own needs or was commissioned, he would ex-

change wooden objects for grain or sell them at the local market. Seasonal activ-

ity regulated both the practice of the craft and the demand for specific products

(Krige & Krige 1943: 32). At the peak of the agricultural season craft-work

came to a standstill and during the latter part of the dry season rebuilding and

rethatching of huts took priority over other craft activity.

Most wooden utensils had a relatively long life-span and seldom required

replacement unless attacked by termites. In the absence of a developed trading

system there was no stimulus for surplus production and local demand was not

great enough to support full-time specialists.

Since the 1930s, wood-working has become less generally practised. Restric-

tions on the felling of most indigenous trees has limited the legal use of certain

woods and the pole-and-daga method of hut-building is no longer used. The

demand for wooden stamping-utensils declined as the economy became increas-

ingly dependent on purchased maize-meal. Furthermore, migrant labour re-

duced the number of potential craftsmen. The men remaining in the villages

tend to be the unemployed, the elderly and the disabled. Through necessity

some of these men have become woodworkers, carving traditional objects for

sale as well as non-traditional objects to supply new demands deriving from

changing social and economic conditions.

104 ANNALS OF THE SOUTH AFRICAN MUSEUM

Uses of wood and reed

Domestic equipment

1. Dishes

Wooden dishes, didelé (Fig. 48A), or platters made from soft woods such

as marula (Sclerocarya caffra) or mohoo (Ficus sycomorus) were traditionally

used for serving cakes of porridge.

The rough shape was carved with an adze and thereafter hollowed out and

finished off with a gouge, lehérd. The ridges made by this tool are clearly visible

on new dishes but they become smooth through use. The rim usually has a lip

and the base may be thickened or raised to prevent it wearing through from use

B

Fig. 48. A. Wooden dish, del6, SAM-—9675. Underside shows the marks left by the tool,

leh6r6. B. Calabash vessel and wooden imitation, SAM—9686 and SAM-9695. (All Krige

Collection, 1936-8.)

LOBEDU MATERIAL CULTURE 105

on the ground. Repeated washing and scouring with abrasive sandstone gives

used dishes a smooth finish.

Demand for wooden porridge dishes has declined but they are still oc-

casionally made and used. Many families use the calabash khethéba, made by

women, for the same purpose. An interesting wooden vessel in the Krige Collec-

tion is carved to resemble the shape of this type of calabash vessel (Fig. 48B).

2. Milk-pails

Milk did not form an important part of the Lobedu diet and the need for

milk-pails, dikhamél6é (Fig. 49), was, therefore, not great. During fieldwork in

Fig. 49. Herd-boy holding a milk-pail, khamélé. Photo: E. J. Krige,

1936-8.

106 ANNALS OF THE SOUTH AFRICAN MUSEUM

the 1970s pails were not seen being made but, according to woodworkers who

had made them in the past, soft woods such as marula or wild-fig were most suit-

able as they did not easily crack. Other woods mentioned were mbidibidi

(Ekebergia capensis), morula moobyani (Lannea discolor) and mobeda-khamélé

(not identified).

Two types of milk-pail are represented in the Krige Collection. A cup-

shaped pail (Fig. 50) referred to as théhdé-ya-tshwene, literally ‘head of baboon’,

is the older form of pail and is similar to early Venda specimens (Van Warmelo,

in addendum to Curson 1932: 58). This type of pail was rarely seen in the 1930s

but the taller, narrower pails (Fig. 50) were still in use concurrently with pur-

chased buckets (Krige notes).

ans

Fig. 50. Two types of milk-pail: khamél6, SAM-9716; and théhé ya tshwene, SAM-9718.

(Both Krige Collection, 1936-8).

3. Mortars

Woods ranging from soft to very hard are used for mortars, mafudu

(Fig. 51). A soft wood has the advantage of being light to move and easy to

carve but has the disadvantage of wearing down and weathering quickly; hard

wood is difficult to work but more durable and resistant to termites. Woods used

by carvers interviewed in the 1970s included marula, motémé (Adina micro-

cephala) and mohoo. A carver often took his tools into the bush to work nearer

the source of the wood. Once the tree had been felled a block of suitable size

was cut from the trunk with a saw and was carved very roughly with an axe. The

mortar was shaped with an adze and gouge, and the hollow was chiselled out

LOBEDU MATERIAL CULTURE 107

Fig. 51. Woman using a wooden mortar and pestle for stamping maize. Note the winnowing-

baskets used during the stamping-process and the grain spread out to dry in the courtyard.

Molototsi valley, 1975.

with a long metal tool often made from part of a ploughshare. The finishing was

done with a small adze and /ehdr6 gouge. Finally a file or sandpaper was used

on the surface.

Where women still stamp maize by hand, mortars have remained in use and

the porridge made from this meal is much preferred to that made of bought

maize-meal. Furthermore, the different stages in stamping and winnowing pro-

duce meal ranging from a coarse to a fine texture, which gives variety to the

staple diet (see p. 78) and the whole-kernel grain is far more nutritious than the

refined product sold in the shops. None the less, pressure on land has resulted in

most families having to depend on bought maize-meal. The frequency of stamp-

ing maize at home is consequently much reduced.

Although ubiquitous in the 1930s, the free-standing mortar, Jefudu, was in-

troduced into the Lobedu area by Tsonga-speakers. Mortars were widely

adopted but until the 1950s were not allowed to be used inside the capital be-

cause of their Tsonga association (Mantwa Modjadji 1981 pers. comm.). During

the 1930s there was a special area outside the capital where women stamped

maize in these mortars (Fig. 52).

The older method was to stamp grain in a hole in the ground in which a hol-

lowed block of wood was set (Krige 1982: 16). The women worked in kneeling

position. This method was well suited to the stamping of sorghum and millet but

was not as effective for stamping hard maize kernels. The practical advantages

of the /efudu mortar for pounding maize seem to account for its widespread

adoption.

108 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 52. Women stamping maize in wooden mortars outside Modjadji’s village. Note the large

number of mortars in use, reflecting the supply of maize, and the use of winnowing-baskets.

Photo: E. J. Krige, 1936-8.

Fig. 53. Women using wooden pestles to thresh millet in the fields. The hollow in the ground

has potsherds embedded in the base to give it a hard surface. Photo: E. J. Krige, 1936-8.

LOBEDU MATERIAL CULTURE 109

4. Pestles and beaters

Long heavy pestles, mese, are used for stamping maize in a mortar and in

the 1930s were used for threshing grain in hollows in the ground (Fig. 53);

shorter pestles are used for stamping nuts in small mortars and for pounding clay

on a stone before use (see Fig. 4A). Woods specially favoured for this purpose

are mokwal6 (Acacia nigrescens and Acacia polyacantha), mothanari (Colophos-

permum mopane) and morididi (Combretum imberbe). In general a pestle is

made of a heavier wood than the mortar. A large club head is shaped at the

upper end to give it additional weight.

A lighter pestle, kherégélé, used for grinding tobacco into snuff (Fig. 54), is

a straight branch tapering slightly at the grinding-end. It is used with a clay pot

that has a striated inner surface (see Fig. 14 and p. 66).

A short-handled beater, khevhad6 (Fig. 55), with a flat working-surface is

used for compacting floors and ledges of huts and courtyards. Some potters use

this tool to shape the pot when it is leather-hard. A khevhado is usually made in

Fig. 54. Snuff being ground using a pestle, kherégélé, and a pot with inner

striations (see Fig. 14), Modjadji’s village, 1973.

110 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 55. Floor beater, khevhadé6, SAM-9726, Krige

Collection, 1936-8.

hard fine-grained woods, such as the Combretum species, so that it does not

wear down easily.

5. Spoons

Spoons, mafo, used for stirring and serving food are made from a variety of

fairly hard woods. Moataba (Pterocarpus rotundifolius, round-leaf kiaat), is

specially favoured as its branches are usually straight and thin; mosese (Pelto-

phorum africanum) and moila khipala (not identified) are also used but marula

and wild-fig are considered too soft to make durable spoons.

Lobedu spoons (Fig. 56) have almost flat bowls and they are not used as

eating-utensils or for handling liquids. The main use of the larger spoons is to

Fig. 56. Spoons and stirrer stored under eaves, Molototsi valley, 1975.

LOBEDU MATERIAL CULTURE east

stir porridge as it thickens and to serve the stiff porridge from the cooking-pot

into calabashes or wooden dishes in smooth, evenly rounded cakes. The rim

round the bowl of a spoon used for serving porridge prevents the hot porridge

from running over the sides and allows a smooth cake of porridge to be formed

(Krige 1981 pers. comm.). Spoons without rims are used mainly for stirring or

scraping down the sides of the cooking-pot. Smaller spoons are used for prepar-

ing relish in fairly small pots.

Wooden spoons are still made and they are in daily domestic use. Washing

with abrasive sandstone eventually gives the spoon a smooth, seasoned patina.

6. Stirrers

The traditional stirrer or twirler, lefhéth6 (Fig. 57), was made from

motomé (Adina microcephala) wood. Young motémé trees have a central

branch from which shoots radiate out in a whorl. The twirler was cut and

Fig. 57. Traditional stirrer, lefhéth6, SAM-9698, Krige

Collection, 1936-8.

trimmed from the growing tops of young trees, which gave the utensil its typical

form. A newer wire stirrer, lefhéth6 la darada, is now popular. This is made

from a straight rod of wood (usually moataba) with two wire rings attached at

one end (Fig. 56).

A lefthéthd is used in the early stages of cooking porridge to prevent lumps

forming when the meal is added gradually to a pot of boiling water. A woman

rotates the stirrer between her palms so that the porridge is mixed with a rapid

twirling action. When the porridge thickens she changes to the heavier /efo.

7. Stools and headrests

Stools, zwidulo, are made from any tree that has branches of appropriate

natural shape. In the khéré men sit on these seats while discussing village mat-

ters (Fig. 58A). In summer they are usually found under a large shady tree and

in winter or cold evenings round the fire-place in the khdr6d. Smaller stools,

112 ANNALS OF THE SOUTH AFRICAN MUSEUM

B

Fig. 58. A. Men sitting on wooden stools in the khéré of Mambeolo,

made ‘riempie’ chairs, Semosa’s homestead. 1973.

1975. B. Locally

LOBEDU MATERIAL CULTURE 113

carved from logs, are found in cooking-huts. Riempie chairs (Fig. 58B), based

on those seen in farmers’ homes, were made by some carvers in the 1930s and

were still made in the 1970s.

The most common headrest, kheseaméld, is simply a block of wood or a

bundle of old rags placed under the sleeping-mat. Two carved headrests

(Fig. 59)A—B) were collected in the Lobedu area in the 1930s but such headrests

Vz

Fig. 59. Headrests, zwiseamél6. A. SAM-9730.

B. SAM-9731, made by a Tsonga craftsman. (Both

Krige Collection, 1936-8.)

were not widely used. One is known to have been made by a Tsonga craftsman

and present-day Lobedu woodworkers confirmed that carving of this nature was

done mainly by Tsonga speakers.

8. Strainers

A wooden strainer for beer, sef6 ya hothéd6 vyalwa (Fig. 60A-B), is a

fairly recent addition to Lobedu household equipment. It can strain much larger

114 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 60. A. Sieve, sef6, made of a wooden frame with wire-gauze stretched across the base,

used for straining large quantities of beer; clay beer-pot in foreground. Modjadji’s village,

1975. B. Sieve in use at Modjadji’s village, 1975.

quantities of beer than the woven strainer, which it has largely replaced. An ex-

ample (SAM-10881), purchased in 1976 from a craftsman who brought these

strainers to sell at the market at the capital, is made of pine and blue-gum wood.

Pine is used for the sides of the box-funnel and gum for the lateral supports and

handle. The pieces of wood are joined with corrugated fasteners and nails, and

strips of metal are nailed over the corners and side joins to give them extra

strength. Wire-gauze and a lattice of pieces of hoop metal are nailed across the

base opening to form the sieve through which the beer is poured.

This type of strainer is a modern innovation related to an increase in the

brewing of beer for sale. In the 1930s the selling of beer was very uncommon

and beer was generally brewed on a relatively small scale, preferably from sorg-

hum or millet. The woven strainer (see p. 86) was used by wringing and squeez-

ing the beer into an open-mouthed pot, lebéda (see p. 66). Beer was consumed

at home and in a number of socially important contexts such as paying tribute,

showing hospitality or gratitude, and honouring ritual obligations. This pattern

of beer-drinking has continued to the present but, in addition, women now brew

LOBEDU MATERIAL CULTURE 1S

beer for sale as it is a way to earn money at home. Maize-meal is bought at the

store, often on credit against later payment when the beer is sold, brewed in

large tin drums, strained through a wire-mesh sefo into another large metal

drum, and served in glass jars or mugs of fixed volume. Thus the change in the

pattern of brewing is reflected in the utensils used in the preparation and con-

sumption of beer.

Hunting-equipment and weapons

Objects having wood or reed as a component are described here and refer-

red to in other sections where relevant.

1. Arrows

Arrows (Fig. 61A—G) are distinguished according to their type of head:

mosévhé wa phagane is an arrow with a flat, unbarbed metal head; mosévhé wa

mogovho is an arrow with a long, straight metal head; mosévhé wa digoébe (lit-

erally a fishing-arrow) is an arrow with a barbed metal head; and thebola is an

arrow with a knobbed wooden head (Fig. 61F-—G).

Arrow shafts were generally made of reed into which the tang of the head

was fitted and the join bound with sinew, bast or stripped creeper. A hard wood

such as mosese (Peltophorum africanum) was used for the knobbed head of the

thebola, which was used in hunting birds and small buck.

2. Bows

Formerly the bow, vhora, was the principal weapon in warfare and in hunt-

ing. Both these activities had ceased by the 1930s.

Mofwata (Brachylaena transvaalensis), which was considered a particularly

good wood for making bows (Krige notes), is described by Palmer & Pitman

(1972: 2149) as ‘strong, elastic and durable’—ideal qualities for the purpose.

Examples in the Krige Collection have staves, rounded in section, which taper

from the grip to the slightly shouldered ends. The string of twisted thong is

fastened to each end with a knot (Fig. 62).

By rubbing the stave with fat it was kept pliable, and in cases when repair

was necessary the break or crack was bound with wire and covered with a piece

of cow-hide (SAM-—9836, Krige Collection).

3. Clubs

Wooden clubs were not part of Lobedu material culture. The term théga, a

club, is a Lobedu rendering of the Tsonga term nhonga, and the weapon was

never widely adopted by the Lobedu (Krige 1983 pers. comm.).

4. Hafts of tools and weapons

Weapons and tools used for a variety of purposes were hafted in wood.

Fairly hard woods were selected for this use as it was important that the wood

should not crack easily. Moataba (Pterocarpus rotundifolius) and mordéd6é

116

ANNALS OF THE SOUTH AFRICAN MUSEUM

Wy BLA

[

(

1K |

LATER

|

e /)

| ALTA AN:

Beene TA MMOL Le

Lf

=

f}

AFA,

}

aa

V. BRANCO

cm

Fig. 61. Types of arrows. A. Unbarbed arrow, mosévhé wa phagane.

B, E. Barbed arrows, mosévhé wa digdbe. C, D. Long-tanged arrows,

mosévhé wa mogovhd. F. Wooden-headed arrow, thebola. G. Butt end

of the thebola.

LOBEDU MATERIAL CULTURE LT;

Fig. 62. Detail of bow-string attachment.

(Pterocarpus angolensis) were often used but many other woods were also suit-

able. The handle, mofhenye, was carved to the required shape and a socket for

the attachment of the head or blade was burnt through the haft with a hot

poker, moro (see Fig. 110A). When a man needed a new tool or weapon he

would commission a specialist smith for the head and make the haft himself.

Hafts were generally undecorated but usually well proportioned in relation

to the head and well balanced for handling. Occasionally a knife, mofhaga, had

a band of decorative carving on the hilt (see Fig. 111A).

> fraps

A cane-rat trap, mothéma (Fig. 63A), was made from a branch of moloha

(Acacia ataxacantha) split into a number of sections for most of its length. The

thicker end was left whole and the split sections were spread apart and fixed in

position by being tied to graded wooden rings (made of pliable wood or creeper

stem). The overall shape was that of a funnel. String was wrapped round the

sticks and passed from one to the other to fill in the gaps that were large enough

for a rat to pass through. Herd-boys placed the trap in the grass where rats,

magwelele, were known to run, beat the bush to scare the rats into the trap and

then killed them with a knife or a small spear.

118 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 63. Traps for cane rats. A. Mothéma, SAM-9862, Krige Collection, 1936-8.

B. Wire mothéma, Modjadji’s village, 1973.

A modern version of the mothéma is made in wire (Fig. 63B). The opening

of the trap has a flap that opens inward only so that the animal gets caught in-

side and cannot escape.

A trap, khethédélé (Fig. 64), for birds consisted of a ring of wood (a pliable

stick bent into a circle and tied with bark) with a mesh of bark strips tied across

it. Slip-knots made of the tail hair of an ox were attached to the cross-strips and

the trap was covered with a thin layer of soil so that only the knots of hair were

LOBEDU MATERIAL CULTURE 119

Fig. 64. Khethédélé, trap with slip-knots for catching birds,

SAM-9863, Krige Collection, 1936-8.

visible above ground. Grain or seed sprinkled between the knots attracted birds

and their legs were caught in the slip-knots.

The mothéma and khethédélé, as well as a number of other traps that in-

volved the use of stones and string or pliable wood under tension, were made by

herd-boys who spent much time in the veld looking for edible wild fruits or

honey and making traps for birds and small animals. They were much less com-

mon in the 1970s than in the 1930s.

The trap for termites made in the side of a termite-heap using sticks and a

clay pot was described in an earlier section (see p. 68).

Musical instruments

1. Drums

Drums played a significant role in public life and were played on many oc-

casions of ritual and social importance. Drums were never individually owned.

Every district head owned a set of drums and was responsible for their upkeep.

When a new district head was appointed the chief handed over a drumstick as a

public sign of his office (Krige & Krige 1943: 126).

Lobedu drums for secular use are of two types: géma (Fig. 65A), a large

drum with hemispherical resonator, played with a drumstick; and thithimedzh6é

(Fig. 65B), a long, cylindrical drum made in various sizes, played with the palms

of the hands. Drums are made by specialists who take orders and receive pay-

ment for their work. In the 1930s in the Venda area the cost of a large drum of

similar type to the Lobedu gdéma, was an ox or its equivalent (Kirby 1934: 36).

120 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 65. A. Large géma and small thithimedzhé drums at headman Mohale’s village, 1975.

B. Large thithimedzhé drums on the verandah of a sleeping-hut in the special part of the

capital used by Modjadji and her wives, 1976.

LOBEDU MATERIAL CULTURE 121

The cost in 1976 was in the order of R20 to R30 but less if the wood was sup-

plied by the buyer. For making drums soft-wooded trees such as mohoo (Ficus

sycomorus), movhamba ngoma (Albizia gummifera) and marula (Sclerocarya

caffra) are used. The carving-tools are adzes and gouges of various sizes, and a

long, flat piece of sharpened scrap metal is used as a chisel.

A goma drum is made in the following way: the resonator is carved from a

solid section of tree trunk and worked so that the long grain of the wood runs

laterally across the drum and not vertically down its length. The resonator is hol-

lowed out until the walls are about 2—3 cm thick, and a small hole is made in the

centre of the base, which is thicker than the sides. A band of relief carving

usually encircles the bowl at its widest diameter. The goma in the Krige Collec-

tion has handles resembling interlaced loops carved at the four ‘corners’ and a

band of carving between the handles. These features are characteristic of many

drums from the Lobedu area and are also typical of the Venda ngéma described

in detail by Kirby (1934: 34-38). Holes round the mouth of the resonator are

burnt with a hot poker and a piece of wet cow-hide is stretched over the head

and pegged into position through the holes. The hair is removed from a circular

patch in the centre of the skin. An additional strip of hide with slits made in

Fig. 66. Drums being played by women at a gdésha dance. The men on the right are playing

reed pipes (see Fig. 70). Photo: E. J. Krige, 1936-8.

122 ANNALS OF THE SOUTH AFRICAN MUSEUM

position of the pegs is fastened round the rim to strengthen the attachment. Be-

fore the wet skin is stretched over the resonator a few river stones are placed in-

side the drum. According to Krige & Krige (1943: 221) the reason for this may

be related to the ‘cooling’ quality of the stones. Coolness is perceived as a posi-

tive quality which can counteract negative forces. It is also likely that the

conventional attributes of the gOma and the manner of beating it have special

meanings and symbolic associations in specific social contexts and that many of

these are related to chieftainship and to fertility.

Two or three cylindrical drums of different sizes are played together with

the gdma. The longest and most important of these is gaedis6, which sounds the

entry of solo dancers; the other two are both called thithimedzho. All are played

by women using the palms of their hands. They are not played in upright posi-

tion but are placed on their sides and tilted upwards when being played

(Figs 66-67).

The construction method for both gaedis6 and thithimedzho is the same.

The wood said to be best for these drums is mordd6é (Pterocarpus angolensis)

but softer woods are also used (Jan Lebiya 1976 pers. comm.). The resonator is

carved from a solid block of wood, the length of resonator foilowing the grain.

The tools used are the traditional adzes and gouges as well as a piece of sharp-

ened scrap-iron used to chisel out the shell. Some, but not all of these drums,

Fig. 67. Women playing drums at a Saturday beer-drink near Molabisana’s shop, 1973. Note

the tin g6ma on the right.

LOBEDU MATERIAL CULTURE 123

have a handle on one side. Holes round the mouth of the resonator are made

with a hot poker, and pegs for attaching the skin are carved in a hard wood. As

for the goma drum a piece of wet cow-hide is stretched over the head and verti-

cal slits are made to coincide with the positions of the holes for the pegs. A band

of hide with horizontal slits in the position of the pegs is wrapped around the

head and the pegs hammered into the rim through the slits in the two layers of

skin. These drums are seldom decorated.

A frame-drum, khethadada (Fig. 68), played during maldb6 spirit-possession

rituals is made in the following way. A piece of wet goatskin, with the hair re-

moved, is stretched over a broad hoop of pliable wood or metal (for example

the frame of a bicycle wheel) and the overlapping edges of skin are cut into

strips, twisted and tied on the underside to fasten the head in position and form

the grip for the player to hold. Two or more khethadada ‘tambourines’, beaten

with sticks, are played together with tshélé and mathodzi rattles (see p. 153) and

other drums to accompany the dancing of a possessed person and to encourage

the spirit to manifest itself (Krige & Krige 1943: 243). The Tsonga cult of

Fig. 68. Frame drum, khethadada, and drumstick used to accom-

pany maldbé dancing. Note attachment of skin on underside.

SAM-9827, Krige Collection, 1936-8.

124 ANNALS OF THE SOUTH AFRICAN MUSEUM

possession from which Lobedu maldbo practices derive has been described by

Junod (1927: 479-504) and the associated musical instruments by Kirby (1934:

41-44). The cult among the Kgaga has recently been described by Hammond-

Tooke (1981: 104-111).

In addition to the drums for secular use there are four sacred drums (dig6-

mana) that are held in great awe and are associated with the chief and the wel-

fare of the people. The following information was recorded by Krige & Krige

(1943: 126-127). The sacred drums are similar in shape to the ordinary g6ma

drum but they are played by hand instead of with a drumstick and their sound is

believed to please the ancestral spirits. The largest of the four, phataje, stands

about 120 cm high, next in size is tanga, then pekahare, and rangwedi, the

smallest but most important. All four drums are said to contain a human skull

instead of the stone that is usually put inside a drum. In addition, rangwedi has

strips of human facial skin placed under the ox-hide head and is smeared on the

outside with khetshila, ‘dirt’ from the corpse of a person of noble blood who was

killed for this purpose. The maintenance of the sacred drums, which involves

treating them with fat and ochre to protect them from infestation by insects, the

renewal of the skins, and the repair of the hut in which they are kept is the

special task of the Rabothada family, which has historical links with Phalaborwa

where the institution of sacred drums was strongly developed (Krige & Krige

1943: 127).

The sacred drums were played only on certain occasions of ritual import-

ance, for example the gOmana ceremonies, and unlike the ordinary drums they

were always played by men. Their beating was associated with asking the ances-

tors for rain and with the seasonal cycle. So important were these drums that

they were looked upon as having supernatural power (Krige & Krige 1943: 127).

Although they are still kept in good repair, the sacred drums of the royal family

have not been beaten regularly since the last decade of the nineteenth century

and not at all since before the 1930s. The reason for this is said to be that killing

for ritual purposes has long been prohibited by law and the necessary parts of a

human body are thus no longer obtainable (Krige 1983 pers. comm.).

2. Stringed instruments

Kirby (1934: 196) grouped southern African stringed instruments into three

general categories according to the fundamental tones produced and the har-

monic sounds generated by these fundamentals. The Lobedu stringed instru-

ments khekhaba, khedzhél6 and kheddlédélé (Fig. 69A—-B) fall into Kirby’s

groups one, two, and three respectively.

The khekhaba is a musical bow in which the fundamental tone of the string

and its harmonics are sounded as a chord. As the string is tied back near the

centre it yields two chords, depending on which part of the string is struck. The

bow is made of a well-seasoned branch with the bark removed. It is fitted with a

string (of wire, twisted sinew or hair) tied back by a loop of wire, sinew or fibre

near the centre. At this point a calabash resonator is attached to the bow. The

LOBEDU MATERIAL CULTURE 125

Fig. 69. Stringed instruments. A. Khedzhdl6, SAM-9819. B. Kheddléddél6, SAM-9821.

(Both Krige Collection, 1936-8.)

string is struck with a thin stick or reed. It is usually played by unmarried men

on its Own as an accompaniment to a song (Kirby 1934: 207).

The khedzholo (Fig. 69A) is an instrument that is played so that certain

harmonics of the string are isolated to produce a melody. The stave is usually

carved from a branch of soft wood about 80—90 cm long and 6 cm in diameter.

The bark is removed, a flat surface formed along one side of the wood, and the

pith hollowed out, leaving one end solid. Through this end a hole is bored and a

tapered peg fitted. A wire is attached to the upper end of the tuning-peg and to

the opposite end of the wooden stave or sounding-board. A notch on the stave

makes it easier to hold the khedzholé steady when in use. A small friction-bow

for playing the instrument consists of a stick about 10 cm long with a string of

tail hairs or plant fibre attached to notches at the ends of the stick.

The player holds the khedzhdl6 with the open end resting on his left

shoulder. A calabash or tin resonator is often placed over the end to increase

the sound produced by stroking the wire with the bow held in the right hand,

and varying the tension of the string with the left hand. Only men play this in-

strument, either alone or to accompany a song.

The khed6éléd6l6 (Fig. 69B) is a musical bow in which the harmonics of the

string are used in conjunction with their fundamentals to produce the melody.

The bow stave, which is always thicker in the centre than at the ends, is con-

structed in one of two ways—it may be carved as a whole from a solid branch of

flexible wood or the central part and the ends may be carved separately and then

fitted together. In the latter case it is usual for two different kinds of wood to be

used—a fairly soft wood for the thick section and pliable wood for the ends. In

both cases the ends are bent to the required curve while green and allowed to

dry in that position.

126 ANNALS OF THE SOUTH AFRICAN MUSEUM

The string is tied to notches cut at the tips of the stave and a piece of cotton

or sinew is attached near the centre of the string to draw it back towards the

stave. It is wound round the stave to secure it in position. The mouth, placed at

one end of the thick part of the stave, acts as a resonator and the string is

plucked in various ways with a plectrum (a small stick or a thorn) to produce a

number of tunes, often wistful or sad. It is played only by men or boys and

sometimes the player sings an accompaniment. Formerly, if a young unmarried

man played this instrument all night people knew he was longing for a wife

(Krige notes).

3. Wind instruments

The playing of the reed pipe ensemble, mothavha, accompanied by drum-

ming and the performance of the gésha dance, was an important form of social

activity and musical entertainment among the Lobedu. Although by the 1970s

_ the inter-district gésha dances associated with individual reciprocities had disap-

peared, those associated with dancing for rain at the capital or at the end of a

period of mourning continued, albeit less frequently than in the 1930s (Krige

1982: 23). The number of reed pipes in a set can vary from more than twenty to

less than ten. Each man plays a single pipe that has a fixed pitch and produces a

single note, the longest pipe giving the deepest note. Each pipe has its own

name denoting the part it plays within the ensemble and how it combines with

other members of the set.

A set of twenty-two reed pipes (SAM-—9812) collected in the 1930s is named

and graded as follows (Krige notes):

1. Ledzekha la pele—the longest pipe.

2. Lembo la dzusi la giba—played after pipe 14 to introduce a song, giba.

3. Ledzekha la gamorao—played with 1.

4. Ledzekha—played with 1 and 3.

5. Lembo la dzusi la vesa—played with 2.

6. Lembo—played with 2 and 5.

7. Ya ga ntau ha thakhudi—played after 12.

8. Lembo—played with 2, 5 and 6.

9. Moholo a khomo.

10. Khomo—played with 9.

11. Khomo.

12. Thakhudi.

13 2Giba:

14. Dzusi ya giba—awakens or heralds 13, starts the giba song.

15. Tateledi.

16. Dzusi.

17. Dzusi ya vesa—introduces the vesa song.

18. Pegudi.

19. Thezwane.

LOBEDU MATERIAL CULTURE 127

20. Ya ga ntau a thezwane—played after 19.

21. Thezwane.

22. Khenthé kha dzusi a giba—high-pitched pipe for ‘rousing’ or introduc-

ing 13.

Some of the pipes (Fig. 70) are made of cultivated reeds, mothékhathébye,

which are stronger than the indigenous mbaseriti reeds formerly used for reed

(\

aU

XX

(\

UvAvaNal

om

LOOK,

A 2

PS

Fig. 70. Some examples of the pipes in a set of twenty-two reed pipes, mothavha,

SAM-9812, Krige Collection, 1936-8.

128 ANNALS OF THE SOUTH AFRICAN MUSEUM

pipes. Not all pipes in the set are made of this type of reed; nine are of a thinner

reed, not identified. Lobedu reed pipes are made in the same way as the Venda

pipes described in detail by Kirby (1934: 162). The reeds are cut to the required

length so that a node forms a natural plug at the lower end of the pipe. To bore

through intervening nodes a length of red-hot iron is used. If ornamented, the

pipes are incised with a sharp knife. The pipes are tuned against each other and

can be cut down to raise the pitch. A complete set of pipes is usually kept in a

special skin bag, saga ya mothavha, and many of the pipes bear some distin-

guishing mark so that the player can easily recognize the pipe that he usually

plays.

Reed pipes are played only by men, who dance in a circle while playing,

causing the melodic illusion of changing sound sequences that would not occur

were the players to remain stationary. The dance-master, malogwani, signals the

start and changes in the pattern of the dance but in actual performance leader-

_ ship changes continually as one dancer after the other takes the centre position

in the circle of dancers. An accompaniment of drumming is played by women on

a goma and three thithimedzo drums (see Fig. 66).

In addition to the reed-pipe ensemble, there are a number of flutes and

whistles made of bamboo or wood that are played on their own, especially by

herd-boys. The term ‘flute’ is used here for pipes that can produce more than

one note, and ‘whistle’ for those that produce only a single note.

The khethodigoé (Fig. 71A) is a transverse flute made of indigenous mbase-

riti reed closed naturally at both ends by nodes in the reed. The embouchure

near one end and three finger-holes near the other end are bored through the

reed with a red-hot poker. A number of tunes can be played on this flute. The

rodigo flute (Fig. 71B) is made from a length of mbaseriti reed open at both

ends. The player blows over two V-shaped notches at one end and varies the

tune by fingering the other end. The /engwale flute (Fig. 71C), made of wood, is

played in the same way. The example in the Krige Collection is made of moro-

gdl6 wood (Carissa edulis) a fine-grained wood that does not split easily. A

whistle, naga, made of horn, reed, or wood, is a pipe closed at one end and

sounded by blowing across the open end. A naga (Fig. 71D) collected in the

1930s is made of wood shaped to resemble a horn and decorated with wire bind-

ing. A feather kept in the open end is used as a pipe cleaner.

A small vibrating wind instrument that has great ritual importance among

the Lobedu is the mantsaakhéda (Fig. 71E), which consists of two pieces of

wood wider at one end than the other and slightly concave on the inner surfaces,

between which is inserted a piece of bark or any material that will vibrate when

blown. Rubber was used in a specimen collected in 1970. The mantsaakhéda is

surrounded by secrecy and is used by old men who impersonate the voices of the

ancestral spirits, zwidajane, in certain religious ceremonies, for example the thu-

gula harvest thanksgiving and the khetshila purification ceremony after the death

of royalty. The spirits are believed to be hidden in the near-by bush and to ex-

press themselves by whistling. Messages are conveyed by varying the tone of the

LOBEDU MATERIAL CULTURE 129

20

Fig. 71. Wind instruments. A. Flute, khethodig6, SAM-9814. B. Flute, rodig6, SAM-9815S.

C. Flute, Jengwale, SAM-9816. D. Whistle, naga, SAM-9817. E. Whistle, mantsaakhdda,

SAM-9824. (All Krige Collection, 1936-8.)

whistling, which can be translated into the tone patterns of the language. The

mphenyi instrument played by khiudogane in the vyali school (see p. 99) oper-

ates on the same principle. The vibration of the membrane across the end of the

reed pipe produces the strange eerie sounds by which khiudogane communi-

cates.

Wind instruments made of horn are described in a later section (see

p. 166). They include whistles and the horn trumpet, phalafhala, which is

associated with authority and is blown to signal important announcements.

4. Idiophones

In the 1930s xylophones, dibela, with calabash resonators (see p. 153) were

used in the Lobedu area by itinerant minstrels who played the xylophone and

sang humorous songs in return for food and lodging (Krige & Krige 1943: 316).

There is no record of xylophones being made by local craftsmen and it is prob-

able that the instruments used were of Venda origin (Kirby 1934: 56).

Also in use in the 1930s was a plucked instrument, deze, consisting of a flat

wooden resonator set under a series of iron tongues that were plucked with the

thumbs to produce the sound. By adjusting the length of the iron tongues the in-

strument could be tuned. This sansa-type instrument was played mainly by

young men (Krige 1982: 22).

The wooden clappers mewasa (Fig. 72) used by vyali girls also fall into this

category of instruments made of inherently resonant materials.

130 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 72. Wooden clappers, mewasa, used by girls during the vyali initiation, SAM—9858, Krige

Collection, 1936-8.

Playthings

In making playthings, children acquired a lot of useful knowledge of the en-

vironment and an early training in craftwork. Herd-boys made everything they

used in hunting and in playing games. They got to know the qualities of the dif-

ferent woods and their toys changed with the seasonal availability of the natural

materials, grasses (see p. 89), fruit shells, or maize stalks. Wooden toys in-

cluded wagons, motor cars, guns, sledges, and clubs for playing khdr6r6, a

game similar to hockey (Krige & Krige 1943: 108). Today boys still make their

own toys but the raw materials seldom come from the veld. Wire, bottle-caps

and tin-cans are used most effectively for making cars, tractors or lorries.

Ritual objects, figures and carved poles

At present little is known about the traditions of figurative carving among

the people of the northern Transvaal, although a thesis based on field research

undertaken in the 1970s is in preparation (A. Nettleton, Department of Fine

Art, University of the Witwatersrand, 1981 pers. comm.).

Carved figures of symbolic significance in initiation rituals are used by the

Venda (Stayt 1931: pl. 31) and have also been collected among the Kgaga of

Maake and the Narene of Sek6r6r6 (Potchefstroom University Collection). A

male and a female figure in the collection of J. Witt of Tzaneen were said to

have been among the Lobedu digéma revealed to initiates during a vyali

initiation (see p. 99). At the end of the vyali the costumes and all associated

objects should be burnt, so it is unusual for them to be represented in museum

or private collections. Furthermore, the secret nature of these objects forbids

their description by those who have seen them. A géma is essentially something

131

LOBEDU MATERIAL CULTURE

1973»

the capital,

’s village,

dji

.

ja

6rd of Mod

oro surrounded by forked poles.

B. The kh

Fig. 73. A. Layout of huts around central kh

——————— ES

132 ANNALS OF THE SOUTH AFRICAN MUSEUM

secret and mysterious but it may be intrinsically very simple. The ritual context

of its revelation imbues a géma with its awesome quality.

The wooden clappers, mewasa (Fig. 72), used by vyali girls to accompany

their songs, are simply two flat pieces of wood carved to a point at one end and

a handle at the other. No significance is attached to their manufacture but, once

used, they acquire significance and they must be burnt at the end of the school.

For the vyali initiation a row of debarked poles is set up in the khdéro in a

specially medicated furrow at the back of a carefully prepared platform, /e-

vhalélé la vyali, on which the initiates stand when singing. In front of this is the

kholon6ni shrine, a circular mound of clay in which a river stone is embedded

and in the centre of which is a forked, medicated pole of musoso wood (Termin-

alia sericea, vaalboom) believed to have protective powers (Krige & Krige 1943:

136).

The khéro of the Queen’s village (Fig. 73A) is surrounded by a palisade of

pointed or forked poles (Fig. 73B) that may only be used in the villages of chiefs

or headmen. Headmen from all the districts are called up to provide poles for

the Queen’s kh6ré when it is renewed. In this way the khdro of the capital ex-

presses the solidarity of the chiefdom. The entrance to the khdro is also the en-

trance to the village and, as such, it is protected with posts of special wood

(kherale or moludu) treated with medicines when they were set up. Beneath the

threshold a medicated wand or a river stone is buried to neutralize or ‘cool’ any

evil that may threaten the village. These measures are but a very small part of a

complex scheme for protection of the village.

Fig. 74. Carved khéro poles at the capital, 1973.

LOBEDU MATERIAL CULTURE 133

Some of the khéro poles are carved in the form of stylized figures (Figs 74,

75A-C). Their meaning is not fully understood. According to Krige (notes) they

were made to honour the Queen. Schlosser (1975: 95) stated that they were ex-

clusive to the Queen’s khoro but had no other significance. According to J. Witt

of Tzaneen who has made a collection of carved poles from Modjadji’s khéré,

craftsmen of particular skill carved the poles to distinguish their contribution to

the khéro.

Fig. 75. Carved khér6é poles. A-B. From the capital, now in private collection of J. Witt.

C. Collected when the poles of the khér6 at the capital were being renewed, SAM-9751, Krige

Collection, 1936-8.

134 ANNALS OF THE SOUTH AFRICAN MUSEUM

Structures

Although the study of settlement pattern, hut forms and _building-

techniques was not a main focus in this study, structures and building-techniques

recorded in the 1970s are briefly described and supplemented by data and photo-

graphs from the 1930s provided by E. J. Krige.

The smallest cluster of huts within a village consists of a sleeping-hut, ndo,

a cooking-hut, khetaga, and a courtyard, mosha, formed by the space between

them. At the capital these pairs of huts are arranged in a circle round a central

courtyard, the khdér6é (Fig. 73A). Sleeping-huts all face the khdér6é and cooking-

huts have their backs to the khdér6. Each wife in a polygynous family has her

own mosha, her own utensils, storage area, granaries, and fields. The household

of the chief wife of the head of the village is usually positioned opposite the

main entrance but, apart from this, there is no ranking of household units within

the village. In smaller settlements there is no khoro and huts are arranged in an

arc or straight line.

Fig. 76. Framework of a hut under construction in the 1930s. Photo: E. J. Krige, 1936-8.

LOBEDU MATERIAL CULTURE 135

i Auts

The oldest huts were built using the pole-and-daga method but even before

the 1930s sun-dried mud-brick construction was replacing the older method that

drew heavily on timber resources. Between 80 and 100 poles were needed for a

single hut (Fig. 76). The following description of the pole-and-daga method was

recorded in 1976 from wood-carver F. Semosa, and examples of the materials

were collected for reconstructing a hut at the South African Museum. Hard, ter-

mite-resistant woods such as moathaba (Pterocarpus rotundifolius), mosese (Pelt-

ophorum africanum), or moréjé (Dicrostachys cinerea) were said to have been

preferred, but many other hard woods were also used. It was often necessary to

straighten the poles before use. This was done ingeniously by making a series of

oblique cuts on the inner curve of the pole while it was still green and then

wedging it in the fork of a tree and bending it straight (Fig. 77A). In 1976 poles

straightened in this way were still visible in one of the oldest huts at the capital

(Fig. 77B). The process, -vhaga, was also used by the Venda (Van Warmelo

1937: 299; Van der Waal 1977: 79) whose hut-building techniques were similar

in many respects to those of the Lobedu.

The straightened poles were carved roughly to a point at one end and

embedded in the ground to a depth of about 40-50 cm, following a circle that

had been traced out on cleared ground. The soil was saturated with water to

facilitate embedding the poles. Flexible withies from trees such as mofwara

tsweni or mothala maligana (both Grewia spp.) encircled the poles horizontally

at regular intervals and were lashed on to the framework with strong bark fibre.

Both the laths and the bark had to be softened in water before use. The ridges

formed by the laths were usually accentuated in the plastering-process and made

into ledges in the walls. A mixture of mud and dung was applied by hand to the

framework both inside and out. This part of the work was done by women and

they also helped with transporting the materials from the bush and collected the

thatching-grass for the roof (Frans Semosa 1976, field records).

The conical roof structure of the oldest huts was made of closely packed

rafters (Fig. 78B). To construct the roof a central post, mosima, was erected in

the middle of the hut, and a working-platform was made by placing planks

across the walls and securing them to the central post. The pointed wooden roof

pinnacle, /Jenodu, was supported by the mosima while the rafters, levhalélé la

thaga, sharpened at the upper end, were tied to the pinnacle. The lower ends of

the rafters rested on and were tied to the poles forming the walls, /evhaléldé la

tshite. Flexible cross-laths tied to the rafters encircled the roof at intervals.

When the roof was complete the mosima was cut away (Krige 1982: 13).

The roof was thatched with bundles of grass placed on top of the rafters

(see p. 93). Krige recorded that in the 1930s the plastering of some sleeping-

huts continued up into the roof to form a plastered dome (Fig. 78A). “This dome

has circular rims marking the cross-laths which, beneath the plaster, hold to-

gether the framework of closely packed rafters’ (Krige & Krige 1943: 21). This

136 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 77. A. Straightening a pole by wedging it between two trees after oblique cuts have been

made on the curve of the pole, Semosa’s homestead, 1976. B. Hut poles showing the cuts,

Modjadji’s village, 1976.

LOBEDU MATERIAL CULTURE 137

Fig. 78. A. Plastered roof of a sleeping-hut. Photo: E. J. Krige, 1936-8. B. Traditional

roof structure of a cooking-hut. Closely packed rafters support the thatch. Modjadji’s

village, 1976.

138 ANNALS OF THE SOUTH AFRICAN MUSEUM

feature, also found in Venda huts (Van der Waal 1977: 88), gave a very neat

finish to the hut interior. These architectural details disappeared when new con-

struction techniques were adopted. The plastered ridges were not made on mud-

brick walls, and ceilings could not be plastered when widely spaced rafters were

used (Fig. 79A).

Wooden doors used in sleeping-huts were carved from solid pieces of soft

wood such as wild-fig and hinged by means of projections on the upper and

lower ends that fitted into sockets in the lintel and threshold beams. A reed door

(see p. 85 and Fig. 16) was used in the cooking-hut.

By the 1930s brick construction, which had been taught by the mission since

the late nineteenth century, was replacing the older pole-and-daga construction

method, and legislation to protect many of the indigenous trees accelerated the

process by reducing the availability of wood (Krige 1941: 18). Bricks were made

by women and men using wooden box moulds (Fig. 79B) or tins and at the capi-

tal bricks were commissioned for 10 shillings per 800—1 000 bricks (Krige 1941:

18). The increased participation by women in hut-building compensated to some

extent for the absence of men working in towns, and building with sun-dried

mud-bricks has continued to the present both for the conventional round huts

described below and for more modern houses.

The ground is cleared and a circle drawn with a stick attached to a string

fixed to a peg in the centre of the circle. A shallow trench of stones is usually

laid as a foundation for the brick coursing. Mud plaster is applied between the

bricks and, when the wall is complete, to the interior and exterior. The initial

plastering is done by hand and a wooden tool, khevhado, is used for compacting

the plaster. The same tool is used at a later stage when the roof is complete for

compacting the mud floors of huts and courtyards, which are smoothed, -ridéla,

with a stone, thidél6, then smeared with dung, and sometimes decorated with

patterns made by running the fingers over the wet dung.

The traditional roof structure of closely packed rafters, which supported the

thatching grass without the need for sewing, was replaced by the ‘rondavel’ roof

of widely spaced rafters, usually of purchased blue-gum timber, on to which the

thatch was sewn. In the 1930s modern window and door-frames were rare except

at the Christian settlement, but by 1940 a few rectangular houses with windows

had been built at the capital (Krige 1982 pers. comm.). Subsequently many rect-

angular-shaped huts with thatch or tin roofs were built but in the 1970s, at the

capital and in smaller settlements, the circular-plan hut remained the most

widely used dwelling.

In the 1970s the apparent trend among non-traditionalists who could afford

the materials and were living within access to transport was toward modern

housing. Many people, however, still lived long distances from main roads and

the most frequently seen dwellings were mud-brick huts, which are low-cost

structures appropriately designed for a hot climate and a living-pattern in which

most activities take place out of doors.

LOBEDU MATERIAL CULTURE

Fig. 79. A. Modern ‘rondavel’ type roof structure with widely spaced rafters, plastered

mud-brick walls. Modjadji’s village, 1973. B. Mud-brick and mould, Ramatswalela’s

homestead, 1975.

139

140 ANNALS OF THE SOUTH AFRICAN MUSEUM

2. Granaries

In the 1930s granaries were very much a part of Lobedu domestic life. Per-

manent granaries were built in the storage area, marala, behind the huts and

temporary granaries were built in the fields at harvest season. By the late 1970s

there was no need for granaries as there was not enough grain to be stored and

most families depended on bought maize-meal for their basic subsistence

requirements. Very few granaries were seen in use during fieldwork and the

traditional grain pit for maize had fallen completely into disuse. The following

descriptions are based on photographs and information recorded in the late

1930s by E. J. Krige (1982: 14-15).

Traditionally decobbed maize was stored for long periods in a grain pit,

kheledese (Fig. 80), usually situated in the cattle kraal. At the royal village the

pits were in the khér6. The pit was a large hole in the ground lined with dung

and grass; it had a fairly narrow opening but broadened out below ground level.

Before being placed in the pit the grain had to be completely dry to minimize

- = : . a

> -...

Fig. 80. Kheledese, pit for storing maize. Note depth of earth above opening of pit.

Photo: E. J. Krige, 1936-8.

141

LOBEDU MATERIAL CULTURE

ee ie

an

3

23

Hu DN

CO

oe

Zo

aA

“fa eh Sip

S58

om

oOo GO

=e

=e

=| SSI?

Ow 2

Bae

So,

eionas

age

Ewe

oO Ay

aa

i

So 6

ae

28

= a.

o ?

~

In

h grass i

it

6, granary formerly used for stor

Fig. 81. A. Disused kheash

of cycads, 1976. B. Kheashé w

have been adopted from Tsonga

142 ANNALS OF THE SOUTH AFRICAN MUSEUM

the chances of it rotting. A stone was placed over the opening, which was then

sealed with dung and covered with earth to ground level. A stone was placed to

mark the stituation of the pit. Storage in the sealed pit changed the flavour of

the maize, which was then called madoni. Certain pits were known to produce

particularly good madoni and were used repeatedly.

The kheashoé (Fig. 81A) was a structure of poles, mabala, encircled by

cross-laths at the top, middle and base, and secured with bark fibre. The base

was raised about 20 cm off the ground on forked posts or large stones. This was

the traditional way of storing maize on the cob while drying. A similar structure,

also referred to as kheashé (Fig. 81B), but with grass woven between the poles,

was introduced by Tsonga immigrants. The grass was said to protect the grain

from being pecked at by fowls. This protection was only necessary after the

Tsonga practice of keeping fowls had gained acceptance among the Lobedu.

The Jedulu (Fig. 82) was a granary of roughly the same shape as the

Fig. 82. Ledulu granary for millet and sorghum, made

of a wooden framework plastered with clay. The

thatched roof must be removed to reach the contents.

Photo: E. J. Krige, 1936-8.

LOBEDU MATERIAL CULTURE 143

kheasho but the walls were plastered with daga and it had a conical thatched

roof that had to be lifted to reach the contents. It could be sealed for long-term

storage. The /edulu was the traditional Tsonga granary for decobbed mealies but

was widely used by the Lobedu and the Venda for sorghum and millet as well as

for the short-term storage of maize.

The /etole (Fig. 83) was a temporary granary erected in the fields for storing

sorghum before it was threshed. Structurally it was an inverted roof supported

by forked upright posts. It was lined and covered with grass to protect the grain

Fig. 83. Lejéle granary in which sorghum was stored in the fields before threshing.

Photo: E. J. Krige, 1936-8.

. from birds. A raised platform moak6 (Fig. 84), was also used in the fields for

temporary storage of millet and sorghum prior to threshing and winnowing. In

spring these platforms were used to stand on when scaring birds from crops. A

smaller permanent platform, vhoald, built inside the cooking-hut, was used for

keeping bags of grain off the ground to protect them from termites and other

insects. In the 1970s these platforms were still seen in some traditional house-

holds.

STONE

Oral traditions record a time in Lobedu history (probably in the late eight-

eenth century) when stone enclosures were built. The description (Krige &

Krige 1943: 7) of high stone walls, upright stone projections, and a stone seat for

the chief clearly link the Lobedu with the Venda and their Karanga ancestors

144 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 84. Moako platform used for temporary storage of sorghum in the fields prior to

threshing. Photo: E. J. Krige, 1936-8.

north of the Limpopo River. The ruins of these stone structures associated with

the reign of Chief Kheale are still standing, although much overgrown. The site

is strategically situated on a hill-top commanding a good view of the surrounding

country. In other areas there are remains of stone-walled enclosures for live-

stock and terraces of former occupation sites (Fig. 85).

In some villages stone-walled enclosures for livestock are still found and

stone is still used for the contour walling on the cultivated hillslopes. Many old

contour walls, maraka, are now overgrown but they show up well on aerial

photographs. Valley cultivation was at first practised only by Tsonga-speaking

immigrants but later shortage of arable land on the hillsides and the introduction

of an effective plough led increasing numbers of Lobedu to cultivate the flatter .

ground. The plough, however, destroyed the protective verges of the fields and

broke down many of the stone contour walls that prevented erosion (Krige &

Krige 1943: 236). Where cultivation of the steep hillslopes has continued, stone

contour walling is still built.

Unworked stones were and still are used in a number of ways for grinding,

crushing, pounding, or smoothing. In a pair of grinding-stones, the lower stone,

lwala, is usually flat, large and fairly rough while the upper stone is smaller,

harder and rounded. Stones are used for grinding sorghum and millet but not for

maize, which is stamped in a mortar. Heads of maize are decobbed on a rough

stone (Fig. 86). During the preparation of clay it is pounded with a wooden

pestle on a large grinding-stone. Smaller, hard stones are used for cracking open

the hard shells of marula pits and generally for cracking nuts. A hard, smooth

LOBEDU MATERIAL CULTURE 145

Fig. 86. Rough stone for removing maize from the cob. Molototsi valley, 1975.

stone, thidél6, is used to compact the surface of a newly made mud floor and

smaller rounded river stones are the traditional smoothers and burnishing-tools

for pottery (see p. 56).

Sandstone, /ehut6, often referred to as the ‘Vim’ stone, is widely used as an

abrasive cleaning agent for utensils. A large grinding-stone on which the lehuté

146 ANNALS OF THE SOUTH AFRICAN MUSEUM

is crushed is generally found near the washing-up area (Fig. 87). A less abrasive

stone is used for scrubbing the body when washing.

Herd-boys in the fields used hollow stones as troughs for goat’s milk which

they curdled by adding acidic fruit juice and ate with spoons made from grass or

leaves (Krige & Krige 1943: 107).

Fig. 87. Grinding-stones for crushing sandstone into an abrasive powder for scouring

utensils. Note pot for water, calabash utensils, and spoon. Modjadji’s village, 1975.

Stone-carving seems to have been confined to the making of soapstone

(steatite) pipe-bowls, gagana (Fig. 88), for smoking dagga, badje (Cannabis sa-

tiva). No technical information was obtained on their manufacture and they are

no longer made. According to Krige (1982: 41) they were used together with a

reed, khetoméla, and a horn water-vessel, naga la khomo. The dagga was placed

in the stone bowl fitted over one end of the reed, the other end was inserted in

the horn filled with water. A glowing ember was placed in the bowl on top of the

dagga and the smoker drew the smoke through the water by inhaling deeply

over the mouth of the horn. This provoked a strong flow of tears and saliva

which could be expelled through another reed, mphasi. This was called maada a

khoho.

By the 1930s dagga-smoking was illegal and the horn water-pipe was no

longer in use. To avoid being arrested people had taken to placing the dagga in

a dampened hole in the ground, and smoking in a lying position. The smoker

filled his mouth with water and drew the smoke from the dagga through a reed,

LOBEDU MATERIAL CULTURE 147

Fig. 88. Stone pipe-bowls, K68, Klapwijk Collection.

khetoméla, positioned to make contact with the dagga which had been placed in

the hole and lit. Nowadays dagga is not smoked through water but in an ordi-

nary pipe or by rolling it in paper.

SUMMARY AND COMMENT

The indigenous Lowveld environment offered a wide range of materials for

possible exploitation. In the above section the uses of clay, grass and other plant

fibres, wood, and stone are described. Although many of the techniques are

termed ‘traditional’ this is not intended to imply a fixed body of skills but to

refer to processes that had developed and continued to develop in a particular

environment and within a self-sufficient subsistence economy.

The form of the material culture at any one time depends on an interaction

of technical, economic, social, and ideological factors, each of which contains a

number of related variables. Technical variables include the types of material,

the tools and the skills; economic variables include the demand for the products,

the division of labour, and the impact of money and mercantile goods on local

production; social variables encompass the range of social contexts for the use of

artefacts and customary patterns of use; ideological variables relate to the sys-

tem of values that gives meaning to certain materials and products.

Natural resources constituted the most important material base in the Lo-

bedu economy. In addition to providing wild foods and land for cultivation and

grazing, the natural environment provided most of the materials on which the

productive technology depended for the construction of settlements and the

148 ANNALS OF THE SOUTH AFRICAN MUSEUM

satisfaction of most other material requirements. Using the Krige Collection as a

sample, over 60 per cent of the inventory of artefacts derive primarily from the

indigenous environment and, if huts and granaries were included, the figure

would be well over 80 per cent.

Knowledge of the natural environment was a very significant aspect of tech-

nology. In the 1930s Krige & Krige (1943: 45-50) recorded the names of over

500 plants used as food and in the technology, and commented that knowledge

of the vegetation and its application in craftwork amounted to a ‘veritable

science’. Selection of appropriate materials for particular purposes was an essen-

tial part of the technical process. This knowledge was acquired from an early age

in play activities and later through participation in subsistence activities. Herd-

boys, for example, developed a sensitive understanding of their environment

and acquired many technical skills. Since children have been receiving formal

education in school, less time is spent in the veld and less is learnt about the

~ natural environment. School education has also tended to undermine a value

system that stressed order in nature and the ritual control of nature by the chief

through the ancestors.

Lobedu orientation toward nature was expressed not only tangibly in the

material culture but intangibly in the rain and fertility cult, and the seasonal

ritual offering of the first termites to the ‘rain-pots’. In the vuhwera and vyali

initiation rituals there was a strong emphasis on the wearing of costumes from

the appropriate natural materials, often associated with rain and damp places

and believed to have positive ‘cooling’ qualities. In the vuhwera costumes there

was an integration of tangible materials and the intangible symbolic values with

which they were invested. These values were further echoed in the songs of ac-

companying vyali girls which were believed to enhance fertility.

Human use of natural resources modifies the environment which, in fact,

becomes partly man-made. The clearing of land for settlement and cultivation

depletes the natural resources and, in the area under study, the accessibility of

many of the traditionally used plant fibres and woods has been appreciably

reduced. Depletion of timber resources prior to the 1930s and consequent legal

restrictions on the felling of trees encouraged a change in the method of roofing

from closely packed to widely spaced rafters, and of hut-building from pole-and-

daga construction to that of mud-bricks. The latter change in building technol-

ogy, which involved a greater use of mud, was also encouraged by an increased

participation of women in hut-building to compensate for the absence of men

working in towns.

Sources of clay, mainly in river-banks, are distributed throughout the area

and in the 1970s were still used extensively by potters. A number of factors have

contributed toward continuity of the craft of pottery. Important among these is

the fact that women are the potters and main users of the products. The technol-

ogy has been passed on from one generation of women to the next, women have

tended to remain in the rural villages, and the pattern of use has remained

relatively constant. Earthenware is generally preferred for cooking on an

LOBEDU MATERIAL CULTURE 149

open hearth and clay pots are less expensive than cast-iron alternatives.

Although water may be carried in lighter tin or plastic containers, it is always

stored in earthenware, which has the practical advantage of keeping water cool

and fresh.

Since the 1930s when only a few of the Christian community made a ‘busi-

ness of pottery’ (Krige & Krige 1943: 320), the manufacture of pottery for sale

has increased. Pottery not only supplies subsistence needs but is one of the ways

in which village women can earn money without having to seek wage employ-

ment. This factor has led to an increase in the scale of production.

Although the staple diet has remained maize-based, there is a qualitative

nutritional difference between home-stamped maize-meal and the less nutritious

refined product bought in the shops. Most people, however, are dependent on

bought maize as over-population has resulted in many people being landless.

The insufficiency of grain is reflected in the disuse of granaries, a reduction in

the frequency of stamping maize and in the decreased use of related utensils

such as mortars, mats for drying meal and pots for steeping grain prior to stamp-

ing.

Changes in basketry since the 1930s seem to relate mainly to reduced

demand for certain products as a result of the changing subsistence base, an

increase of migrant labour, and related changes of values. An example of this is

the kherodwana basket that has become almost obsolete following a change in

the pattern of marriage as a result of migrancy. Formerly this serving-basket was

symbolic of the willingness of a bride to honour her husband and to serve him

respectfully. As migrancy increased many marriages were contracted in the

absence of the man with the result that there was little need for the basket. The

wooden porridge dish, del6, which fitted inside the basket was also used less

frequently. As demand decreased, so did the motivation to transmit the related

technical skills, with the result that they are unknown to the younger genera-

tion.

The replacement of woven beer-strainers by a large wooden sieve was re-

lated to the brewing of maize-meal beer for sale. Formerly woven strainers were

used when beer was brewed in relatively small quantities, mainly from sorghum,

and was used to fulfil social obligations or to show hospitality. This type of

strainer was not suited to straining large volumes nor was it suitable for beer

made from maize-meal. On the other hand, the fine-meshed, wooden-framed

sieve is well suited to straining finely ground maize-meal and can handle large

volumes. As beer-making for sale increased, the woven beer-strainer fell into

disuse.

As migrancy increased, women took over the making of mothatha baskets

and pate mats, traditionally made by men, and they continue to be made by

women for domestic use, although the mothatha baskets have largely been re-

placed by enamelware.

Working with wood, including the weaving of baskets from wood-slivers,

remains exclusive to men and has tended to become more specialized. Many

150 ANNALS OF THE SOUTH AFRICAN MUSEUM

new tools have been added to the tool-kit but the traditional adzes and gouges,

well adapted to their functions, have not been replaced. Tool marks made by a

lehoér6 gouge are visible on a number of items in the Krige Collection and this

type of tool was still used by woodworkers in the 1970s. In general there is a vis-

ible relationship between the tools, the motor action of the user, and the form of

the product. A change in the use of tools is usually detectable on the artefact

and this is particularly so when modern knives, files, and sandpaper are used.

Traditionally many natural materials were used without modification by for-

mal tools. Stones, selected for their particular shape or texture, were used as

nut-crackers, grinders, and smoothers. The abrasive Jehut6 was and still is

widely used for scouring utensils, and the ‘loofah’ plant, leshokhoté (Luffa

aegyptiaca) provides an effective scourer. Many plant fibres can be used as lash-

ings without any technical preparation and the natural shape of the Bauhinia

seed-pod, thema, is well suited to its use as a pottery tool. Thus natural

- materials continue to be used in a variety of ways and have not been replaced by

purchased alternatives.

Despite the transition to a money economy, in many respects the natural

environment remains an important, if diminishing, source of materials and,

where there is still a demand for traditional products, the related technical skills

remain alive.

USE OF DOMESTIC RESOURCES

The products of cultivation and livestock husbandry form the main content

of this section but fruit shells, although natural resources, are included with

gourds, and horns of game animals with those of cattle, as technically they are

treated in similar ways.

GOURDS

Gourds belong to the family Cucurbitaceae, as do pumpkins and melons.

They are cultivated for their fruits and for their leaves which are made into a rel-

ish. The mature fruits of the indigenous species Lagenaria siceraria (formerly

Lagenaria vulgaria) are widely used for making calabash utensils. The smooth-

skinned gourd, thaga, grows in a variety of shapes that lend themselves well to

the making of ladles, scoops and flasks; the small rough-skinned gourd, leraka,

is used mainly for making rattles. The seeds from gourds of useful shape are

kept for the following season.

Seeds are sown during the summer rains between November and January.

Most women plant gourds among the other crops and leave them in the fields

until they are the required size and hardness. By June the mature gourds are

usually considered ‘strong enough’ to be picked and made into utensils. There is

no specialization in this activity—all the village women know how to prepare

calabashes and make those that are needed for domestic use. Some women are

LOBEDU MATERIAL CULTURE Sul

expert at cutting a neat rim and are asked to do this for others. If a woman has

made too many calabashes for her own use she may exchange them for produce

or sell them.

Gourd-working techniques

A gourd of suitable shape is selected for the product in mind. The method

of manufacture is. essentially the same for all calabash vessels. The process ob-

served during fieldwork in the 1970s did not differ from that described by Krige

for the 1930s. A small hole is cut in the gourd which is then cooked in boiling

water for about an hour. When cool, the hole is enlarged, the soft inner flesh

removed with a spoon (formerly with a seed-pod scraper, thema) and the edge of

™

Fig. 89. Repair techniques for calabash utensils. A-B. With wire. C. With string.

152 ANNALS OF THE SOUTH AFRICAN MUSEUM

the rim cut neatly with a knife. The outer skin is scraped off, -thala, and the

gourd is then scoured, -huta, inside and out with powdered sandstone, lJehuto. It

is left outside to dry and is then ready for use. Constant washing with lehuto re-

moves any rough edges from the rim and gives the calabash a smooth finish. In

this way the quality of a calabash utensil improves with use.

The method of preparation differs slightly if the gourd used is too big to fit

into a pot of boiling water. After the hole has been cut, the gourd is inverted

and left for about a week so that the flesh can rot. Thereafter the inside is

cleaned out, the outer skin scraped off, and the surface scoured.

Rattles are made simply by letting a /eraka gourd dry out thoroughly. It is

not boiled. Holes are branded at opposite ends and the inside is cleaned out

with a twig. Small stones or seeds are put into the cavity and a stick longer than

the gourd is pushed through so that it stops the holes and forms a handle at one

end.

Cracked or broken calabashes may be repaired by making a row of tiny

holes on either side of the crack with a fine awl made from a sharpened nail or

umbrella-spoke. The crack is then sewn together with fibre string or fine wire

(Fig. 89A—C). The stitch may be simple oversewing or a more complicated

stitch such as herring-bone. During the months when gourds for making new ca-

labashes are not available, repairs become a practical necessity.

Fig. 90. Calabash utensils: bowl, khethéba, SAM-—9686; scoop, khefhahé, SAM-9711; and

ladle, kheh6, SAM-9693. (All Krige Collection, 1936-8.)

LOBEDU MATERIAL CULTURE 153

Uses of gourds

Utensils and containers

Gourds are made into a number of domestic utensils including bowls,

ladles, scoops and flasks (Fig. 90). A calabash bowl, khethéba, is in daily use for

the staple maize-meal porridge, vhoswa. A smaller khethébana is used for soft

porridge, khepye, and for serving beer individually to an honoured guest. When

used for beer it is referred to as phafhana.

Ladles, zwihd, of various shapes are in regular use. Khehé kha vyalwa is a

long-handled ladle for serving and drinking beer. The long handle makes it easy

to pass from one person to another. Khehé kha mabudu is a short-handled ladle

for light beer not usually drunk in company. Kheho kha meezi is a large ladle for

transferring water from a storage pot to other containers.

An open scoop, khefhahd, is used for handling dry meal. When making

porridge the meal is added to the water with this utensil.

A large vessel, mokhaba (Fig. 91A), is used for transporting liquids. The

small mouth prevents spillage and it is much lighter and less likely to break than

a clay vessel used for the same purpose. In the 1930s a waisted calabash, khe-

khébo (Fig. 91B), was in use as a portable flask for carrying thin porridge.

Seeds or dried beans were kept in similar calabash vessels. Modern containers

have largely replaced calabashes for these functions.

Small-waisted calabashes fitted with stoppers were used as containers for

herbs, medicines, or snuff (Fig. 92). To some extent these have been replaced

by small tins and bottles but they are still used, especially for medicines.

Musical instruments

The resonators of certain musical bows and xylophones consist of suitably

shaped calabashes. The musical bow, khekhaba, (see p. 124) which derives its

name from the North Sotho word for a gourd, has a single calabash resonator

that is held against the player’s body during performance. A xylophone, bela,

(see p. 129) has a resonator below each wooden bar. They are oval in shape,

graded and positioned so that the largest calabash is the resonator for the

deepest note. Near the closed end of each resonator there is a small hole with

resin on the rim and covered with pieces of spider’s web which vibrate when the

instrument is played and add to its tone (Kirby 1934: 50-56). Xylophone players

Kept the seeds of successful resonator gourds so that they could cultivate gourds

of similar shape.

Rattles, tshélé (Fig. 93A), used in rituals of spirit possession are made from

the rough-skinned gourd, /eraka, in the manner described above (p. 152). Leg

rattles, mathodzi (Fig. 93B), are made from the smooth hard-shelled fruit of the

mothodzi tree (Oncoba spinosa). When the lethodzi fruit has turned a dark

brown on the tree it is picked, a small hole is made at each end and the seeds re-

moved with a stick. Small stones are put inside the dry empty shells and a stick

is passed through the holes to close them. Three or four sticks each fitted with

154 ANNALS OF THE SOUTH AFRICAN MUSEUM

B aes

Fig. 91. A. Calabash flask, mokhaba, SAM-9732. B. Waisted flask, khekh6b6,

SAM-9847. (Both Krige Collection, 1936-8.)

LOBEDU MATERIAL CULTURE 155

Fig. 92. Small calabash vessels for snuff or medicines, SAM-—9918 and 9855.

(Both Krige Collection, 1936-8.)

Fig. 93. Rattles. A. Tshélé SAM-9830. B. Mathodzi SAM-9831. (Both Krige Collection,

1936-8.)

156 ANNALS OF THE SOUTH AFRICAN MUSEUM

four or five fruit shells are held together with thongs, or sisal string, threaded

through holes at the ends of the sticks. The thongs are long enough to be used to

tie the rattles to the wearer’s leg. Sometimes slits are made in the shells to en-

hance their sound. They are worn by mal6b6 dancers during spirit-possession

rituals and at recreational dances.

Playthings

Young boys make humming-tops, mankényana, out of lethodzi fruit shells.

A stick is fitted into a hole at one end of the shell and fixed with beeswax. The

string is wound round the stick and when pulled it causes the top to spin with a

humming sound. According to Krige (notes) this is one of the oldest Lobedu

toys. They were seen occasionally during the author’s fieldwork.

Ritual uses

Melons, maguadzi (Citrullus lanatus), were among the first green foods

each year to be ritually ‘bitten’ and offered to the ‘rain-pots’ (Krige & Krige

1943: 274). A small calabash, phafhana, was considered the most appropriate

vessel in which to offer beer to the ancestors.

In former times a doll with a body made of a calabash was given to a girl

when she reached puberty. This doll was kept by the girl until her marriage

when it was carried in a kherodwana basket by the bridal party. If she were not

a virgin the calabash was pierced. According to Krige & Krige (1943: 113) this

custom had fallen into disuse by the 1930s.

SKINS AND HIDES

Livestock husbandry was less important in the Lobedu subsistence economy

than agriculture and this is reflected in the material culture. Goatskins, however,

were important in the traditional technology providing the main material for

clothing as well as for the bellows used in metal-working. Goats were slaugh-

tered on occasions of ritual or social importance and they were also used for the

payment of debts or fines and in many other economic exchanges (Krige &

Krige 1943: 45). Cattle, on the other hand, were rarely sacrificed and the use of

their hides was less significant in the technology.

Skin-working techniques

Skin-working involved a number of processes: flaying the animal, fleshing,

dressing, cutting, and sewing. Men did most of the work but women helped with

the dressing of the skin (Fig. 94). The following description is based on accounts

by men who no longer prepared skins in the 1970s but remembered the tech-

nology.

A skin was removed from the carcass, pegged out with wooden pegs,

dinédo, and rubbed with an abrasive stone or dry maize cob to scrape away the

LOBEDU MATERIAL CULTURE 157

Fig. 94. Woman applying fat to a goatskin. Photo: E. J. Krige 1936-8.

fat and clean the flesh side (Fig. 95). This removed matter that would putrify

rapidly if the skin were left untreated. Thereafter it could be allowed to dry out

before being softened or dressed. At this stage the skin was still very susceptible

to infestation by insects or bacteria and the experienced skin-worker knew the

value of certain plant oils and repellents that could be used to protect the raw

hide (Krige & Krige 1943: 47). The dressing-process that followed rendered the

skin less prone to decay but did not make it completely stable under unfavour-

able conditions.

The pelt was dampened with warm water for the softening-process, -suha,

which was done by rubbing the skin briskly by hand. This was an arduous pro-

cess and a number of men would work together on a single skin or take turns in

158 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 95. Goatskin drying after the flesh side had been cleaned, Molototsi valley, 1976.

working on the skin. Pips of the ditshidi fruit (Ximenia caffra) were ground be-

tween two stones to extract the fat, which was then applied to the skin. This

acted as a dressing and kept the skin supple. The hair was usually left on

goatskins but could be removed before softening by soaking the skin in water

and scraping the skin with a rough stone. Cattle-hides used for sandals, shields,

and quivers were not softened but covered with dung and buried so that the hair

would rot and fall off.

For certain purposes a prepared skin had to be cut and sewn to a conven-

tional pattern. This was the work of a specialist craftsman. A woman’s skirt,

motshe (Fig. 96), required particular skill. Gussets were cut to give the skirt the

correct shape; flaps of skin, dinyedi, were cut and sewn on later for decoration;

a triangular ‘tail’ was cut and sewn on to the waist so that the hairy side faced

outward when folded over the back of the skirt. Pieces of skin were formed into

small studs and attached so that the hair showed up against the flesh side on

which a nap was raised by making small incisions with a special metal tool,

mosédo (see p. 177), with a cutting edge that could be sharpened. The nap-

raising process, -déga, gave the skirt its characteristic texture and without this

finish a skirt was not considered complete. Sewing the pieces together was done

by passing sinew through holes made with a small awl. Ties at the waist of the

skirt were traditionally made of plaited wild-cotton thread (/etshida) and later of

cotton thread bought from traders.

LOBEDU MATERIAL CULTURE 159

\ VAN ‘

t\ \ yi

\ \ \B

\s DAN na ayy

Vad SAV yet

i. \\

oe

Fig. 96. Skirt, motshe, for a married woman, SAM-9762, Krige Collection, 1936-8.

A number of factors led to a decrease in demand for skin products and a

decline in skin-working technology. The early adoption of European dress by

migrants, the influence of the mission, and the popularity of cloth meant that

fewer people wished to wear skin clothing and fewer men learnt the traditional

skills. Many young men left the rural villages to work in town and the money

they brought back to the villages had a compounding effect on the decline of

skin-working by replacing goats in many traditional transactions and thus reduc-

ing the supply of skins (J. D. Krige 1934: 137, 145; Krige 1941: 17).

Uses of skins and hides

Clothing

Traditionally men wore a loin-covering, thséte, which was a triangular piece

of goatskin that passed between the legs, and strings from the three ends were

fastened in a knot at the back of the waist. A second piece of skin, called

bokone, was attached to the waist-string to cover the buttocks. No other cloth-

ing was worn except a goatskin cloak in winter. In the 1930s it was very rare for

men to wear skin clothing although it was not uncommon for a skin loin-

covering to be worn under a pair of trousers (Krige 1982: 29).

Women’s clothing consisted of a small triangular front apron, thethd, and a

back apron, thebyana (worn mainly by unmarried girls), or a skirt, motshe, worn

by married women and by unmarried girls (Fig. 97). A girl initiate coming out of

seclusion at puberty was given a front apron, a back apron, and a skin cape to

wear over the shoulders. A photograph taken in the 1930s (Krige & Krige 1943,

160 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 97. Young girls wearing skin skirts and waist ornaments. Also note ear ornaments and

hair-styles. Photo: E. J. Krige, 1936-8.

pl. 8a) shows that the hairy side of the skin was worn toward the body and the

other side has a raised nap similar to that found on the motshe.

Skin aprons were important in the traditional exchanges that preceded a

marriage. Among the gifts from the boy’s to the girl’s family was a goat, pudi ya

kebolao, that provided the skin for the groom to prepare a thebyana for his fu-

ture wife. This and a thethd, also prepared by the groom, was worn until after

the birth of her first child. Thereafter the thebyana was replaced by a motshe.

These skin aprons were more than simple items of clothing, they expressed ma-

terially the relationship between the two families. Thus ‘People not only create

their material culture and attach themselves to it, but also build up their re-

lationships through it and see them in terms of it’ (Evans-Pritchard 1940: 89).

By the 1930s cash payments had replaced many of the traditional gifts

(J. D. Krige 1934: 137). Payments of money were explained in terms of ‘thank-

ing’ the girl’s parents for consenting to the marriage. In many cases the groom

was working in town and absent from his village at the time when the prelimi-

nary arrangements for his marriage were being undertaken.

The production of skin clothing steadily declined. Imported cotton clothe

such as salempore and morina had been introduced by traders in the area from

the mid-nineteenth century onward and, together with mission influence and

LOBEDU MATERIAL CULTURE 161

changing economic conditions which encouraged migrancy, led to the eventual

replacement of skin clothing by cloth wraps for women, and trousers and shirts

for men. In the 1970s salempore and morina cloths were still worn by conserva-

tive women. Small skin aprons were retained by old women as under-garments

but most young women wore modern dresses in keeping with their modern life-

style. Skin clothing for general use was something of the past but the tradition

was preserved to some extent in initiation schools. Skin skirts were also worn as

special costume for dances or celebrations.

Hide sandals, zwiéta zwamphashane, were rare in the 1930s and today they

are not made at all. A piece of hard hide was cut to the size of the foot and

strips of skin were passed through slits in the sole at the toe and at the heel to

fasten the sandal to the foot.

Sandals made of rubber from motor-car tyres were popular in the 1930s and

were said to be durable and more comfortable than hide sandals. In the 1970s

these sandals were still made and worn fairly widely.

Baby-slings

A skin sling, thari, was used for carrying a baby on the back (Fig. 98). Tra-

ditionally the skin used for making a thari came from the goat that a husband

slaughtered in honour of the new-born baby. The skin was dressed in the man-

ner described above and no cutting was needed. A nap was raised on the flesh

side using the metal tool, mosédd. The hairy side was worn on the inside against

the body and a flap was folded over on to the side with the raised nap. To this

flap various things might be attached. A thari collected in the 1930s has attached

to it a soft fibrous bulb, /e6me, which was used to wipe the baby’s nose, a small

skin bag of protective medicines, and strings of coloured glass beads, which ex-

pressed the mother’s joy in her child. Most of the skin baby-slings seen in use in

the 1970s had been handed down from previous generations but they were oc-

casionally still made when a child was born. Most women, however, used blan-

kets for carrying their babies.

Bags and pouches

Bags for general use and for special purposes were made of skin. A reptile-

skin pouch, khesigo, was used to keep divining-bones when not in use. Medi-

cines were often carried in small skin bags, dithévhélé, which could be attached

to a girdle or baby-sling. Large skin bags used by men served the purpose of

pockets before trousers were widely adopted (Krige 1982: 30). A set of reed

pipes was Kept in a skin bag made specially for this purpose.

Traditionally snuff was kept rolled up in a piece of soft goatskin, moséha

(Fig. 99), or in a calabash snuff-container (see p. 153) carried in a small pouch

made of dassie skin. Horn snuff-boxes (see p. 167) were introduced by Tsonga-

speakers and in the 1930s were widely used by Lobedu men (Krige notes). Small

tins and bottles were and still are used as snuff containers. In the 1960s a piece

162 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 98. Young girl carrying a baby in a thari. Photo: E. J. Krige,

1936-8.

Fig. 99. Roll of skin for holding snuff, moséha, SAM-9854, Krige Collection, 1936-8.

LOBEDU MATERIAL CULTURE 163

of plastic sheeting rolled up in the manner of the moséha was a popular way of

carrying snuff (Krige 1982 pers. comm.).

Poisoned arrow heads were kept in a hide quiver, khéd6b6, and sharp

knives were kept in a hide sheath, khelada.

Shields

An ox-hide shield, khetagu, was described by an old Lobedu man as being

roughly of hour-glass shape with the narrow section at the grip. No specific in-

formation on making shields was obtained and no examples were seen in the

field or in museum collections.

Skin components of weapons

A sleeve of tail skin from an ox or cow was used for the attachment of a

spear-head to the shaft (see Fig. 112C). The wet skin was pulled over the join

between head and shaft and when dry it formed a tight band that secured the at-

tachment.

The string of a bow was made of a strip of twisted hide (see p. 115).

Thongs

Thongs or riems, mazwa, were used mainly for tethering animals and for

inspanning cattle, a practice adopted from white farmers. They were also used

generally for purposes that needed a strong binding-material. They are still

made when suitable hide is available.

To make a riem, an ox-hide is put into water to soak and when thoroughly

wet is put flat on the ground and cut round and round to give a very long strip of

hide. The strip is thrown over a strong branch or a frame and a heavy stone is

tied to the ends to weight them. In the 1930s bored stones were used for this

purpose (Krige 1982: 31). Using a long stick, the thong is wound up tightly and

then allowed to unwind. This is repeated a number of times while the skin dries

out. The friction causes most of the hair to fall off and fat is applied to help pre-

serve the skin and prevent the thong from becoming brittle.

This method of making thongs is widespread throughout southern Africa.

Other uses of skins and hides

Bellows

Goatskin bellows were used for forcing air into the furnace or forge during

the smelting and smithing process. The two ‘bags’ of the bellows were made

from the uncut skin of the hind legs of a goat (Krige 1982: 31).

Drums

Cow-hide was used for the heads of géma and thithimedzhé drums and

goatskin for the tambourine, khethadada (see p. 123).

164 ANNALS OF THE SOUTH AFRICAN MUSEUM

Mats

Unworked but cleaned goatskins were used by some as sitting-mats. They

were not as common as grass mats (Krige 1982: 31).

HORN AND BONE

The care of livestock was predominantly a male activity as was work on

horns and hides. No specialized skill was required to convert horns into con-

tainers of various kinds and most horn artefacts utilize the natural form of the

horn without any additional working. No artefacts were carved entirely from

solid horn. Horn snuff-boxes were made by using both the solid end-section and

the natural hollow of the horn.

Bone was not used to a large extent and when it was used the natural shape

of the bone, or piece of bone, was modified to suit the particular function. Solid

bone was not used as a carving medium.

Uses of horn and bone

Medical equipment

1. Cupping-horn

A cupping-horn, mothoho (Fig. 100A), consists of a small horn or end-

section of a larger horn perforated at the solid tip so that it is open at both ends.

Beeswax (modu) is used to seal the narrow opening during the cupping-process.

Cupping or blood-letting, -loméha, is practised to cure headaches or pains

believed to be caused by impure blood. Most people know how to use a cup-

ping-horn and the operation does not have to be carried out by a diviner. Two

small incisions are made in the affected area over which the wide end of the

horn is placed. Air is sucked through the narrow end and blood is drawn into

the horn by the vacuum. The tip is sealed with wax and blood continues to flow

into the horn until the operator breaks the vacuum and removes the horn. The

blood is buried in a secret place for fear of it being used in witchcraft.

2. Medicine-horns

Herbalists store medicines in ox-horns (Fig. 100B). Rain-water for use in

special rituals is also kept in horn vessels. Horns used in this manner are not

carved but many have holes drilled near the rim for the attachment of a thong.

3. Divining-bones

A set of Lobedu divining-bones consists of about forty pieces, the most im-

portant of which are the four flat carved bone or ivory pieces, two male and two

female, which fall in sixteen different combinations, the cardinal positions. The

set includes a number of pieces not made of bone, for example, shells, river

pebbles and marula pits, as well as pairs (male and female) of bones from ani-

mals associated with totems. The basic principles of divination are outlined by

Krige & Krige (1943: 226-227).

165

LOBEDU MATERIAL CULTURE

ine-horns at

1c

. Med

B

llage, 1973

horn in use, Headman Mohale’s vi

Headman Mohale’s village, 1973.

Fig. 100. A. Cupping

S Vl

166 ANNALS OF THE SOUTH AFRICAN MUSEUM

Musical instruments

1. Trumpet

The traditional horn for making a phalafhala trumpet (Fig. 101A) was that

of the sable-antelope from which the instrument derives its name, but kudu

horns were also used. The horn was left to dry out until the core became loose

and was easy to remove. An embouchure was carved in the side of the horn at

the point where the hollow is narrowest, i.e. just before the solid tip starts. If

the horn showed signs of splitting it was strengthened with a skin sleeve to pre-

vent further cracking. Horn trumpets are seldom made today but a number of

old ones, belonging to district headmen, are still in use.

The phalafhala produces only one note but when blown hard it is loud and

carrying. It was and still is associated with authority and it sounded as a signal or

to make important announcements, for example to warn of a veld fire, to sum-

-mon people to meet at the capital, to herald the arrival of a party of gosha

dancers, or to announce the arrival of tribute sent to the Queen.

2. Whistles

Small antelope horns were made into whistles, dinaga (Fig. 101B). The

solid core was removed, leaving the horn-sheath closed naturally at one end and

open at the other. This formed a simple stopped pipe and sound was produced

by blowing across the open end. Horn whistles were used in hunting and by

herd-boys. They are seldom made today as game is protected and boys spend

relatively little time herding or playing in the veld. Whistles were also made

from bone, often the hollow tibia of a small animal. One end was stopped and

the whistle was played by blowing across the other end. According to Krige &

Krige (1943: 135) during the vyali the Bird of Muhale communicates by

whistling through an instrument made of bone but the Bird is also said to use the

mphenyi made of reed (see p. 99).

Fig. 101. A. Horn trumpet, phalafhala SAM-9822. B. Horn whistle, nagana, SAM-9818.

(Both Krige Collection, 1936-8.)

LOBEDU MATERIAL CULTURE 167

Pipes

In the past dagga, badje, was smoked through a water-pipe that consisted of

a stone pipe-bowl, gagana (see p. 146), a reed, khetoméla, and an ox-horn

water-vessel, naga la khomo. The ox-horn did not require special preparation.

The dagga was placed in the bowl and the reed formed the pipe-stem that was

inserted into the horn of water so that the smoker could draw the smoke

through the water when he inhaled deeply at the mouth of the horn.

Snuff-boxes

Horn snuff-boxes were introduced by Tsonga-speakers and their use was

adopted mainly by Lobedu men. The form of the snuff-boxes (Fig. 102A-—B) fol-

lows the natural shape of a horn. A hole is bored from the existing hollow

through the solid tip of the horn. The wide part of the opening is then closed

with a plug of wood or gourd and the narrow mouth is fitted with a small stop-

per. It is not uncommon for a horn snuff-box to have a carved ridge near the

mouth and a thong tied below the ridge. In an unusual example (Fig. 102A) col-

lected in the 1930s a human figure has been carved following the natural shape

of the horn.

Tools

1. Rolling-tool

An ox-horn, flattened on one side, is used as a tool, khehdthod, to coil wire

round a foundation of tail hair in making wire arm- and legrings, maséga.

Fig. 102. Horn snuff-boxes.

A. SAM-9850. B. SAM-—9851.

(Both Krige Collection, 1936-8.)

168 ANNALS OF THE SOUTH AFRICAN MUSEUM

The tail hairs, maditsi, are first cut, twisted, and the ends joined to form a

circle. The wire is passed through a small split stick and wound a few times

round the hair circlet, which is then placed on a wooden plank and stroked,

-hota, in a downward movement with the horn tool so that the wire rolls evenly

on to the base of hair (Fig. 103). In the 1930s brass or copper wire was bought

from traders and today this had been replaced by aluminium wire. Before wire

could be bought it was made by an arduous process of drawing hot metal

through a perforated plate (see p. 172).

Traditionally women wore large numbers of maséga on their ankles (see

Fig. 109) and wrists but these ornaments have become less common. Legrings

are said to decrease mobility and there is an increasing tendency for young

women to adopt European dress.

Fig. 103. Horn tool used to roll wire on to foundation of tail hair in making maséga bangles,

Molototsi valley, 1976.

2. Marula tool

A bone tool, /efhéd6 (Fig. 104), is used to remove the tough skins of the

‘marula fruits from the flesh prior to making marula-cider, mokhobe. The exist-

ence of a special tool for this purpose and of a special bowl (see p. 64) for

drinking marula-cider is an indication of the importance of the marula fruit to

the Lobedu. The fruit ripen in late January and February when crops are imma-

ture and there is little grain left over from the previous harvest. The marula

season is, therefore, eagerly awaited. Marula-cider is rich in vitamin C and a

source of nourishment as well as being a refreshing drink. It is usually made by

women in the fields under the marula trees where it is also consumed with much

enjoyment.

LOBEDU MATERIAL CULTURE 169

Fig. 104. Bone tool, lefhéd6, for removing the skin from marula fruit,

SAM-9746, Krige Collection, 1936-8.

The ripe fruits that have fallen to the ground are used to make the cider.

The tough skins are removed with the /efhéd6 and the juice is squeezed from

the flesh with a kneading action, -soha. The pulp and the pits are placed in a

large clay pot, thukhwana. Water is added and the fruits are rubbed, -suha

marula, to extract all the juice, which is then ladled into another pot, leaving the

pits behind.

At this stage the juice, lesebeta, is left to ferment. After one day it is still

very sweet and is known as /edutu. A layer of scum, ledobe, which has risen to

the top, is removed. A bowl is placed over the mouth of the pot and the edges

are sealed with some of the /edobe. Another layer of scum rises to the top of the

liquid and is removed before the cider is ready for drinking. The longer it is al-

lowed to ferment, the less sweet and more intoxicating the cider becomes. It is

usually allowed to ferment for about three days before it is drunk.

Water is added to the pits that remained behind and is decanted on the fol-

lowing day to provide a dilute children’s drink, malithele. Thereafter the pits are

left to dry and are cracked open later for the kernels, which are ground and

added to vegetable relish.

3. Plastering-tool

The flat rib bones of cattle are recorded as having been used at Medingen as

trowels when plastering walls (Klapwijk Collection, Transvaal Provincial Mu-

seum Service K108A).

DUNG

Cattle-dung was mixed with mud to form a kind of cement used in hut con-

struction for the plastering of walls and floors and for hardening the floor of the

courtyard. The mixture of dung and mud was spread over the floor surface and

compacted with a wooden beater (see Fig. 55) or a smooth stone, thidéld. In the

1930s decoration of floors and walls was rare (Krige 1978 pers. comm.) but in

the 1970s finger-patterns made while the last layer of dung was still wet were

very popular at the capital and elsewhere in the capital district, and gave an at-

tractive finish to the floor.

170 ANNALS OF THE SOUTH AFRICAN MUSEUM

Formerly dung was used to line the inside of grain pits (see p. 140) and to

fill in the spaces between the reeds of the door of a cooking-hut (see p. 85).

Dung was not generally used as fuel nor was it widely used as a fertilizer.

Although it was realized that adding dung to the soil improved the crop yield, it

was said also to encourage the growth of weeds, which discouraged its use

(Muneri Modjadji to Holloway Commission 1930).

SUMMARY AND COMMENT

Domestic resources, the products of cultivation and livestock husbandry,

are secondary resources in that they depend on the natural resources in the form

of arable land and grazing, and on agricultural and pastoral technology. In the

above section the uses of gourds, skins, horn, and bone are described. Fruit-

shells and horns of game animals, although natural resources, are included as

their technical preparation is similar to that of gourds and cattle-horns respect-

ively.

In general the processes for which women were traditionally responsible

have remained more stable than those carried out by men, and this is so in the

case of calabash-making. There is no specialization in this craft as most women

know how to cultivate gourds and make a variety of utensils. The continued use

of calabash vessels for serving porridge seems to be related, as in the case of

cooking-utensils, to the constant dietary pattern and the continued demand for

calabash utensils well-suited to their varied domestic uses. It is not uncommon

for enamelware to be used concurrently with calabashes for storing dry foods

but calabash vessels have not been replaced for serving porridge, decanting

water and, in some situations, for drinking beer.

The long-handled beer-ladle, which is passed from person to person at

social beer-drinks, can be seen as expressing generosity and good social rela-

tions. At beer-drinks where bought beer tends to be consumed by individuals,

the calabash ladle is not appropriate and the glass jars or tins of standard

volume that are used instead reflect materially the commercial principle under-

lying this kind of beer-drinking.

. The most important product of livestock husbandry that was used in the tra-

ditional technology was goatskin. This provided the basic raw material for skin

clothing. Men processed the skins and manufactured the garments for domestic

use. By the early twentieth century, however, production of skin clothing had

started to decline. Migrancy imposed the need for men to adopt European

dress, the mission influenced the clothing of those who adopted the faith, and

these early pressures coincided in the late 1890s with the rinderpest, which

almost annihilated all the livestock of the area.

Migrant labour had a further effect on the decline of skin-working by reduc-

ing the number of craftsmen. Money that migrants brought back into the area

had a compounding effect by replacing goats in many traditional transactions,

thus reducing the supply of skins.

Well into the twentieth century women continued to wear skin clothing as

LOBEDU MATERIAL CULTURE 7A

well as the cotton cloth wraps that had become ‘neo-traditional’ since their wide-

spread adoption and popularity in the nineteenth century. Even as the use of

cotton clothing increased, older women often wore a skin apron underneath a

cloth skirt or wrap. By the 1970s, however, skin clothing had fallen almost en-

tirely into disuse. Old skin skirts were brought out for ceremonial occasions but

new ones were seldom made and few men remembered the former techniques of

making skin garments.

Among the changes that paralleled the move from skin to cotton-cloth

clothing were the replacement of the goatskin baby-sling by a lighter cloth ver-

sion that did not have the disadvantage of being smeared with fat, which was

absorbed by the wearer’s cloth wraps, and the decline in use of skin bags by men

for carrying personal possessions in favour of trouser pockets.

Horns traditionally used as containers or flasks, especially for medicines,

were still used for this purpose in the 1970s. The rain medicine, mofugo, used in

a ritual of appeasement in 1973 (see p. 101) was carried to the site of the cer-

emony in an ox-horn flask and returned afterwards to the home of Modiga who

was responsible for its safe-keeping.

Horn also continues to be used in the form of the tool khehdéthé, made

from a flattened horn-core and used for coiling wire round a circlet of hair to

make a bangle. This tool is so well adapted to its function that it is difficult to

think of an alternative that would work as well. Equally well adapted is the bone

tool, /efhéd6, used to remove the tough outer skin of marula fruit prior to mak-

ing mokhobe cider. The use of a specific tool for this purpose and of a special

bowl, kheritswana (see Fig. 13C) for drinking mokhobe is an expression of the

social and dietary importance of the marula fruit.

Although not as extensively exploited in the technology as the natural

environmental resources, domestic resources supplemented the indigenous

materials and were essential to the traditional material culture.

USE OF IMPORTED MATERIALS

METAL

Metal ores were not mined in the Lobedu area. The closest source of iron

ore, gwedzi, is some distance to the north-east, and when metal-working was an

active industry ore had to be carried on foot to the Lobedu smelting-sites. Krige

(1941: 9) recorded that unfortunates who had no other way of earning hoes or

cattle to acquire a wife carried iron-ore on foot-paths from a ‘long way to the

north’ (probably the Tshimbupfe area) in return for a hoe per journey. The

location of smelting-sites some distance from the ore source is not unusual in the

Transvaal Lowveld. Archaeological evidence from the Phalaborwa area indi-

cates that ‘it is clear that smelting-sites were located near the centres of demand

and distribution, rather than near the ore-supply’ (Van der Merwe & Scully

L971 182).

2 ANNALS OF THE SOUTH AFRICAN MUSEUM

Copper came into the Lobedu area from Messina and Phalaborwa, which

were important centres of copper production in the pre-colonial period (Wilson

1969: 173-174). According to Van Warmelo (1940: 4) until the mid-nineteenth

century there was ‘a constant stream of travellers going to and coming from the

copper mines at Messina, where they bartered for the precious metal’. Tin is

known to have been mined in the Rooiberg area and may have been traded over

a wide area.

Early trade through Delagoa Bay was a source of iron, copper, and brass,

and in the late nineteenth and early twentieth centuries European trade goods,

including iron hoes, tools, and wire, were brought into the area in much larger

quantities than before. During this period and later into the twentieth century

another source of iron and metal goods was importation by migrant labourers

returning from work on the mines.

_ Metal-working techniques

‘In the old days . . . iron making was one of the great industries’ (Krige &

Krige 1943: 47). By the 1930s this industry had ceased largely because of two re-

lated factors—the appropriation of the ore sources by whites and the introduc-

tion of cheap factory-made metal objects, often imitations of the indigenous

products, which rapidly undermined local production.

A number of Lobedu smelting-sites are said to exist. So far none has been

excavated but pieces of iron-ore, gwedzi, and slag from the furnaces were col-

lected by Krige (SAM-—9889, 9890) from a terraced site near Modjadji’s Nek. At

this particular site there is also clear evidence of habitation in the form of hut-

floors and daga. No historical record of Lobedu smelting-technology has been

found by the author and men interviewed during fieldwork had only the vaguest

memory of the processes involved. It was said that the siting of a furnace was

important both in relation to the fuel supply and the prevailing winds. It was

also necessary to know the types of wood that would produce very hot coals.

Goatskin bellows with clay nozzles were said to have been used to keep the fire

burning for many hours but technical details were not remembered.

Metal-working was a specialized industry and there is no record of the ex-

tent of Lobedu involvement in the production of metal goods. The Lobedu ob-

tained many of their metal artefacts through trade with the Phalaborwa, Venda

and Lemba, and locally from Lemba craftsmen living in the Lobedu area. The

tools (Fig. 105A—D) of a Lemba smith who was over 80 years old in the 1930s

are preserved in the Krige Collection. They include a hammer, nddo, used for

forging metal implements; tongs, Jomand, for handling the red-hot metal; a

draw-plate, Jeg6khd, used in wire-drawing; and a vice, bagd. Also in possession

of the smith was a tin ingot, /ekatea, which was said by him to have been used in

‘joining metals’.

Wire was made by heating a piece of soft metal, usually copper or bronze,

and shaping it so that it could pass through a hole in the draw-plate. The vice

securing the piece of metal was wedged in the fork of a tree while the wire was

LOBEDU MATERIAL CULTURE 16/3

D

Fig. 105. Tools that belonged to a Lemba smith. A. Vice, bag6, SAM-—9898. B. Draw-

plate with holes of different size for wire-making, legd6khé6, SAM-9896. C. Hammer,

nod6, SAM-9895. D. Tongs, lomanéd, SAM—9894. (All Krige Collection, 1936-8.)

drawn out with the plate. According to Stayt (1931: 65) a forked stick was used

to lever the plate as the wire was drawn out and wound round the tree. For thin-

ner wire the process was repeated using a smaller hole in the plate.

Uses of metal

Hoes

Before the introduction of the plough, hoes, mazhébé, were the most

important agricultural implements and their use has continued in areas unsuit-

able for ploughing. In addition, hoes were important as exchange media especially

in marriage contracts. Hoes so used established legally binding obligations

between the kin groups involved (Krige 1964: 162).

Hoes are said to have been forged locally but many were traded from neigh-

bouring people. In the early nineteenth century the Venda were exporting hoes

174 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 106. Woman hoeing in the fields. Photo: E. J. Krige, 1936-8.

‘in thousands to neighbouring districts’ (Beuster 1879: 239) and the Phalaborwa

people were also trading extensively in hoes. It is possible that specialists con-

trolled the production of hoes as in the case of a Phalaborwa iron-worker who

‘taught others to produce iron but retained exclusive monopoly over the knowl-

edge of hoe-making’ (Scully 1978: 254). Although there must have been fluctua-

tions in the exchange rate of hoes through time and in different social and |

economic circumstances, it was only after the large-scale introduction of cheap |

factory-made hoes in the late nineteenth century that the traditional value of

hoes seriously declined.

The form of hoe-heads collected in the Lobedu area in the 1930s is roughly

diamond-shaped with a faint midrib running down the centre. Venda hoes illus-

LOBEDU MATERIAL CULTURE 7S

trated by Stayt (1931, pl. 29) are of similar shape. For use in the fields (Fig. 106)

hoe-heads were attached to a long wooden handle. In general it seems that hoes

used in agriculture did not differ in shape from hoes used for exchange, except

that the latter were new and not worn down through use. There is some evi-

dence, however, of heavier, elongated spade-shaped hoes being used solely for

exchange. “These hoes were too large and heavy for tilling the soil and were

merely kept in a hut’ (Krige 1964: 162).

By the 1930s hoe-heads (Fig. 107) of local manufacture had come to be re-

garded as heirlooms that had links with the ancestral spirits. They were often

partly embedded in small clay-mound shrines where offerings were made to the

ancestors.

Fig. 107. Old hoe-head kept for use in ancestor

rituals, Mohale’s village, 1973.

176 ANNALS OF THE SOUTH AFRICAN MUSEUM

Ornaments

Copper and its alloys were used for making solid neckrings and bangles,

mefhiri (Fig. 108A), beads, zwirogaana (Fig. 108B) and coiled wire arm- and

legrings, masega (Fig. 109). Ornaments made of hand-forged metal were rare

even in the 1930s and are now used as amulets and regarded as ‘thugula’ objects.

Kherogaana beads of the ‘wrap-around’ type, i.e. made from a narrow strip of

hand-worked copper cut and bent to form a circle, are among the oldest Lobedu

ornaments. Individual metal beads in the Krige Collection (SAM-—9910) are said

to have been handed down from generation to generation as highly valued

objects. Similar beads, dithotho (SAM-—9911), made in the same way from factory-

produced copper and brass were much more plentiful and less valued than the

ancient copper beads. Dithotho beads were strung in large numbers on cotton

thread to form beaded strands with tassels at the ends and were worn attached

Fig. 108. A. Solid metal bangle and neckring, SAM—9908 and 9909. B. Wrap-around metal

beads on thread foundation, SAM-9913. (All Krige Collection, 1936-8.)

LOBEDU MATERIAL CULTURE M77

Fig. 109. Rolled wire legrings, maséga, Modjadji’s village, 1973.

to skin skirts. More rarely the older kherogaana beads were worn in strings by

those who had enough of them (Krige 1983 pers. comm.).

The mefhiri neckrings and bangles collected in the 1930s are made of brass,

which suggests that they are not as old as the copper kherogaana beads although

they are said to have been of local manufacture. They were rare in the 1930s and

in the 1970s were not seen in use at all.

Coiled arm- and legrings are still made (see p. 167) but wire of copper and

brass has been replaced by aluminium wire and these ornaments are not worn in

such large numbers as in the past.

Tools and weapons

When metal-working was an active industry the iron heads for tools and

weapons were locally produced. Later scrap metal was used and forged into the

required artefacts. More recently, purchased tools and weapons have come into

use. Many of the traditional tools used in wood-working (see p. 101) and skin-

working (see p. 158) had characteristically shaped iron heads that were related

to their particular technical uses. The action of the craftsman in handling the

tool was taken into account when forming the head and when hafting it. The

long-handled gouge, /ehdr6 (Fig. 46), is a good example of this.

The following iron-headed tools and weapons (Figs 110-112, see also Fig.

61) were collected in the 1930s: adze, bédwana; gouges, mahdér6; poker, moro;

awl, morogola; nap-raiser, mosédo; knife, mofhaga; spatula, lehobd, for wiping

perspiration from the face; spear, lerumd; axe, zaga; and arrow-heads,

mesévhé. Artefacts, locally smelted and forged, were rare in the 1930s. They

178 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 110. A. Poker, mor6, SAM-9725. B. Awl, morogola, SAM-9742.

(Both Krige Collection, 1936-8.)

ue

Aut

Fig. 111. A. Knife, mofhaga, SAM-9848. B. Sweat-scraper, leh6b6, SAM-9887.

C. Skin-working tool, moséd6, SAM-9764. (All Krige Collection, 1936-8.)

were highly valued as thugula heirloom objects even if rusted and in bad repair

and they were thought of as links with the ancestors. These beliefs are still held

by many Lobedu.

GLASS AND SHELL

A distinction should be made between Lobedu ancestral beads, which prob-

ably pre-date the eighteenth century, and glass beads of more recent origin.

Ancestral beads are said to have been brought from north of the Limpopo River by

the founding ancestors, an oral tradition which is substantiated by comparative

chemical analysis of similar beads excavated at the Rozwi ruin at Dhlo-Dhlo

(Zimbabwe) which was occupied around ap 1700 (C. Davison & Desmond

Clark 1976: 132-137). Lobedu ancestral beads include those known as modala,

modzike and thanganwe, that are similar in appearance and in chemical

LOBEDU MATERIAL CULTURE 179

Abe?

Ret Se

EG Wa

“Qs

ee

eG

= "Ge .

Fig. 112. A-B. Axe-heads, SAM-—9884 and 9885.

C. Spear-head and butt, SAM—9843. Note skin sleeve

for attachment of head to shaft. (All Krige Collection,

1936-8.)

180 ANNALS OF THE SOUTH AFRICAN MUSEUM

composition to the Venda ‘beads of the water’, vhulungu vha madi, described by

Van Riet Lowe (1937: 368), Dicke (1937: 406), and Schofield (1958: 189-192).

Ancestral beads, presented by Krige in 1937 to the Archaeology Department of

the University of the Witwatersrand, were seen by the author. They include

blue-green, modala, cobalt blue, modzike and thanganwe, red, mokhwibidu,

green, lethodi, yellow, thaha, and opaque white, khefhago. Many are small cane

or snap beads having a waxy, translucent quality. The dark cobalt-blue beads

tend to be larger. These beads and other ancient beads were and still are

regarded by the Lobedu as thugula or sacred objects.

Fig. 113. Beaded ornaments. A. Necklets, khekhadwa, SAM-9781, and leda-

bara la dithséhéla, SAM-9780. B. Armbands, zwifhédd6, SAM-9782. (All

Krige Collection, 1936-8.)

LOBEDU MATERIAL CULTURE 181

Glass beads acquired through trade in the late nineteenth and twentieth

centuries differ in appearance from ancestral beads in lacking the distinctive

translucent quality and characteristic colours; they are also far more abundant.

Whereas heirloom beads were usually treasured individually and used as amu-

lets, later beads were very plentiful and were made into a variety of ornaments.

Nineteenth-century cane beads were made into a many-stranded necklet,

khekhadwa (Fig. 113A), which was worn traditionally by a young bride when

she first came to her husband’s home. This necklet was also worn by recently in-

itiated boys and girls in recognition of their new status (Krige 1982: 34). Charac-

teristically the khekhadwa is strung so that it has four bands of colour composed

of many strands of beads, usually green, madalayane, blue, moddgwa, yellow,

khirulwana, and black, lesoli. Also worn traditionally by a young bride (Krige &

Krige 1943, pl. 12b) were zwifhéd6 armbands (Fig. 113B). The beads used were

small, rounded oblates, mainly black, white, and dusky pink. They were strung

in rows between pieces of hide about 13 cm in length. Old examples were strung

on sinew; in more recent specimens cotton string has been used. Zwifhéd6 arm-

bands were also worn for dances and celebrations, together with waist orna-

ments, zwibébédana, which consist of multi-coloured beadwork flaps worn over

a skin skirt (Krige & Krige 1943, pl. 7b). These bead ornaments are no longer

made but many Lobedu women still have old pieces in their possession. The

large number of examples in the Klapwijk Collection suggests that they were

widely used well into the twentieth century.

In the late nineteenth and early twentieth centuries ostrich eggshell beads

were brought by traders into the Lobedu area. Migrants returning from the

Fig. 114. Girdle of ostrich eggshell beads, mofthaga, K102A, Klapwijk Collection.

182 ANNALS OF THE SOUTH AFRICAN MUSEUM

mines could buy these beads in the Munnik and Soekmekaar areas which were

said to have been centres of manufacture (Klapwijk Collection K102A, C) as

were the Phalaborwa and Venda areas. Beads could be bought loose or in

strings, some were roughly cut, others smoothly polished, and they were made

up into mothaga girdles and worn by Lobedu women and young girls. The mo-

thaga (Fig. 114) is a single or double string of ostrich eggshell beads with a single

seed of the wild banana (modolo) at the centre back. Worn over a skin skirt tied

in the front with the seed always at the back, the girdle was thought to have pro-

tective qualities for the wearer and to enhance fertility. They were made in large

numbers and from 1900 onwards factory-made white glass beads were imported

by traders to replace ostrich eggshell beads (Klapwijk Collection K102A).

Cowrie shells from the coast were traded in the Lowveld and worked into

ornaments (Fig. 113A). They were thought to bring good fortune to the wearer.

The cylindrical central spire of the cone shell was made into an ear ornament,

dibadana (Figs 115-116). It was worn by women through the ear lobe in the

manner illustrated by Tyrrell (1968: 62). Early dibadana are said to have been

made of ivory (Krige 1982: 35); the shell version replaced the ivory, and porce-

lain copies were introduced by traders in the early twentieth century. Most, but

not all, dibadana have a hole through the centre. They were popular in the

1930s and many photographs from that period show women wearing them. In

the 1970s they were not seen in use but a few old women still had them in their

possession.

Ee

Fig. 115. Ear ornaments, dibadana, K68D, Klapwijk Collection.

(yo

LOBEDU MATERIAL CULTURE 18

x

Fig. 116. Woman and child. Note hair-style, ear ornaments and clothing, also

reed door of hut. Photo: E. J. Krige, 1936-8.

CLOTH

Machine-woven cotton cloth from India and Europe was brought into the

Lowveld by traders following the well-established routes from the port of Dela-

goa Bay to the interior. Striped salempore gained early acceptance among the

Lobedu as it did among the Venda who termed it nwalukambu. German print

184 ANNALS OF THE SOUTH AFRICAN MUSEUM

introduced by the mission became popular and has remained so among older

women. Plain black morina was also an early trade cloth and it is still worn by

conservative Lobedu women. In the 1970s morina and salempore wraps, dugu

yanogeni and dugu yahopaketa, worn around the waist and across the shoulders

(Figs 116-117) were often braided with coloured cloth at the edges and worn

with a cloak, seyi, of coloured cotton. Local tailors took orders for machine

braiding (Fig. 118). There was a wide range of cotton cloth available in the local

stores but salempore and morina were still in demand by women who had not

adopted modern dress.

Fig. 117. Young girl with partly shaved head wearing salempore cloth,

and child wearing bead necklet, grass armbands and metal anklets.

PhotosE ve Ktiees 193658:

LOBEDU MATERIAL CULTURE 185

Ee ie al

Fig. 118. Local tailor, Lebiya’s village, 1975.

Members of the malé6b6 cult wore distinctive wraps of red and white cloth.

They said that these colours were favoured by the spirits.

Cotton head-cloths, mothini, common in the 1970s, were not worn in the

1930s except among the Christian community. Most women shaved their heads

completely but some allowed their hair to grow and styled it in particular

fashions. Some young women favoured leaving a crown of hair on the top of the

head, others shaved the hair to form certain styles (Figs 116-117), or parted it in

a number of bands (Tyrrell 1968: 63, 65).

SALT

A small amount of salt was extracted locally from the leaves of a certain

plant that were dried, mixed with water, filtered and the filtrate boiled away

until the salt remained. Most of the salt used by Modjadji’s subjects came from

the salt-pans in the Soutpansberg to the north and from Tsonga territory to the

east and was acquired through trade. Salt from pans was extracted in a filtration

process that can still be observed at Sautini, a hot spring on the confluence of

the Klein Letaba and Mamostwapi rivers.

The process is similar to that recorded by Junod in the early twentieth cen-

tury. The salt-laden earth is filtered through large grass-lined filters (Fig. 119A)

and the filtrate is evaporated, leaving the salt crystals, which are shaped into

small mounds for trading (Fig. 119B). This salt is much preferred to the factory-

processed product. Salt is added to the many relishes that accompany the rather

bland unsalted porridge.

186 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 119. A. Salt-filter made of mopane bark and branches, Sautini, 1973. B. Small mounds

of salt drying over warm coals, Sautini, 1973.

LOBEDU MATERIAL CULTURE 187

SUMMARY AND COMMENT

The above section is concerned with materials that were not indigenous to

the environment and therefore had to be obtained through trade or by travelling

outside the area. The most important material in this category is iron.

Iron was essential to the subsistence economy in that the exploitation of

both natural and domestic resources depended on the use of iron tools. Iron was

needed for the production of the basic agricultural implement, the hoe, as well

as for weapons and tools on which the technology was dependent. The fact that

this essential resource was not found in the Lobedu area gave it a high social and

economic value. Long distances were travelled to obtain ore, and iron objects

were valued items of trade. Throughout the Lowveld hoes were used as media

of exchange in trade and in marriage contracts.

Before the local metal-working industry had been largely undermined by

the appropriation of ore sources and the large-scale introduction of factory-

made copies of indigenous products, iron was imported from the Phalaborwa

and Venda areas both in the form of ore to be smelted and forged locally, and in

the form of complete artefacts ready to be hafted for use. By the late 1930s the

smelting-process was no longer practised in the Lobedu area (Krige 1982 pers.

comm.) and by the 1970s it was not even possible to record the complex pro-

cesses formerly used. Forging iron objects, -thula tsipi, from any available pieces

of metal continued for much longer and is still practised on a small scale by a

few craftsmen (often Tsonga-speaking immigrants) using scrap metal and

modern tools (Krige 1982 pers. comm.).

In the 1930s objects made of locally smelted and forged iron were rare and

by the 1970s they were extremely rare. Iron tools and weapons that have not

rusted away completely have been kept for religious purposes as sacred thugula

objects believed to have links with the ancestral spirits. When beer is offered at

a family shrine, it is poured over the thugula object, which is either embedded in

the shrine or placed on it to represent its deceased owner and, by association,

the ancestors in general.

By the 1970s purchased metal tools such as knives, saws, axes, nails, files,

drills, hammers, and pliers, were widely used, especially in making non-

traditional furniture such as chairs and benches, and in construction-work.

Window-frames, doors and roof structures were made using modern carpentry

tools and techniques. Although from the late nineteenth century carpentry

skills had been taught to the Christian community at Medingen, these skills had

not been widely implemented among traditionalists. Huts did not have win-

dows, and the doors were either of reed or carved from a solid piece of wood.

As mud-brick construction replaced the older pole-and-daga method, the struc-

ture of huts was subject to further innovation in the form of modern door-

frames and windows. The construction of these fittings required the use of

modern tools.

Traditional tools, however, were appropriate to their specific functions and

bédwana adzes and leh6r6é gouges (usually with blades forged from scrap metal)

188 ANNALS OF THE SOUTH AFRICAN MUSEUM

are still used for carving drums or utensils from a solid piece of wood. The

important and particular relationship of tool form to artefact is apparent in this

continuity of technique, despite the availability and concurrent use of modern

tools.

Copper, which came from ore sources in the Venda and Phalaborwa areas,

was not essential to subsistence activities but was valued as a material for

drawing wire and for making beads. As in the case of ancestral iron objects,

copper ornaments that pre-date their mercantile equivalents have become

sacred thugula objects. Also in this category are the very old glass beads, some

of which are believed to have been brought to the area by the founding ances-

tors.

Krige (1982: 32) observes that a marked characteristic of Lobedu

material culture is the lack of elaboration in dress and ornamentation. She

relates this to the fact that the decorative aspect of ornamentation is

_ secondary to the association of beads with the ancestor cult and to their

protective and healing powers. Secondly, she draws attention to the fact

that, in contrast to the practice among exogamous Nguni groups, dress and

ornamentation was not an important part of courtship as, following the pre-

ferred pattern of cross-cousin marriage, most people grew up knowing

exactly whom they would marry. Ornamentation was thus only partly dec-

orative in intent and consequently modern glass beads have had relatively

little popular appeal.

Similarly simplicity characterized traditional dress even after trade cloths

were adopted. The striped salempore, German prints, and black morina cot-

ton cloths, which came into use during the nineteeth century, seem to have

been accepted more for their practical advantages in a warm climate than

for their decorative qualities.

The replacement of skin clothing by cloth wraps and modern garments

meant that people were dependent on an imported purchased commodity for

their basic clothing requirements. The traditional cloth wraps required little

or no sewing but, as modern dress became popular firstly among the Chris-

tian community, there was a need for local tailors and dress-makers (Fig.

118). This involved the use of a whole range of other imported goods from

sewing-machines to needles and cotton.

In the 1930s, with the important exception of metal goods, the balance

between locally produced and imported material items favoured local pro-

duction. Material innovations were not widely accepted except by the Chris-

tian community, and the subsistence economy was still viable. By the 1970s

the transition from a subsistence to a money economy was clearly reflected

in many aspects of the material culture. In addition to purchased alterna-

tives for many traditional objects, motor cars, bicycles and a wide range of

brand-name products were ubiquitous. It is relevant to consider the resili-

ence of the traditional technology in the face of fundamental economic

change.

LOBEDU MATERIAL CULTURE 189

MATERIAL CULTURE AND SOCIAL CHANGE

The present study, undertaken in the mid-1970s, aimed not only to describe

the material culture but to analyse changes that had occurred during the 40 years

since the Krige Collection had been made. The post-war period coincided with

an unprecedented economic development of the area, growth of light industries,

improvements in roads and communications, and increased white settlement fol-

lowing the virtual eradication of malaria. The black population of the area in-

creased far beyond the carrying capacity of the land and the domestic economy

became increasingly monetized. It is against this background of changing social

and economic conditions that changes in the material culture will be viewed.

In discussion of these conditions the author is indebted to E. J. Krige (1975,

1981) for her analysis of social change among the Lobedu, as well as for exten-

sive personal interviews and correspondence over the duration of the study.

Without her analysis based on data from both the 1930s and the period 1960-80,

it would not have been possible to gain insight into the interrelated nature of

material and social change.

HISTORICAL, ENVIRONMENTAL AND SOCIAL CHANGE, c.1880—1980

The final decades of the nineteenth century were significant in Lobedu his-

tory. During this period conflict over land alienation and taxation imposed by

the Volksraad of the Zuid-Afrikaansche Republiek increased, and led to con-

frontation and eventual Lobedu submission to imposed territorial arrangements.

The fixing of the boundaries of the Modjadji Location in 1892 greatly reduced

the arable land and grazing available to Lobedu farmers. White farmers settled

along the Mokeetsi River and on the land south and west of the Location

boundaries toward present Tzaneen. This dispossession of land coincided with

the loss of cattle through diseases, the rinderpest of 1896 being the most serious.

Few cattle survived this and a subsequent drought further strained the subsis-

tence economy and gave impetus to labour migration which had occurred on a

small scale from as early as the 1860s (Delius 1980: 296). Taxes payable in cash

forced men to enter the labour market, which had far-reaching social and ma-

terial implications. European dress was soon adopted by most men and the ac-

quisition of expensive items such as guns, ploughs, bicycles or sewing machines

was an additional incentive to earning money.

At first men stayed away from their home villages for fairly short periods,

and the percentage of men absent from the reserve at any one time was rela-

tively low, but the length of time away and the number of migrants increased

steadily. In the late 1930s migrants formed approximately 37 per cent of the

adult male population, by the end of the 1970s the figure was higher than 65 per

cent (Krige 1981: 181) and, although figures are not available for the 1980s,

there is no doubt that the number of migrants far exceeds the 1970s estimate and

continues to increase and includes growing numbers of women. In the 1980s

population density is such that the vast majority of people have no land to culti-

vate and are entirely dependent on the earnings of migrants.

190 ANNALS OF THE SOUTH AFRICAN MUSEUM

From the turn of the century onward migrancy not only brought money into

the rural economy and created new material needs but changed the balance of

labour within the Lobedu domestic economy by reducing the number of male

workers. Unlike the division of labour in Nguni cattle-based economies, in the

predominantly agricultural Lobedu economy men participated actively in crop

production. The reduced agricultural work force was to some extent compen-

sated for by the use of the plough, rejected in its early heavy form but accepted

later in a lighter form (Krige & Krige 1943: 325). Hoe-cultivation, however,

continued in many places as much of the Modjadji Location is too steep for

ploughing; furthermore, the plough had certain disadvantages—it destroyed

contour walling, and was dependent on draught animals and on good rainfall to

soften the soil prior to its use.

As the number of migrants increased, women played increasingly active

roles in the productive economy by becoming involved in the structural aspects

of hut-building, the moulding of sun-dried bricks and the making of string, bas-

kets and mats formerly made mainly by men. The importance of women’s roles

in the economy as well as in social, political, and ritual organization minimized

the disruptive effects of migrancy and was a stabilizing factor in the material cul-

ture, both in terms of the technology and the demand for its products.

In the late 1930s there was virtually no labour migration of women and even

by the late 1970s, despite the increasing tendency for younger women to work as

daily and weekly labourers on neighbouring farms or in light industries, there

were still relatively few women migrants (Krige 1981: 155). In general the tech-

nical processes such as pottery and calabash-making, for which women were tra-

ditionally responsible, continued to be practised to supply domestic needs. In

the case of pottery, production exceeded subsistence requirements as many

skilled potters made pottery for sale as well as for their own use. This was one of

the ways in which women responded to the transition to a money economy. Far

from dying out, by the 1970s the practice of this traditional craft had increased

to meet new needs. It is noted, however, that although in the 1970s there was

widespread use of traditional technology in the domestic sphere, this was per-

petuated mainly by older women and by those living at the capital which is

notably conservative. Young educated women showed little interest in acquiring

or practising traditional skills and looked to employment outside the area.

Post-war economic change

During the decade after World War II the previously unhealthy Transvaal

Lowveld was opened up to intensive white settlement following the virtual eradi-

cation of malaria (achieved by the aerial spraying of insecticides). The conse-

quent development of the rich mineral and agricultural resources of the area and

the growth of industries ended a period of relative isolation for the local black

inhabitants. There was an unprecedented demand for labour in the newly estab-

lished factories, mines, and plantations. A network of new roads made possible

a new form of migrant labour, that of weekend commuting to and from the

LOBEDU MATERIAL CULTURE 191

reserves. Migrants returning at weekends with cash to spend brought much

needed money into the economy and also provided a welcome market for the

sale of home-brewed beer, which enabled village women to supplement the

wages of their husbands.

By the 1970s women had responded to the inevitable shift from subsistence

agriculture (resulting from shortage of land) by seeking ways of earning money

at home. Apart from brewing beer for sale, women were involved in other

money-making activities such as sewing, the moulding of mud bricks, the

plastering of huts and floors, and decorating of courtyard walls. By the 1980s

growing numbers of women had become wage-labourers on tea-plantations,

fruit-estates, and in light industry. These economic changes have had far-

reaching social implications (see Krige 1981: 153-155).

Mission influence

During the late nineteenth century Lobedu values and traditions were sub-

ject to the impact of mission doctrine. In 1881 Reuter, of the Berlin Missionary

Society, established a mission station at Medingen and started working among

Modjadji’s people. Mission teaching was opposed to polygyny, bridewealth pay-

ment, and cross-cousin marriage; it stressed free choice of marriage partners and

discouraged the extended family, thereby threatening the essential character of

Lobedu marriage and social structure (Krige 1981: 151). Although relatively few

people were converted, the mission had a strong educational influence and pro-

vided training in technical skills, especially in carpentry and building methods,

and encouraged individual enterprise and private ownership of property. By the

mid-1930s it was predominantly the small Christian community (less than 5 per

cent of the population) that had accepted western dress and was involved in

money-making activities to enable them to meet the costs of education and the

material requirements of their new way of life (Krige & Krige 1943: 320, 326).

Although early educational influences had some impact on values among a

small section of the community, belief in the power of the ancestral spirits was

not much undermined by the ideology of mission teaching. The institutionalized

rain and fertility cult with Modjadji at its centre continued to command wide-

spread observance of custom, reflecting the tenacity of ancestor beliefs and the

importance of rain in a subsistence economy primarily dependent on crops and

indigenous plant foods. Even converts to Christianity believed that Modjadji

had the power to make rain, albeit by the will of God (Krige & Krige 1954: 81).

Persistence of belief in the ancestors, in the power of medicines to control the

forces of evil, and in the fertility and rain cult was still evident in the 1970s but

the relevance of a fertility cult in an increasingly monetized economy was open

to question. 7

Pressure on environmental resources

Pressure on natural resources and on land in particular increased signifi-

cantly in the period 1930-80. This was primarily due to an unprecedented

192 ANNALS OF THE SOUTH AFRICAN MUSEUM

increase in population following the implementation of National Party government

policies regarding influx control of blacks in urban industrial centres and on

white-owned farms. The population density more than doubled between the late

1930s and the 1980 census when the average density had reached over 200 per-

sons per square kilometre. Of necessity much of the land formerly cultivated

was taken over for settlement. Such was the population increase that many

households had no fields at all. By the end of the 1970s the whole population

was dependent on wage labour and on buying their staple food, maize-meal.

The productivity of the remaining farm-land has been reduced by soil ero-

sion caused by over-grazing and the cutting back of protective bush and forest.

Plant resources basic to the traditional technology have been seriously depleted

and long distances must be travelled to obtain the materials required for special

purposes such as initiation costumes.

The increasing demand for fire-wood and the extensive use of indigenous

~ woods, including marula, for the firing of mud bricks has had a harmful effect on

timber resources. In the 1930s marula trees were never cut for fire-wood as the

nutritive content of the fruit and kernels was so highly valued (Krige 1983 pers.

comm.). In times of grain shortage the marula season in February and March

brought welcome relief to the hungry and bridged the lean months before the

reaping of the grain harvest. The present trend reflects both the breakdown of a

subsistence economy and the depletion of the environment to the extent that the

basic need for fire-wood outweighs all other considerations.

Investigation of the material culture in the 1970s showed a number of

responses to diminished natural and domestic resources, but the reaction to

changing economic conditions was more complex than a simple substitution of

new objects for old or the introduction of commodities previously unknown.

The 1970s inventory of material goods included many of the objects recorded in

the 1930s by Krige, but this did not mean that the pattern of their use had not

changed. In some examples, such as mortars for stamping maize, the frequency

of use had decreased considerably, in other cases substitute materials such as

sisal had replaced the indigenous fibres. Certain traditional objects had become

obsolete in response to social change, while others were used only on occasions

of ritual significance. Some of the technical skills were still practised, as in the

case of pottery, while others had disappeared completely. It became clear that a

number of variable factors came into play in accounting for the observed

changes and continuities. These included social, economic, technical, and ideo-

logical factors, the interaction of which resulted in the particular form of the ma-

terial culture at the time of observation. From a closer examination of specific

examples of change seen in relation to these factors, it is possible to gain some

insight into the nature of change in material culture.

CHANGE IN MATERIAL CULTURE c.1940—80

During the 1930s when the Krige Collection was assembled, the Lobedu

domestic economy was still largely self-sufficient and there was widespread

LOBEDU MATERIAL CULTURE 193

knowledge of the natural environmental resources utilized in the technology.

During the 40 years under discussion, for reasons outlined above, there was an

inevitable transition to a money economy. Some of the ways in which this funda-

mental economic change affected the pattern of use of particular utensils and

structures is discussed below.

Granaries and utensils associated with harvest

In a grain-based economy full granaries may be seen as indicators of pro-

ductivity and symbols of prosperity. In the 1970s, however, the disuse of grana-

ries reflected the converse. According to Krige (1982 pers. comm.) the 1930s

situation was strikingly different. Every homestead had kheasho or ledulu (see

p. 142) granaries for storing maize on the cob and many also had an under-

ground pit, kheledese (see p. 140), in the cattle-kraal or in the khdér6é for the

long-term storage of the decobbed maize. Large khesého grain baskets (see

p. 74) for more accessible grain storage were widely used and at harvest-time

temporary storage structures, moako and letole (see p. 143), for millet and sor-

ghum were set up in the fields prior to threshing (see Fig. 53). By the 1970s crop

yields could not satisfy basic grain requirements and there was no surplus to be

stored. The few granaries observed by the author were empty and in disrepair;

grain-pits had become obsolete.

Also prominent in the 1930s was the large mothatha basket (see p. 69)

used for carrying the harvest from the fields. As harvests declined, so did the

need for this basket, added to which was the fact that enamel basins proved use-

ful substitutes when carrying-utensils were needed. None the less, in the 1970s

mothajtha baskets were still occasionally seen in use in cooking-huts and court-

yards for holding maize-meal or other foods and, in areas where cultivation con-

tinued, they were still made prior to the harvest to supply a seasonal demand.

In the 1930s harvests were large enough to require transport by wooden

sledges. Fairly widely used after their early adoption from white farmers in the

Lowveld, these triangular wooden sledges were no longer in use in the 1970s.

Considering the soil erosion that they caused their use had been discouraged by

agricultural extension officers but, in addition, the diminished harvests rendered

them unnecessary.

Utensils associated with the hand-milling of maize

Ubiquitous in the 1930s, the wooden mortar, Jefudu, and pestle, mose (see

Fig. 52), were used two to three times a week throughout the year, the pre-dawn

thud of stamping being one of the characteristic village sounds (Krige 1983 pers.

comm.). By the 1970s the use of mortars and pestles for producing maize-meal

was limited to a few months after the harvest. Thereafter most people were

dependent on buying their staple food and the mortar and pestle were seldom

used.

With the reduced domestic stamping of maize, the varied dietary pattern

associated with different textures of meal produced at the successive stages in

194 ANNALS OF THE SOUTH AFRICAN MUSEUM

the process of stamping (see p. 78), was largely replaced by a monotonous and

less healthy diet based on highly refined, bought maize-meal.

Also related to the decrease in frequency of stamping maize was the decline

in the use of the khegdy6 mat (see p. 84) used for drying the fine meal, and the

wide-mouthed /ebéda pot (see p. 66) used for steeping maize during the stamp-

ing-process.

Interestingly, the /esél6 winnowing-basket (see p. 74), widespread in the

1930s for use during the stamping-process, remained in regular use not only for

winnowing when necessary but for a number of other domestic uses. It served as

a general-purpose kitchen utensil, a tray, a lid to cover the mouth of a pot, a

base for drying relish, or for holding and carrying garden produce. This multi-

purpose utility seems to have ensured a continued demand for the lesél6.

Utensils associated with cooking and the mosha

Although there was a widespread transition from home-grown to bought

maize, maize-meal still remained the basis of the diet and the manner in which it

was prepared and served remained relatively unchanged. Most of the traditional

domestic utensils associated with cooking were still much in evidence even in the

early 1980s. The khelalél6 cooking-pots of various sizes and morifhi bowls (see

p. 64) were used in almost every household for preparation of the daily meal, as

were the wooden spoons, maf6é (see p. 110), and stirrers, mafhéthd (see

p. 111), used at different stages during the cooking of vhoswa, the staple maize-

meal porridge. Most cooking still took place on an open hearth in the mosha or

inside the cooking-hut, the cooking-pot being supported by three suitably pos-

itioned stones or, in some cases, by a purchased triangular metal stand known as

a ‘drievoet’ (‘three-foot’). Having been prepared in the traditional manner, the

porridge was served into well-used and often carefully repaired calabash vessels,

zwithéba (see p. 153). The continued use of these utensils reflects the value

placed on them by village women. Although used side by side with enamelware

and plastic utensils, many of the locally made domestic utensils were preferred

for their particular qualities and usefulness. Thus the pot, modzéha (see p. 60),

used for storing drinking-water at home had not been replaced by tin or plastic

containers because the porous, cooling quality of the earthenware made it pref-

erable to alternative containers. For transporting water, however, lighter and

less breakable containers have advantages over pottery and consequently, since

the 1930s, tins and plastic buckets had entirely replaced pottery vessels for this

purpose.

In addition to particular material properties that give certain utensils practi-

cal advantages, characteristic posture while working is a pertinent factor when

considering the usefulness of particular objects or substitutes. Thus the hand-

broom lefielo (see p. 89), has not been replaced by the long-handled European

broom because it is more convenient to have at hand for sweeping while cooking

at the open hearth. Furthermore, as Krige has observed (1983 pers. comm.),

Lobedu women do not see bending down as a disadvantage because most

LOBEDU MATERIAL CULTURE 195

activities of the mosha are carried out at ground level in kneeling or bending

position.

Earning money at home: beer-brewing for sale

The transition from a subsistence to a money economy deprived village

women of their economic self-sufficiency. In most cases their husbands’ wages

were insufficient to meet the basic needs of food and clothing. A response to this

situation was that women became increasingly involved in money-making activi-

ties at home. Whereas in the 1930s it was mainly the Christian community that

made ‘a business’ of sewing or selling pottery, by the 1970s an informal sector of

the cash economy was developing rapidly. One of the most successful ways in

which women could earn money was to brew beer for sale.

A number of changes in material culture were related to the change from

brewing beer largely for social or hospitality purposes, as was the case in the

1930s, to the more recent brewing of beer for sale, mainly to supply weekend

commuters and visitors (access being facilitated by improved roads and trans-

port). The increase in scale of brewing demanded larger utensils—the large clay

nkhoé (see p. 60), traditionally used in brewing beer, was replaced by the even

larger “44 gallon’ drum, and the woven beer-strainer, leth6d6 (see p. 86), was

replaced by a capacious wooden-framed wire-gauze sieve, sef6 (see p. 114). As

neither the traditional beer-strainer, nor any commercially sold strainer was big

enough to cope with very large volumes of beer, enterprising craftsmen devel-

oped the sef6 (term taken from Afrikaans ‘sif’) to meet the needs of large-scale

brewing. It is an interesting example of a borrowed idea modified to suit specific

requirements.

By the 1970s, in addition to the time-honoured pattern of sharing beer

among kinsmen and neighbours on occasions of social or ritual importance, an-

other form of social beer-drinking had developed in response to the selling of

beer. In the older ‘hospitality’ beer parties the utensils used were the traditional

beer-pots (see p. 62) and calabashes. The long-handled calabash khehé kha

byalwa (see p. 153) in which beer was served was passed freely from person to

person, showing generosity and bonding social ties. The etiquette of the oc-

casion was dignified and ordered, the formality being expressed in the conven-

tional seating of men and women on either side of the server and in the use of

traditional utensils. By contrast, social occasions which centred around the sale

of beer were informal and traditional utensils were inappropriate. The need for

containers of standard volume in which the beer could be sold to individuals was

met by glass jars or tins of regular size. Instead of the beer circulating among a

group of friends, the buyer tended to consume the beer for which he had paid

and, as a customer, his behaviour was not subject to the etiquette appropriate to

a guest. Thus the difference between these two forms of beer-drinking, both of

which occurred in the 1970s (the conventional form being rare and usually as-

sociated with ritual occasions), was clearly reflected in the related utensils and in

the pattern of their use.

196 ANNALS OF THE SOUTH AFRICAN MUSEUM

The effect of migrant labour on the use of certain utensils

As noted previously, the migrant-labour system brought money and a wide

range of imported commercial goods into the domestic economy. It also had di-

rect and indirect implications for the manufacture and use of certain traditional

utensils.

A direct response to the decrease in the number of men in rural villages was

an increased participation of women in craft activities formerly undertaken

mainly by men. This trend, which was noted in the 1930s in relation to hut-

building and the manufacture of mats, certain baskets and cordage, continued in

the following decades. Where demand for a product continued, as in the case of

sleeping-mats, women became the main producers both for domestic use and for

informal sale.

One of the crafts, however, that remained exclusive to men was wood-

working. Women were unaccustomed to using carving-tools and the arduous

_ work of felling trees or cutting large branches, often deep in the bush, was more

appropriate to men. Furthermore, the rough carving of large objects often took

place in the bush to avoid having to transport the heavy timber. Although

migrancy reduced the potential number of craftsmen, there always had been

some degree of specialization and enough carvers remained to supply domestic

needs. Many of the carvers active in the 1970s were older men or those unable

to find employment in town, who could earn some money locally by making

wooden utensils for sale. Carpentry skills, which had been taught at the mission

since the late nineteenth century, were in demand for the construction of new

houses and other buildings within the reserve. Thus both traditional wood-

working technology and introduced carpentry techniques were utilized to supply

specific demands.

The direct effect of migrant labour on the customary pattern of marriage

(see Krige 1981: 148-157) had indirect effects on the material culture by chang-

ing the social context in which certain objects had been of practical and symbolic

significance. The decline in use of the kherodwana serving-basket (see p. 80) is a

good example of a utensil that became obsolete as an indirect result of migrant

labour. It was customary practice for a bride to take a kherodwana to her new

home as a symbol of willingness to serve and honour her husband. In this

basket, covered with a lid, she placed the daily meal that she prepared for her

husband. As it became increasingly common for marriages to be contracted

while the groom was working in town, the need for this special-purpose basket

declined. By the 1970s it was no longer in use at all and only a few older men

could recall the intricate way in which it had been made.

Material expressions of beliefs and values

In the context of ritual, objects and the materials from which they were

made were invested with symbolic meaning and their use was prescribed by cus-

tom. For example, the vuhwera and vyali costumes (see p. 93), made from

materials associated with rivers or damp places, were believed to enhance fertility

:

LOBEDU MATERIAL CULTURE 197

and prosperity. The choice of materials was based as much on a complex system

of values as on technical considerations. Similarly, river stones and certain

plants were believed to have ‘cooling’ properties and to embody positive, pro-

tective forces. The material dimension of ritual in the form of medicines, thugula

objects, shrines, certain living plants, or offerings of beer made from Eleusine

formed part of a much wider cognitive system that communicated meaning to

those who understood the symbolic associations.

In order that meaning could be understood clearly, material symbols had to

be conventional and not subject to idiosyncratic variation. The effectiveness of a

non-verbal language depends on the continuity of the symbolic associations. In

the Lobedu case there was evidence that knowledge of the symbols necessary to

communication through ritual action and the associated objects remained alive

even in the 1970s. This could be accounted for by the continued observance of

customs, by the tenacity of ritual practices and by the importance of women in

the sphere of ritual. In the long term, however, school education and the conse-

quent change in values cannot but undermine the relevance of a traditional

world view.

CONCLUSIONS REGARDING THE NATURE OF CHANGE

From the examples outlined above it is clear that an interplay of economic,

social and ideological factors accounted for change and continuity in the ma-

terial culture and that the nature of material change was more complex than

simple replacement of hand-crafted artefacts by machine-made substitutes. Dif-

ferent kinds of change related to particular circumstances and social contexts—

by the 1970s some artefacts, such as granaries, had disappeared completely in

response to economic change; others, such as calabash drinking-vessels for beer,

were retained for traditional social or ritual purposes, whereas modern substi-

tutes were used when beer was sold; the physical properties of certain utensils,

such as earthenware pots for storing water, made them preferable to substitutes,

thereby ensuring their continued use, and the multi-purpose utility of some

objects, as in the case of winnowing-baskets, made them less likely to become

obsolete than utensils of more restricted utility.

Early changes in dress related to mission and school influence as well as to

the requirements of town life. Social and religious distinctions were expressed

outwardly in differences in style of dress and, in some cases, in colour and type

of cloth. The expression of consciousness through clothing or ornamentation

continues to adapt to changing circumstances. In the 1980s a trend toward ‘neo-

traditional’ dress, arising from national consciousness, was apparent among edu-

cated women. This was particularly evident at the coronation of Modjadji V in

1982 (Krige 1983 pers. comm.).

Numerous examples confirmed that change in material culture could only

be understood with reference to the wider social and economic context. The

undermining of the metal-working industry that occurred before the period

under discussion was a direct result of colonial appropriation of the ore sources

198 ANNALS OF THE SOUTH AFRICAN MUSEUM

and smelting-sites, concurrent with the large-scale importation of cheap iron-

ware from Europe. In addition to the fact that the termination of local produc-

tion forced people to buy alternatives was the fact that many of the imported

tools, such as knives, axes or saws, had practical advantages both in terms of

availability and function. The saw, saha, for example, had no traditional equiva-

lent and met a real need; imported knives were very efficient and easier to

handle than the older mofhago. There was not, however, a total replacement of

old tools by new. Traditional adzes and gouges (see p. 101) were still the most

appropriate tools for carving certain wooden objects and the demand for them

remained. From the 1930s onward this demand was met by the forging of iron

(-thula tsipi) from scrap metal, often undertaken by Tsonga-speaking immigrants

from the Phalaborwa area (Krige 1983 pers. comm.). Thus, although in the

1970s there was an apparent continuity in the use of certain tools, the method of

their manufacture had changed considerably.

In the case of pottery the technology had remained remarkably constant

but, in response to changing economic conditions, the scale of manufacture had

increased far beyond subsistence needs. Conversely, the decline of skin-working

presents an example of diminished production linked to changes in clothing cus-

toms, decreased demand for articles made of skin, and consequent decline in the

related technology.

Basketry, in many respects, seemed to reflect a continuity of use and manu-

facture, but on investigation it turned out that women had become increasingly

involved in the making of certain baskets and mats in response to the reduction

of craftsmen caused by migrancy. An unqualified observation of continued use

of these items might have masked the significant social change that underlay the

increased participation of women in this form of production.

In view of the fundamental economic changes that have occurred over the

past 40 years it is perhaps surprising that so many of the traditional technical

skills have remained useful and relevant. Women in particular have proved both

conservative and adaptable in this changing situation by using traditional skills

to generate income and meet new economic needs. It is predictable that, as

increasing numbers of women enter the labour market, the previously conserva-

tive influence of women will be reduced.

In attempting to gain some understanding of the nature of change in

Lobedu material culture the limitations of a purely technological approach became

apparent and it was fortunate that a wealth of anthropological data, collected

and analysed by E. J. Krige, could be drawn on to inform the discussion of

change. The general conclusion that can be drawn from this study supports the

premise that change in material culture is inextricably bound up with social and

economic factors and with ideological values; the precise nature of the links and

the way they articulate depend on particular circumstances and are part of a

continuous process of change.

LOBEDU MATERIAL CULTURE 199

ACKNOWLEDGEMENTS

Part of the data published here formed part of a Master’s thesis in Anthro-

pology under the supervision of Professor G. M. K. Schitler, University of Stel-

lenbosch.

My warmest appreciation goes to Professor E. J. Krige for her assistance

throughout the study and for her constructive comments on the manuscript; to

Mr Simeon Modjadji and the people of Modyjadji’s village for their willing

co-operation; to Miss E. M. Shaw for her comments on the manuscript and

permission to consult her unpublished manuscript on basketry; to Dr N. J. van

Warmelo for advice on the vernacular terms; to the Compton Herbarium for

identification of botanical specimens; to Mrs June Hosford for photographic

work; to Mr V. Branco for the line drawings, and to Mrs Marcelle Scheiner for

typing the manuscript.

The Department of Co-operation and Development and the Lebowa Govy-

ernment are thanked for permission to work in the area. Financial assistance

from the Human Sciences Research Council is gratefully acknowledged. Opin-

ions expressed and conclusions drawn are those of the author and should not be

regarded as representative of the Human Sciences Research Council.

REFERENCES

Acocks, J. P. 1975. Veld types of South Africa. Memoirs of the botanical survey of South Africa

40: 1-128.

ALPERS, E. A. 1970. Dynasties of the Mutapa-Roswi complex. J. Afr. Hist. 11: 203-220.

BLUE Book oN Native Arrairs. 1910 (U.17-1911). Cape Town: Government Printer.

BEusTER, C. L. 1879. Das Volk Der Vawenda. Z. Ges. Erdk. Berl. 14: 236-40.

Curson, H. H., THomas, A. D. & Neirz, W. D. 1932. Native milking pails. Argeol. Navors.

Mus. Bloemfontein 1: 55-58.

Das Neves, D. F. 1879. A hunting trip to the Transvaal. London: George Bell.

Davison, C. & DESMOND CLARK, J. 1976. Transvaal heirloom beads and Rhodesian archaeolo-

gical sites. Afr. Stud. 35: 123-137.

Davison, P. & Hosrorpb, J. 1978. Lobedu pottery. Ann. S. Afr. Mus. 75: 291-319.

De.ius, P. 1980. Migrant labour and the Pedi, 1840-80. In: Marks, S. & Atmore, A., eds.

Economy and society in pre-industrial South Africa: 293-312. London: Longman.

DickeE, B. H. 1937. The bush speaks. Pietermaritzburg: Shuter and Shooter.

Du Torr, A. P. 1968. Die plek van die materiéle skeppinge van die Baphalaborwa in hul kultuur.

Unpublished Ph.D. thesis, University of South Africa.

EVANS-PRITCHARD, E. E. 1940. The Nuer. Oxford: Clarendon Press.

GRIMSEHL, H. W. 1955. Onluste in Modjadjiland, 1890-1894. Argiefjaarboek vir Suid-Afri-

kaanse Geskiedenis 11: 193-252.

HAMMOND-TOOKE, W. D. 1981. Boundaries and belief. Johannesburg: Witwatersrand University

Press. (Human Sciences Research Council Publication Series No. 74.)

HAtTTINGH, P. S. 1975. Population pressure on resources: the case of the Modjadji location. S.

Afr. J. Afri. Affairs. 2: 57-65.

HoLLoway Commission Report 1930. Unpublished MS, University of Cape Town libraries,

Herbst Papers (BC 79 E1 415).

Junop, H. A. 1927. The Life of a South African tribe. 2 vols. 2nd ed. London: Macmillan.

Kirsy, P. R. 1934. The musical instruments of the native races of South Africa. London: Oxford

University Press.

Krice, E. J. 1931. Agricultural ceremonies and practices of the bal.obedu. Bantu Stud. 5:

207-239.

200 ANNALS OF THE SOUTH AFRICAN MUSEUM

KRriGE, E. J. 1932. The social significance of beer among the baLobedu. Bantu Stud. 6: 343-357.

KRIGE, E. J. 1938. The place of the north-eastern Transvaal Sotho in the south Bantu complex.

Africa 11: 265-293.

KriGE, E. J. 1941. Economics of exchange in a primitive society. S. Afr. J. Econ. 9: 1-21.

KricE, E. J. 1964. Property, cross-cousin marriage and the family cycle among the Lobedu. In:

Gray, R. & GULLIVER, P., eds. The family estate in Africa: 155-192. London: Routledge &

Kegan Paul.

KricE, E. J. 1974. Woman-marriage, with special reference to the Lovedu—its significance for

the definition of marriage. Africa 44: 11-32.

KriGE, E. J. 1975a. Traditional and Christian Lovedu family structures. In: Wuisson, M. G. &

West, M., eds. Religion and social change in southern Africa: 129-152. Cape Town: David

Philip.

KricE, E. J. 1975b. Divine kingship, change and development. Jn: Fortes, M. & PATTERSON,

S., eds. Studies in African social anthropology: 55-74. London: Academic Press.

Krice, E. J. 1975c. Asymmetrical matrilateral cross-cousin marriage—the Lovedu case. Afr.

Stud. 34: 231-257.

KriGE, E. J. 1981. Lobedu marriage and social change. In: KricE, E. J. & Comarorr, J., eds.

African marriage in southern Africa: 148-157. Cape Town: Juta.

KricE, E. J. 1982. Table of objects of Lovedu material culture in the period 1930-40 and its im-

mediate past, from the point of view of process, change and development (with special refer-

ence to the Krige Collection in the South African Museum). Unpublished manuscript in the

South African Museum Library.

Krice, E. J. & Krice, J. D. 1943. The realm of a Rain-Queen. London: Oxford University

Press.

Krice, E. J. & Krice, J. D. 1954. The Lobedu of the Transvaal. In: ForbE, D., ed. African

worlds: 55-82. London: Oxford University Press.

KriGE, E. J. & Krice, J. D. 1956. The Tomlinson report and the Lovedu. Race Relations 23:

12-25.

KricE, J. D. 1934. Bride-wealth in baLobedu marriage ceremonies. Bantu Stud. 8: 135-151.

KricE, J. D. 1937. Traditional origins and tribal relationships of the Sotho of the Northern

Transvaal. Bantu Stud. 11: 321-356.

KRuGER, F. 1936. The Lobedu. Bantu Stud. 10: 89-105.

Kuper, A. 1975. The social structure of the Sotho-speaking peoples of southern Africa. Africa

45: 139-143.

Kuper, A. 1982. Wives for cattle. London: Routledge & Kegan Paul.

Lawton, A. 1967. Bantu pottery of southern Africa. Ann. S. Afr. Mus. 49: 1-440.

Leacu, E. R. 1961. Rethinking anthropology. The structural implications of matrilateral cross-

cousin marriage: 54-104. London: Athlone Press.

Monnic, H. O. 1961. Lobedu kinship terminology. Afr. Stud. 20: 226-236.

Monnic, H. O. 1963. The structure of Lobedu social and political organisation. Afr. Stud. 22:

49-64.

NELson, W. 1973. Some letters by William Nelson. Africana Notes and News. 20: 195-227.

PALMER, E. & PITMAN, N. 1972. Trees of southern Africa. Cape Town: Balkema.

REPORT OF THE TRANSVAAL NATIVE ADMINISTRATION. 1904-5. Pretoria: Government Printer.

REPORT OF THE ZOUTPANSBERG COMMISSIONER FOR NATIVE AFFAIRS. 1902-3. Pretoria: Govern-

ment Printer.

REUTER, F. 1905-6. Modjadje, a native queen in northern Transvaal: an ethnological study.

Rep. S. Afr. Ass. Adv. Sci.: 242-250.

SANSOM, B. 1974. Traditional rulers and their realms. In: HAMMOND-TOOKE, W. D., ed. The

Bantu-speaking peoples of southern Africa: 246-283. London: Routledge & Kegan Paul.

SCHLOSSER, K. 1975. Bantuktinstler in Stidafrika. Z. Ethnol.: 38-98.

SCHOFIELD, J. F. 1958. Beads: Southern African beads and their relation to the beads of In-

yanga. In: SUMMERS, R. Inyanga: 180-229. Cambridge: University Press.

ScuLLy, R. T. K. 1978. Phalaborwa oral tradition. Unpublished Ph.D. thesis, Binghamton Uni-

versity.

SpieR, R. F. G. 1973. Material culture and technology. New York: Houghton Mifflin.

StayT, H. A. 1931. The Bavenda. London: Oxford University Press.

TYRRELL, B. 1968. Tribal peoples of southern Africa. Cape Town: Books of Africa.

LOBEDU MATERIAL CULTURE 201

VAN DER WAAL, C. S. 1977. Die woning en woonwyse onder die Venda. Unpublished MA

thesis, University of Pretoria.

VAN DER MeErWE, N. J. & ScuLity, R. T. K. 1971. The Phalaborwa story: archaeological and

ethnographic investigation of a South African iron age group. World Archaeol. 3: 178-196.

VAN RteT Lowe, C. 1937. Beads of the water. Bantu Stud. 11: 368.

VAN WaRMELO, N. J. 1937. Tshivenda—English dictionary. Ethnol. Publs Un. S. Afr. 6: 5-345.

VAN WARMELO, N. J. 1940. The copper miners of Musina and the early history of the Zoutpans-

berg. Ethnol. Publs Un. S. Afr. 8: 4-209.

Wacne_r, R. 1980. Zoutpansberg: the dynamics of a hunting frontier, 1848-67. In: Marks, S. &

Atmore, A., eds. Economy and society in pre-industrial South Africa: 313-349. London:

Longman.

WIixson, M. 1969. The Sotho, Venda and Tsonga. Jn: Witson, M. & THompson, L., eds. The

Oxford history of South Africa: 131-182. Oxford: Clarendon Press.

‘ nial

fy ie

1 4 ~

Dy ae i

tor

,

®

ii}

x

5

4 Tae

lh

za

Nant

t

a

t

'

m é

oa

nu

{

ae

2 ;

( :

J

5 ite,

i 1

i ins

y

nf

i

iy

se

}

ime

a

4

6. SYSTEMATIC papers must conform to the /nternational code of zoological nomenclature

(particularly Articles 22 and 51).

Names of new taxa, combinations, synonyms, etc., when used for the first time, must be

followed by the appropriate Latin (not English) abbreviation, e.g. gen. nov., sp. nov., comb.

nov., syn. nov., etc.

An author’s name when cited must follow the name of the taxon without intervening

punctuation and not be abbreviated; if the year is added, a comma must separate author’s

name and year. The author’s name (and date, if cited) must be placed in parentheses if a

species or subspecies is transferred from its original genus. The name of a subsequent user of

a scientific namé must be separated from the scientific name by a colon.

Synonymy arrangement should be according to chronology of names, i.e. all published

scientific names by which the species previously has been designated are listed in chronological

order, with all references to that name following in chronological order, e.g.:

Family Nuculanidae

Nuculana (Lembulus) bicuspidata (Gould, 1845)

Figs 14-15A

Nucula (Leda) bicuspidata Gould, 1845: 37.

Leda plicifera A. Adams, 1856: 50.

Laeda bicuspidata Hanley, 1859: 118, pl. 228 (fig. 73). Sowerby, 1871: pl. 2 (fig. 8a—b).

Nucula largillierti Philippi, 1861: 87.

Leda bicuspidata: Nicklés, 1950: 163, fig. 301; 1955: 110. Barnard, 1964: 234, figs 8-9.

Note punctuation in the above example:

comma separates author’s name and year

semicolon separates more than one reference by the same author

full stop separates references by different authors

figures of plates are enclosed in parentheses to distinguish them from text-figures

dash, not comma, separates consecutive numbers

Synonymy arrangement according to chronology of bibliographic references, whereby

the year is placed in front of each entry, and the synonym repeated in full for each entry, is

not acceptable.

In describing new’species, one specimen must be designated as the holotype; other speci-

mens mentioned in the original description are to be designated paratypes; additional material

not regarded as paratypes should be listed separately. The complete data (registration number,

depository, description of specimen, locality, collector, date) of the holotype and paratypes

must be recorded, e.g.:

Holotype

SAM-—A13535 in the South African Museum, Cape Town. Adult female from mid-tide region, King’s Beach

Port Elizabeth (33°51’S 25°39’E), collected by A. Smith, 15 January 1973.

Note standard form of writing South African Museum registration numbers and date.

7. SPECIAL HOUSE RULES

Capital initial letters

(a) The Figures, Maps and Tables of the paper when referred to in the text

5]

e.g. “... the Figure depicting C. namacolus ...’; ‘. .. in C. namacolus (Fig. 10)...’

(b) The prefixes of prefixed surnames in all languages, when used in the text, if not preceded

by initials or full names

e.g. DuToit but A.L.du Toit; Von Huene but F. von Huene

(c) Scientific names, but not their vernacular derivatives

e.g. Therocephalia, but therocephalian

Punctuation should be loose, omitting all not strictly necessary

Reference to the author should be expressed in the third person

Roman numerals should be converted to arabic, except when forming part of the title of a

book or article, such as

‘Revision of the Crustacea. Part VIII. The Amphipoda.’

Specific name must not stand alone, but be preceded by the generic name or its abbreviation

to initial capital letter, provided the same generic name is used consecutively.

Name of new genus or species is not to be included in the title: it should be included in the

abstract, counter to Recommendation 23 of the Code, to meet the requirements of

Biological Abstracts.

PATRICIA DAVISON

LOBEDU MATERIAL CULTURE

A COMPARATIVE STUDY

OF THE 1930s AND THE 1970s

PART 4 SEPTEMBER 1984 ISSN 0303-2515

df

aa

te

ie

i

ib

"OF THE SOUTH AFRIC

CAPE TOWN

INSTRUCTIONS TO AUTHORS

1. MATERIAL should be original and not published elsewhere, in whole or in part.

2. LAYOUT should be as follows:

(a) Centred masthead to consist of

Title: informative but concise, without abbreviations and not including the names of new genera or species

Author’s(s’) name(s)

Address(es) of author(s) (institution where work was carried out)

Number of illustrations (figures, enumerated maps and tables, in this order)

(b) Abstract of not more than 200 words, intelligible to the reader without reference to the text

(c) Table of contents giving hierarchy of headings and subheadings

(d) Introduction

(e) Subject-matter of the paper, divided into sections to correspond with those given in table of contents

(f) Summary, if paper is lengthy

(g) Acknowledgements

(h) References

(i) Abbreviations, where these are numerous

3. MANUSCRIPT, to be submitted in triplicate, should be typewritten and neat, double spaced

with 2,5 cm margins all round. First lines of paragraphs should be indented. Tables and a list of

legends for illustrations should be typed separately, their positions indicated in the text. All

pages should be numbered consecutively.

Major headings of the paper are centred capitals; first subheadings are shouldered small

capitals; second subheadings are shouldered italics; third subheadings are indented, shouldered

italics. Further subdivisions should be avoided, as also enumeration (never roman numerals)

of headings and abbreviations.

Footnotes should be avoided unless they are short and essential.

Only generic and specific names should be underlined to indicate italics; all other marking

up should be left to editor and publisher.

4. ILLUSTRATIONS should be reducible to a size not exceeding 12 « 18 cm (19 cm including

legend); the reduction or enlargement required should be indicated; originals larger than

35 x 47 cm should not be submitted; photographs should be rectangular in shape and final

size. A metric scale should appear with all illustrations, otherwise magnification or reduction

should be given in the legend; if the latter, then the final reduction or enlargement should be

taken into consideration.

All illustrations, whether line drawings or photographs, should be termed figures (plates

are not printed; half-tones will appear in their proper place in the text) and numbered in a

single series. Items of composite figures should be designated by capital letters; lettering of

figures is not set in type and should be in lower-case letters.

The number of the figure should be lightly marked in pencil on the back of each illustration.

5. REFERENCES cited in text and synonymies should all be included in the list at the end of

the paper, using the Harvard System (ibid., idem, loc. cit., op. cit. are not acceptable):

(a) Author’s name and year of publication given in text, e.g.:

‘Smith (1969) describes .. .’

‘Smith (1969: 36, fig. 16) describes...’

‘As described (Smith 1969a, 19696; Jones 1971)’

‘As described (Haughton & Broom 1927)...’

‘As described (Haughton et al. 1927)...’

Note: no comma separating name and year

Dagination indicated by colon, not p.

names of joint authors connected by ampersand

et al. in text for more than two joint authors, but names of all authors given in list of references.

(b) Full references at the end of the paper, arranged alphabetically by names, chronologically

within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year, e.g. Smith (1969a, 19695) and not Smith (1969, 1969a).

For books give title in italics, edition, volume number, place of publication, publisher.

For journal article give title of article, title of journal in italics (abbreviated according to the World list o,

scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses, volume number, part

number (only if independently paged) in parentheses, pagination (first and last pages of article).

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FISCHER, P.—H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris 88: 100-140.

FISCHER, P.-H., DuvAL, M. & RAFFy, A. 1933. Etudes sur les échanges respiratoires des littorines. Archs

Zool. exp. gén. 74: 627-634.

Konn, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee region of Ceylon.

Ann. Mag. nat. Hist. (13) 2: 309-320.

Koun, A. J. 19606. Spawning behaviour, egg masses and larval development in Conus from the Indian Ocean.

Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L. Zoologische

und anthropologische Ergebnisse einer Forschungsreise im westlichen und zentralen Siid-Afrika 4: 269-270.

Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16: 269-270. ;

(continued inside back cover)

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 94 _ Band

September 1984 September

Part = 74 Deel

MORPHOLOGICAL AND BIOLOGICAL NOTES

ON SOME SOUTH AFRICAN ARTHROPODS

ASSOCIATED WITH

DECAYING ORGANIC MATTER

PART 3

THE FAMILIES DERMESTIDAE,

CANTHARIDAE, MELYRIDAE,

TENEBRIONIDAE, AND SCARABAEIDAE

(COLEOPTERA)

By

A. J. PRINS

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town 8000

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na gelang van die

beskikbaarheid van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad 8000

OUT OF PRINT/UIT DRUK

1, WOES, SO) GED 28 aa, SUES, 5, IED)

AG, ipa), CED), ©, a2, Wy, 1OGES).

111-2, 5. 7, tapi), 14(1=2), 15(425), 240), 27, 3103), 326), 33, 362) 5)

Copyright enquiries to the South African Museum

Kopieregnavrae aan die Suid-Afrikaanse Museum

ISBN 0 86813 061 3

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

MORPHOLOGICAL AND BIOLOGICAL NOTES ON

SOME SOUTH AFRICAN ARTHROPODS

ASSOCIATED WITH DECAYING ORGANIC MATTER

PART 3

THE FAMILIES DERMESTIDAE, CANTHARIDAE, MELYRIDAE,

TENEBRIONIDAE, AND SCARABAEIDAE (COLEOPTERA)

By

A. J. PRINS

South African Museum, Cape Town

(With 24 figures)

[MS accepted 2 April 1984]

ABSTRACT

The immature stages of twenty species of polyphagous beetles belonging to the series

Bostrychiformia, Elateriformia, Cucujiformia and Scarabaeiformia were collected along a nar-

row strip between Mossel Bay and Elands Bay and in small areas around Laingsburg and

Tulbagh. Their morphology is illustrated and their biology and association with decaying organic

matter are noted. Observations on some families of minor importance are also included. Most of

the species discussed are directly involved in the breakdown of organic matter.

CONTENTS

PAGE

Introduction and observations on minor families............ 204

Families Coccinellidae and Chrysomelidae............. 205

Families Elateridae and Anthicidae................... 206

Families Ptiliidae and Endomychidae ................. 206

amilyiClenidacr Wii ers Sook te ince ee ae ana ae 206

Families Colydiidae, Cucujidae and Ptinidae ........... 208

RamilyaNitidulidacie «25 Avi i. See 8 a Seinen eee 209

Families Cerambycidae and Curculionidae............. 209

Morphological descriptions and biological notes ............ DAZ

amilyMenmestidaer 2.4 1e steerer. oe eee tes INDE

Dermestesimaculats de Geena sens ne een DAS

DermestesiaremdeGect tas heme a orc ZA)

Dermestes peruvianus Castelnau.................. DOM

anmlyaCanth anda’ Atatt re eer re es ae 228

Avjronyeha picta (Wicdeniann))sa. ane sees eee 224

army NMC lw ae age es ce vee seu ayer tatnsy: taps acy 226

IMelyrisnvinidis i abniCiUS)}.4 324. eee oe cere D2

Familygkenebmomidacre, sy scs co. soon spoon ee JEM |

Gonocephalum arenarium (Fabricius) ............. 258

Pachyphaleria capensis (Castelnau) ............... 236

Tribolium destructor Uyttenboogaart ............. 239

Alphitobusidiaperinus (Ranzen)..5 2.445220" 20 en 243

203

Ann. S. Afr. Mus. 94 (4), 1984: 203-304, 24 figs.

204 ANNALS OF THE SOUTH AFRICAN MUSEUM

PAGE

Fanuly Scarabacidae ta seca. ce cee ee es 249

Apnodmsilacius Wiedemann ae eee juss)

Apnodiusimoesius Fabricius #225944. 6e sale ee 263

VAD NOGIUSIPLOCERUSIAatO dhe ae ran ere 265

Aphodiusiconsimilis Boneman — 4-4 4e.45 ee nee 269

Onis aygidus (Fabricius): 22... -e.4606 sae nee ae 272

Oniucellus picms(Hausmann))).9)..522 2.5505. 555- 274

Oniticellus planatus Castelnau.................... 278

Rachnodasinuata(Habticius)ia se eae aan 281

Rhinocoeta cornuta\(Fabricius)).....)..5...4. 25900 285

Temnorrhynchus retusus (Fabricius) .............. 290

roxy paroideslaroldin eee ne 296

ACKn owled eemiemtse es nea ricm rt Nn celts one eeu te rca 300

IRCICRENCESIS hs aati d casas hes surance ehh, ale, Sadar age a 300

AND DEC VIAUONS¥e 7 one, Sei erate Rete reer eee cat on ane 302

INTRODUCTION

AND OBSERVATIONS ON MINOR FAMILIES

During the surveys of arthropods associated with decaying organic matter

thirty-four species of beetles were collected in a narrow strip along the southern

and western coastal regions of South Africa and also a small strip between

Laingsburg and Beaufort West as well as between Worcester and Tulbagh.

Fourteen species which are mostly predacious were discussed in the second part

of this series; the immature stages of the remaining twenty polyphagous species

belonging to the series Bostrychiformia, Elateriformia, Cucujiformia, and Scara-

baeiformia are treated here.

The Scarabaeiformia include the coprophagous, phytophagous, myrme-

cophilous and termitophilous scarabs, the termitophilous valgines as well as the

myrmecophilous clambids, and two other small families, the helodids and dascil-

lids. The Elateriformia, Cucujiformia, and Bostrychiformia contain species of

diverse habits and all have members that are of economic importance, being

either entomophagous, phytophagous, xylophagous, or pests of stored products.

The curled-up scarabaeiform larvae of the cockchafers, dung-rollers, fruit-

beetles and monkey-beetles are common in most soils containing plant roots or

decaying organic matter. Predacious species, however, such as ladybirds and

leather-winged or soldier-beetles (family Cantharidae) have caraboid or cam-

podeiform larvae, whereas the eruciform types of the skin-and-hide beetles and

those of melyrids and clerids are often found in dried vegetable or animal matter.

The trogositids (family Trogositidae) (Fig. 13C), which are near the clerids as far

as larval characters are concerned, are often attracted to carcasses and cadavers

during the later post-mortem stages of decay, particularly after rains.

The adult beetles are very often gregarious or congregate in large numbers,

such as the orange and black-coloured ladybird, Lioadalia flavomaculata

(de Geer), of which vast numbers are occasionally found on fresh and semi-fresh

kelp strings in the supratidal and also in the intertidal zones during the early

summer. There is no explanation for this behaviour, but it perhaps coincides with

SOUTH AFRICAN ARTHROPODS 205

a shortage of its main food supply, aphids, at this time of the year, and they are

then forced to search for new resources. Masses of dead and dying ladybirds,

mainly the predacious L. flavomaculata, together with Adonia variegata (Goeze),

Cheilomenes lunata (Fabricius), and Scymnus moreletti Mulsant, and various

other beetles and stink-bugs are often found on the intertidal zone along the west

coast. These insects have been blown out to sea by strong winds and are then

washed on to the beaches.

Various beetles, some of which may be of economic importance, are

attracted to dry or decaying organic matter; they may play an important role in

the destruction of such material, or they may merely utilize it as shelter (e.g.

under large dry cow-pats), particularly in open veld with little vegetation.

In the cases mentioned below, no morphological data on the larvae could be

compiled. The adults of most of the species, however, were in some way associ-

ated with, or are related to, species attracted to organic matter.

Families Coccinellidae and Chrysomelidae

Apart from the above-mentioned ladybirds, another very small predacious

species, Cranophorus varius Weise, which, according to collection data, is

endemic to the Cape Province, was attracted to carcasses where a second and

third sarcophagous stage occurred after repeated rains during the late autumn and

early winter. It is about 1,9 mm long, shiny brown to piceous and covered with

short fine hairs.

The adults of the phytophagous ladybird Epilachna andulata Thunberg, and

of the red and black-striped leaf-eating beetle Chrysomela fasciata de Geer, are

sometimes found to shelter under semi-dry cow-pats. Adults and larvae of both

species feed on the leaves of plants of the daisy family (Compositae) (Skaife

1953). C. fasciata often causes extensive damage to foliage of Arctotheca populi-

folia on the sand-dune system along the south-western coastal belt during late

autumn and early winter. All the developmental stages of E. andulata may at

times be collected on the same plant.

Pupae of another phytophagous ladybird, Henosepilachna elaterii (Rossi),

were occasionally collected under semi-dry cow-pats along the coastal areas

during autumn, together with adults of E. andulata. Henosepilachna elaterii were

often associated with the black and orange-coloured H. gibba (Thunberg) in the

Saldanha— Vredenburg area, also under cow-pats. The pupae of H. elaterii are

yellowish with blackish spots, and the adult beetles, which emerged during March

and April, are 5,8—8 mm long, cadmium yellow, with large oval black spots.

The larvae of some beetles such as certain halticids or flea-beetles (family

Chrysomelidae) contain a toxalbumin, which may cause severe pain and even

death in warm-blooded animals. Death by paralysis is apparently caused by a

lethal saponin in the blood of the larvae of the bushman-poison beetles, Diamphi-

dia and Polyclada spp. (Shaw et al. 1963). These are all yellowish beetles with

larger or smaller blackish spots or marks on the elytra, and they vary from 10 to

15 mm in length. The larvae of the carabids of the genus Lebistina, particularly

206 ANNALS OF THE SOUTH AFRICAN MUSEUM

L. peringueyi Liebke, L. holubi Peringuey, and L. subcruciata Fairmaire, which

are ectoparasitic on the larvae of the above-mentioned poison-beetles, are even

more poisonous than their hosts (Koch 1958). These carabids, about 15 mm in

length, are also yellowish beetles, some with oblong blackish markings on their

elytra. A few of these parasites and their hosts have been observed by the author

under semi-dry animal droppings.

Families Elateridae and Anthicidae

Orthosomatic larvae are found in many families of which the adult beetles

are associated with decaying organic matter; only in a few instances, however,

have the larvae been observed in this medium. Certain click-beetles (Elateridae)

such as Lacon amplicollis Boheman, a brownish to blackish species about 6,6 mm

long, and some Cardiophorus spp. are often found in the veld under dry cow-pats

which do not harbour carabids. Larvae of these beetles, also known as wire-

‘worms, are elongate and cylindrical with a prognathous head, and may be

predacious, phytophagous, and even liminous, as is the case with certain can-

tharids. Ant-beetles (Anthicidae) are common under dry cow-pats almost

throughout the year, particularly the dull black Anthicus stygius la Ferté, and the

larger shiny black Formicomus caeruleus (Thunberg) (about 4,2 mm long). Only

one species, the shiny, dark-brown, wingless Anthicus apterus van Hille

(Fig. 8M), was observed on carcasses during the ceratophagous stage of decay,

but it was never seen on animal droppings in the veld.

Families Ptiliidae and Endomychidae

One of the most common beetles found in semi-fresh to almost dry dung is a

black unidentified species of ptiliid, about 0,8 mm long. These insects, easily

recognized by their feather-like wings, are some of the smallest of beetles known.

Their campodeiform larvae resemble those of the rove-beetles in shape as well as

in the pseudopodium-like tenth abdominal segment. They are active and are

often found together with the adults even on compost heaps and decaying kelp on

the beaches. According to Dybas (1976) the larval mouth-parts are not adapted

for a carnivorous mode of life and they apparently feed on the organic matter and

on fungus spores as in the case of Mycetaea ovulum Wollaston (Endomychidae).

The latter is a small, oval, shiny brown fungus-beetle (about 1,3 mm long) often

found in association with this feather-wing beetle, and which is, according to

collection data, endemic to the Cape.

Family Cleridae

The chequered or clerid beetles are elongate, sombre coloured to metallic

greenish or bluish or even brilliant red and yellow. Elytra are usually entire and

tarsal formula is 5—5—5. The larvae are orthosomatic with a prognathous head;

epicranial halves are separated by a gular area as in melyrids and the molar area

of mandibles is lacking; dorsolateral glands are apparently absent in members of

this family.

SOUTH AFRICAN ARTHROPODS 207

About 300 species have been recorded from the southern African subregion.

Most of them are predacious in both the adult and larval form and are therefore

beneficial; only three species, as far as is known, have become pests of stored

products with a high fat content. Of these, two cosmopolitan species, the red-

legged ham-beetle, Necrobia rufipes (de Geer), and the red-shouldered ham-

beetle, N. ruficollis (Fabricius), occur in South Africa. Both these beetles are

attracted to decaying carcasses of land and sea mammals and human corpses

during the casein fermentation period. They are both long-lived and survived in

the laboratory from March to January or February (a period of 10—12 months),

copulation being observed during October. Simmons & Ellington (1925) give the

longevity as 14 months. Larvae of both species were found to prey on the larvae

of the skin-and-hide beetle Dermestes maculatus de Geer, and those of the

cheese-skipper Piophila megastigmata McAlpine.

Necrobia rufipes is a metallic greenish-blue beetle (4,4—6,6 mm long) with

light reddish legs. The eggs are banana-shaped, yellowish white, and those

collected on a decaying turtle measured 1,10 by 0,24 mm. The incubation period

in the laboratory (25-26 °C) was 2-3 days and the life-span of five larvae occu-

pied 90-272 days on dried beef, the pupal period being 9-13 days (21-25 °C).

According to Simmons & Ellington (1925) the shortest developmental period is

30 days, which includes 17 days as growing larva and 13 days within the cocoon,

until emergence of adult (daily mean temperature 24,4 °C).

Necrobia ruficollis, on the other hand, is somewhat smaller (4,5-5,5 mm

long, blackish blue in colour with reddish pronotum and elytral shoulders, and

also with reddish legs. Eggs are similar to those of N. rufipes; those collected on a

seal carcass measured 0,88 by 0,28 mm. The incubation period in the laboratory

(21-25 °C) was 2-5 days, the larval life-span being 104—203 days under similar

conditions as those of N. rufipes. In both species a papery cocoon is produced by

the larva in which pupation occurs. The pupal period occupied 9-14 days.

Descriptions of the larva and pupa are given by Heeger (1848) and Scott (1919),

and a general description of a clerid larva by Peterson (1967).

Both the above species as well as the smaller Opetiopalpus collaris Schonherr

(3,6-4,4 mm) and Corynetinus fimetarius Wollaston (3,2-3,9 mm) have been

collected under semi-dry to semi-fresh cow-pats on the beach and further inland.

Both these latter species are blackish blue, the first with reddish and the second

with bronzy pronotum.

In Cape Town at least four species of clerids have been recorded in timber,

mostly Eucalyptus and kiaat, with the adult beetles feeding on the larvae of

powder-post beetles (Lyctus species). The largest of these is Cylidrus fasciatus

Castelnau (7,9—9,2 mm long), a somewhat shiny dark-brown beetle with a fairly

broad yellowish transverse band across the middle of its almost impunctate elytra.

It is widely spread in the Subsaharan and Malagasy regions (Corporaal 1950).

Paratillus carus (Newman), the white-banded clerid, which is of Australian

origin, has piceous elytra with a very narrow whitish-yellow transverse band, the

head and pronotum being reddish brown. It is slightly smaller than C. fasciatus

208 ANNALS OF THE SOUTH AFRICAN MUSEUM

(5,3-6,1 mm long). Its adults are often observed in the Cape Town docks in

Japanese oak, where it feeds on both adults and larvae of powder-post beetles. It

has also been introduced into England (Winkler 1960). According to Fisher

(1944) it is a common predator on Lyctus beetles and is also one of the most

numerous and active predators upon larvae of the smaller timber-borers in

Australia (Froggatt 1927).

Pallenis misella Boheman (4,9 mm long) is dark brown with a narrow yellow-

ish transverse band over the punctate elytra, but it differs from the other three

species in the pronotum, which is deeply and widely excised on each side of its

posterior part. It is widely distributed in South Africa, including records from

Natal.

The cosmopolitan small white-blotched clerid, Tarsostenus univittatus

(Rossi), is of similar coloration to P. misella, but has a narrower yellowish

transverse band on the punctate elytra and is smaller (3,2—4,6 mm long); it is

_ widely distributed in South Africa as indicated by collection records from the

Sandveld along the west coast and from Transvaal. According to Laing (1928) its

adults appear at almost the same time as the adults of its host. Froggatt (1927)

obtained this species from wood infested with larvae of Heterobostrychus aequalis

Waterhouse.

A metallic-blue clerid Cylidrus wallacei Thomson, almost 12 mm long, was

found in Bellville, Cape Province, during March 1983 in Koompassia malaccensis

logs imported from Malaysia, and which were infested with larvae and adults of

the yellow-legged auger-beetle, Xylothrips flavipes (Illiger) and the previously

mentioned shot-hole borer, Heterobostrychus aequalis. Both bostrychids are

widely distributed in the Far East (Froggatt 1927).

Trichodes aulicus Klug, another beautiful metallic-blue species about 8 mm

long, but with three orange patches on the elytra, was bred during October from

the nest of a small megachilid bee in the Calvinia district. According to collection

data it is widely spread in the north-western Cape Province; otherwise its habits

are unknown.

Families Colydiidae, Cucujidae and Ptinidae

The cylindrical bark-beetles (Colydiidae) and the flat-beetles (Cucujidae),

particularly Euxestus phalacroides Wollaston and Monotoma spinicollis Aubé,

which are common during the last or ceratophagous stage of decay of carcasses,

were observed to be very numerous in dry cow-dung along the west coast of the

Cape Province. They were found to feed on the dry particles and were often

associated with the shiny black indigenous spider-beetle, Pseudomezium coc-

quereli (Fairmaire), which is about 2 mm long. Both the first-mentioned colydiid

and cucujid are widely distributed in South Africa and overseas (Hetschko 1930a,

1930b) and are small (1,9—2,3 mm long). Euxestus phalacroides is oval and shiny

brown. Monotoma spinicollis is elongate and dull dark brown with serrate lateral

prothoracic margins.

Spider-beetles (Ptinidae) are common in the dry decay stage of carcasses and

SOUTH AFRICAN ARTHROPODS 209

cadavers, and at least four cosmopolitan species occur in South Africa, of which

Mezium americanum Castelnau and Gibbium psylloides (Czempinski) are almost

glabrous and shiny. The latter, also known as the storehouse-beetle, usually

occurs in ships carrying dried foodstuffs and is often reported as damaging maize

kernels under such conditions. This species, together with the drugstore-beetle,

Stegobium paniceum Linnaeus, and the tobacco-beetle, Lasioderma serricorne

Fabricius, has been found in the tomb of Tutankhamun (Hinton 1945). The other

two cosmopolitan ptinids, Tipnus unicolor (Piller & Mitterpacher), found in

walnuts, and Pseudoptinus lichenum (Marsham), introduced from Europe, are

hairy. Larvae of P. lichenum are known to bore in dry wood and bark of fig trees

(Hinton 1941).

Another small, brown, hairy indigenous spider-beetle, Stethomezium

squamosum Hinton, was found in the roots of the shrub Maerua racemulosa,

which are eaten by the local people during food shortage; this shrub occurs from

the eastern Cape Province to Swaziland. Hinton (1943) also records Ptinus tectus

Boieldieu together with other beetles in the roots of this plant, as well as Mezium

natalense Peringuey breeding in blood-meal in Zimbabwe. A rather small,

unidentified, shiny-brown Meziomorphum sp. covered with fairly long spinose

hairs and which is about 1,9 mm long, has been observed to feed on owl pellets in

the vicinity of Cape Town.

Family Nitidulidae

Various undetermined sap-feeding beetles often appear on decaying animal

matter, mostly on carcasses during the late dermatophagous and early cerato-

phagous stages of decay. Two well-known members of this family, the dried-fruit

beetle, Carpophilus hemipterus (Linnaeus) and the corn-sap beetle, C. dimidiatus

(Fabricius), are, according to collection data, widespread in South Africa. Carpo-

philus hemipterus is dark brown with large brownish-yellow marks on the elytra

and was common in soil where fish-meal was used as fertilizer, together with

C. humeralis (Fabricius), which is piceous brown with a small yellowish-red spot

on the base of each elytron; this latter species is also widespread in South Africa.

Carpophilus dimidiatus is a reddish-brown species, often found in dried figs.

These are all small beetles varying in length from 2,3 to 3,5 mm.

A larger (about 4,8 mm long) and more robust dark-brown species, Aethina

castanescens (Fairmaire), was common in chicken manure in fowl pens in the

Cape Peninsula, together with larvae of house-flies, false stable-flies and lesser

house-flies, and the millepede Ommatoiulus moreleti (Lucas), but it was absent in

manure which harboured large populations of the lesser meal-worm Alphitobius

diaperinus (Panzer), and the skin-and-hide beetle Dermestes maculatus de Geer.

Families Cerambycidae and Curculionidae

Apodous larvae are found among long-horned and snout-beetles. In the

long-horned beetles (Cerambycidae) the larvae are elongate and somewhat

210 ANNALS OF THE SOUTH AFRICAN MUSEUM

depressed. In the snout-beetles (Curculionidae), however, they are C-shaped

and usually have a hypognathous head. Larvae of long-horned beetles live in

wood and some are well-known pests. The adults, however, feed on flowers,

leaves, and bark or pollen of plants, and only one species belonging to the genus

Stenauxa was sometimes found among leaf litter and debris on the sand-dune

system of the coastal belt, together with the coppery oedemerid, Melananthis

senex Blair.

Adult snout-beetles, on the other hand, were common under cow-pats as

well as in leaf litter throughout the entire survey area, and were represented by at

least five subfamilies, of which the Eremninae predominated. At least three

well-known Eremnus species (Fig. 1D), viz. E. cerealis Marshall, E. setulosus

Boheman and E. atratus (Sparrman) (all endemic to the Cape Province according

to collection data) were found to be plentiful under dry pats, mainly in the

western Cape Province during the summer and autumn. Both E. cerealis

_ (speckled snout-beetle or grain-worm) and E. setulosus (grey snout-beetle) were

mostly observed under pats in fallow land and were collected only occasionally in

the veld. Eremnus atratus was also found at Hermanus during midwinter.

Eremnus pilosus Boheman, which resembles E. setulosus but is smaller, seems to

be restricted to the south-west of the Cape Province and was prevalent through-

out the spring. A small dark-brown Eremnus sp., about 4,5 mm long, was very

common near Saldanha Bay and Paternoster, but was never collected in any other

area.

The vegetable-weevil, Listroderes costirostris Schonherr, and Neocleonus

sannio (Herbst) (Fig. 1A) (both widely distributed in South Africa and overseas

(Csiki 1934, Voss 1973)), as well as Rhytirrhinus inaequalis (Fabricius) ((Fig. 1C),

which is endemic to the Cape Province, according to collection data), were mostly

found under dry cow-pats in the southern Karoo during the summer. However,

N. sannio was often found along the west coast under similar conditions, together

with Rhytirrhinus acerbus Boheman. During the late summer and early autumn

very large numbers of the vegetable-weevil, L. costirostris, may be present under

totally dry cow-pats in the sandveld area along the Cape west coast, particularly

during periods of drought. In certain areas up to three hundred specimens were

counted under a single large pat. Rhytirrhinus and Listroderes are more or less the

same size (7,5-7,8 mm long), but these two species of Rhytirrhinus are easily

distinguished by the strong carinae on the elytra. Neocleonus sannio is much

larger (12-13 mm), light grey in colour and marked with four to six brown

patches on the elytra.

A tiny, dull-black species of Ocladius, about 3,7 mm long and almost

globular in shape, with rugosoreticulate elytra and the pronotum with rows

of parallel, longitudinal, thin carinae connected with cross-ridges, was found

under dry pats in the Karoo during the late summer and autumn. At least one

species of Rhysoderes (Fig. 1B) was found to shelter under cow-pats near

Saldanha. It is 6,6—8,0 mm long, brownish grey, with head and pronotum rugoso-

punctate.

peli

SOUTH AFRICAN ARTHROPODS

lis.

inus inaequa

irrh

B. Rhysoderes sp. C. Rhyt

Neocleonus sannio.

A

Curculionidae.

Fig. 1.

D. Eremnus sp.

IAL ANNALS OF THE SOUTH AFRICAN MUSEUM

MORPHOLOGICAL DESCRIPTIONS AND BIOLOGICAL NOTES

The systematic arrangement followed is that of Britton (1970).

SERIES BOSTRYCHIFORMIA

Family Dermestidae

Small to medium-sized oval to somewhat elongate beetles, usually of sombre

coloration, often with coloured pattern due to the arrangement of scales on the

body. Head small and deflexed with clubbed antennae. Hind wings well de-

veloped, elytra covering the abdominal segments. Tarsal formula 5—5—5.

Of the nearly 500 species described, only about forty have been recorded

from southern Africa, of which at least ten species are of economic importance.

Hinton (1945) gives a good account (including illustrations) of the various species.

Dermestids are scavengers and feed on dried animal and vegetable matter with a

- high protein content, and some species are therefore attracted to decaying or-

ganic matter such as carcasses or cadavers, particularly during the formation of

butyric acid and other volatile fatty acids. Many species are found in nests of

insects and other higher animals, including birds and small mammals, often killing

the young, and a few are even predacious, feeding on the immature stages of

other insects such as the larvae of the wax-moth, Galleria mellonella Linnaeus, in

beehives.

Apart from the fact that the hairy larvae may cause dermatitis in sensitive

people, due to the liberation of histamine stimulated by urticating hairs (Smith

1973), there is also the danger of diseases such as anthrax (Bacillus anthracis

Koch) being spread by these beetles. The hairs of some species may remain in the

grain on which the larvae were feeding and may have serious effects when

swallowed (Morison 1925).

The larvae are easily recognized by their long body hairs, which may be

simple or modified, and members of the genus Dermestes have, in addition,

paired sclerotized urogomphi. They do extensive damage to stored foodstuffs

(grain and cereal products, dried fish, bacon, biltong, dried sausages, ham, etc.),

carpets, leather, feathers and hair, beeswax, clothes, skin and hides, bone,

insulating materials and museum specimens. Larvae and adults recovered from

Egyptian mummies had caused considerable damage to the embalmed bodies

(Strong 1981). At least eleven species of beetles were found in the skull of an

Egyptian mummy (Hope 1834), of which four were skin-and-hide beetles, viz.

Dermestes maculatus de Geer, D. peruvianus Castelnau, D. frischii Kugelann,

and D. roei Hope. Dermestid larvae will attack laminated boards or even

plaster-of-paris models if they contain glue of animal or vegetable origin, and

when ready for pupation they will perforate roof timbers, wooden panels, and

even lead sheeting. Under adverse conditions they may turn cannibalistic.

There are at least three vegetarian species of Trogoderma, of which the well-

known cosmopolitan Khapra beetle T. granarium Everts, is the most important.

It is a serious pest of stored grain elsewhere in the world, but is apparently not

SOUTH AFRICAN ARTHROPODS ZS

established in South Africa. However, it was reported from a brewery in Pieter-

maritzburg in malt imported from England as early as 1953, and from near

Pretoria in 1955 when a heavy infestation of this beetle and some other grain

insects occurred in a store in sweepings that had been allowed to accumulate for

about three years.

The carpet-beetles include at least six cosmopolitan or nearly cosmopolitan

species that have been recorded from South Africa, including the widely distrib-

uted varied carpet-beetle, Anthrenus verbasci (Linnaeus), also known as the

museum-beetle, which is so destructive to dried museum specimens, and the true

museum-beetle, A. museorum (Linnaeus), reported in the past only from South

West Africa. The furniture carpet-beetle, A. vorax Waterhouse, which is very

similar to A. verbasci, but is easily separated from the latter by the inner or

mesal margin of the eyes, which is excised or interrupted in the middle (entire

in A. verbasci), has been observed in nests of the social weaver-bird (Phile-

tairus socius (Latham)). The larvae feed on feathers and other debris. One of

the more common species found in carpets, bird carcasses and poultry manure in

the western Cape Province is the cosmopolitan black carpet-beetle, Attagenus

piceus (Olivier), which is more elongate than species of Anthrenus; it is occasion-

ally found in association with the indigenous Attagenus jucundus Peringuey

and the nearly cosmopolitan A. gloriosae (Fabricius) in carpets in the Cape

Peninsula.

Skin-and-hide beetles are common in decaying carcasses, both on the

beach and inland, in fish-meal, dried fruits, animal droppings, and even horn,

feathers, and biscuits. Four cosmopolitan species are recorded from South Africa,

of which Dermestes maculatus and D. peruvianus are the most numerous in the

western Cape. Dermestes ater de Geer, although widespread, is a more northern

species, whereas the larder-beetle, D. lardarius Linnaeus, is only found occasion-

ally in the Cape Town docks, mostly in dried foodstuffs such as macaroni and

walnuts.

Dermestes maculatus de Geer

DESCRIPTION

Adult

Previously described by Hinton (1945). Blackish, dorsally clothed with de-

cumbent fulvous to greyish hairs, with mostly head and a broad band along each

lateral margin of pronotum covered with longish white hairs. Apex of each

elytron serrate and sutural angle produced in form of a strong tooth, absent in the

other two species described here. Lateral impressed line of first sternite curved

inwards near base. Length of specimens examined varying from 5,5 to 10 mm.

A native of Europe and widely spread in most parts of the world (Hinton

1945). It is the most common skin-beetle around Cape Town and is also well

established on most of the islands along the South African south and west coasts,

including Dyer, Bird, Marcus, Jutten, Ichaboe, and Malgas islands.

214 ANNALS OF THE SOUTH AFRICAN MUSEUM

Larva (Fig. 2A)

General descriptions are given by Rees (1943), Hinton (1945), and Peterson

(1967). Body form of Dermestes maculatus larva elongate, semi-circular; tergites

sclerotized and piceous to dark brown with broad, pale yellowish-white median,

longitudinal line over dorsum, representing ecdysial suture. Head hypognathous.

Tergal plates covered with medium-long to very long spinulate setae, some hairs

on third to eighth abdominal segments somewhat longer than others. Spinulate

hairs on ventral side paler and shorter. Integument smooth, without spinules

except for short branched (spinulate) spines on certain areas. A transverse row of

strong pointed tubercles present on anterior margin of dorsal sclerotized plates of

abdominal segments 4—9 (Fig. SA—B). Spiracles well developed, mesothoracic

one the largest, oval (Fig. 2C); first three abdominal spiracles less oval and

situated on a non-sclerotized membrane below sclerotized tergal plates, those on

fourth to eighth segment smaller than other abdominal spiracles, almost circular

- and situated on the sclerotized tergal plates (Fig. 5A); in pale-coloured larvae

they are often less sclerotized than others (Fig. 2D). Legs pale brownish to

brown, tarsungulus (Fig. 3L) dark brown, and with short narrow keel-like projec-

tion ventrally on each side of which is a long fine seta. Setation of posterior face of

tibia as in Figure 3L; characteristic of D. maculatus is the long fine posterodorsal

preapical seta on tibiae. Length of full-grown larva about 15 mm.

Head (Fig. 2E—G)

About as wide as long, or only slightly wider than long, sides and hind margin

dorsally fairly convex. Frontal and epicranial (coronal) sutures distinct; frons

strongly sclerotized except for narrow strip along anterior border and on each side

in front of antennae; each side of frons with strong conical tubercle. Epicranium

also strongly sclerotized, except for small area around ocelli; latter six in number,

on each side arranged in two almost vertical parallel rows. Setation more or less

as illustrated. Antennae (Fig. 2B) three-segmented with whitish cone-shaped

basal articulating membrane; first segment slightly more than half the length of

second, wider than others; with about six setae situated dorsally some distance

from its apex; second segment with sensory appendix (sensory cone) at apex and

an oval sensory spot near middle. Third segment small, about one-third the length

of second, with sensory pegs at apex.

Labrum and clypeus

Labrum slightly more than half the length of clypeus, narrower than latter,

its anterior margin fairly deeply and angularly excised in middle. Clypeus

trapezoidal, much narrower in front than behind; anterior margin slightly con-

cave, preclypeus without setae, except for medium-long seta on each side near

lateral margin; postclypeus with about six to eight medium-long setae.

Epipharynx (Fig. 2H) with proximal sensory area composed of about six

large and ten small sensory papillae arranged in two almost parallel rows.

Epipharyngeal scleromes parallel, tapering and somewhat diverging posteriorly.

SOUTH AFRICAN ARTHROPODS 215

=

Y, |

Fig. 2. Dermestidae. Dermestes maculatus larva. A. Full-grown, left lateral view. B. Antenna.

C. Mesothoracic spiracle. D. Abdominal spiracles, segments 1-3, 4, 8. E. Head, dorsal view.

F. Head, ventral view. G. Head, left lateral view. H. Epipharynx. I. Left mandible, dorsal

view. J. Right mandible, dorsal view. K. Left maxilla, dorsal view. L. Labium, ventral view.

M. Labium, dorsal view. N. Labium, left lateral view. O. Left maxilla, ventral view.

216 ANNALS OF THE SOUTH AFRICAN MUSEUM

Broad median lobe setae about five in number on each side, lateral lobe setae

longer and finer and covering each lateral margin to a point well beyond its

middle. Also two sensory pegs situated in median V-shaped excision below two

dorsal median setae.

Mandibles (Fig. 2I-J)

Broad, almost triangular and slightly less than one-third the length of head.

Cutting edge tri-dentate, apical tooth (ti) somewhat larger than others. Molar

area absent, but represented by large stiff process of retinaculum and brush of

hairs or penicillus. Lateral face with demi-scrobe with about four setae of which

the distal two are longer than others; also transverse row of hairs near base of

mandible on lateral face, with median hairs the longest.

Maxillae (Fig. 2K, O)

Galea and lacinia united, only free at extreme apex, the boundary lines

clearly indicated on both sides. Lacinia with forked spur at apex and dorsally also

with brush of fine fulvous hairs, otherwise devoid of setae. Galea lobe-like with

brush of fine fulvous hairs apically and four to five short setae on external lateral

margin just below brush of hairs, as well as one somewhat longer seta on ventral

side near palpifer. Stipes broad, almost quadrate with some setae on ventral side

of which one is much longer than rest; dorsally without setae; also some

medium-long setae along its ridge-like exterolateral margin. A narrow juxtastipes

present, demarcated by sclerotized margin of stipes. Palpifer without setae.

Maxillary palp four-segmented, the segments somewhat tapering towards apex;

first and second segments of about equal length and somewhat shorter than third

and fourth in specimens examined; first segment with a single ventral seta, the

penultimate one with two setae, one lateral and one ventral; apical segment with

some sensory pegs at apex. Subcardo and precardo (alacardo) clearly indicated by

sclerotized demarcation line; former with only about three, latter with about

seven setae.

Labium (Fig. 2L—N)

Submentum and gula fused, mentum almost rectangular, longer than wide

with setal pattern as indicated in figure, including longitudinal row of setae on

lateral face (paramentum). Ligula cleft in middle, ventrally with about five setae

on each lobe as indicated; hypopharyngeal sclerome as a curved sclerite and

raised above the palpi in all specimens examined, superlinguae beset with minute

setae. Labial palpi two-segmented, tapering towards apex, segments of about

equal length, without setae except for sensory pegs on apical one.

Pupa (Fig. 3A-B, E-G, J)

Somewhat elongate, about two and a half times longer than wide when seen

from above, creamy white in colour; completely covered with fine medium-long

fulvous hairs, each hair growing from a small rounded tubercle, which is coloured

SOUTH AFRICAN ARTHROPODS 2h

¢

G 0 ae ree mo vy

Fig. 3. Dermestidae. A. Dermestes maculatus pupa, left lateral view. B. Dermestes maculatus

pupa, pronotum, dorsal view. C. Dermestes ater pupa, pronotum, dorsal view. D. Dermestes

peruvianus pupa, pronotum, dorsal view. E. Dermestes maculatus, antenna of pupa, left lateral

view. F. Dermestes maculatus, abdominal apex of pupa showing developing genital capsule,

ventral view. G. Dermestes maculatus, head of pupa, dorsal view. H. Dermestes peruvianus,

urogomphus of pupa, left lateral view. I. Dermestes ater, urogomphus of pupa, left lateral view.

J. Dermestes maculatus, urogomphus of pupa, left lateral view. K. Dermestes peruvianus, head

of pupa, dorsal view. L. Dermestes maculatus, tarsungulus and tibia of left front leg of larva,

posterior view. M. Dermestes ater, tarsungulus and tibia of left front leg of larva, posterior view.

N. Dermestes peruvianus, tarsungulus and tibia of left front leg of larva, posterior view.

218 ANNALS OF THE SOUTH AFRICAN MUSEUM

yellowish red; tubercles and their coloration more prominent on median dorsal

side of pupa, particularly in newly formed pupae; those on pronotal disc some-

what larger than others. The five so-called ‘gin-traps’ on third to seventh abdomi-

nal terga coloured reddish and very distinct, even in old pupae. Spiracles large,

almost circular on abdominal segments 1-6, that on seventh smaller and less

obvious. Two urogomphi vertical and almost parallel. Developing genital capsule

in form of two swellings as indicated in figure.

Characteristic of the pupae of all three species of skin-and-hide beetles

examined are the six to ten setae on the second ringlike segment of the antennal

sheath (Fig. 3E). These setae are also present in the pupae of Attagenus piceus, in

which case there are six ‘gin-traps’. In all the pupae of Anthrenus verbasci that

were examined these setae and the ‘gin-traps’ were absent.

BIOLOGY

In most species of dermestids the adult beetles need food and access to a

water supply for satisfactory reproduction. This was also found by Dick (1937),

who states that if beetles were allowed to drink they would continue to lay eggs

for 72 days. This is not the case with the smaller carpet-beetles such as Anthrenus

verbasci, which are able to complete their adult life and lay eggs without any

feeding.

The eggs of Dermestes maculatus found in decaying carcasses were dirty

white and minutely and longitudinally striate and also reticulate when viewed at a

certain angle. Most eggs examined varied from 1,30 by 0,54 to 2,04 by 0,92 mm

and were fairly dull.

In larvae reared in the laboratory on their natural food supply, the larval

life-span varied from 31—42 days during January to March, the pupal stage being

7-11 days. During April to May they needed about 50 days to mature and the

pupal stage was about 35 days. Pupation occurs in the larval skin in the soil

without the formation of an earthen cell.

The newly emerged beetle is pale yellowish white with the abdominal apex

reddish; pronotum pale reddish surrounded by a yellowish-white, broad border.

Legs pale, trochanters, tarsi and articulation points of femur, and tibiae reddish;

head also pale reddish, eyes brown, antenna reddish with the apical three seg-

ments piceous brown, and apices of mandibles black. After a few hours they

assume their normal dark coloration.

As indicated previously, during the act of pupation the larvae will perforate

any suitable material for this purpose and may cause severe damage to timber in

buildings. In the past various instances have been recorded where larvae of this

species caused damage amounting to thousands of rands where laminated boards

and wooden panels contained animal glue. It is also possible that the larvae may

consume some of the wood as part of their diet. This fact is also stressed by

Bedwell (1931) when he refers to the smooth hollowed-out parts of the timber

and the presence of the larvae in various stages of development.

During the surveys it was observed that the larvae of D. maculatus very often

SOUTH AFRICAN ARTHROPODS 219

feed on their own pupae and they in turn are heavily preyed upon by larvae of the

ham-beetles, particularly Necrobia rufipes, which sometimes occur together with

this species.

Dermestes ater de Geer

DESCRIPTION

Adult

Previously described by Hinton (1945). Dark castaneous to piceous, integu-

ment somewhat more shiny than in Dermestes maculatus; body hairs fulvous.

Lateral impressed line on first sternum widely curved inwards, only parallel to

lateral margin for about half its length (in D. maculatus it is parallel for most of its

length). Elytral apices without serrations and without sutural teeth. Length

varying from 8,3 to 9,5 mm.

Indigenous to America (Fauvel 1889) and widely spread in the world (Hinton

1945). Only occasionally observed in the Cape Town docks.

Larva

Very similar to that of D. maculatus (see Fig. 2) in size, body form and

setation including spiracles, except for pro-mesothoracic spiracle (Fig. 4B), which

in D. ater bears about twelve almost equally developed ventral setae, whereas in

all specimens of D. maculatus that were examined, there were only one strong

and about five weaker setae. In this species, as in D. maculatus, the first three

abdominal spiracles are situated on a non-sclerotized membrane below sclero-

tized tergal plates, but the abdominal segments are devoid of a transverse row of

strong tubercles (Fig. 5C-D). Urogomphi straight (Fig. 5D—in D. maculatus

slightly curved upwards and anteriorly), and tibiae (Fig. 3M) each with stout

dorsal spine on posterior face near apex (in D. maculatus it is replaced by long

slender seta); short spine-like ventral seta present on each side of base of tarsun-

gulus (in D. maculatus these setae are long). In most specimens examined the

median, longitudinal yellowish band over the dorsum in the form of a thin line,

forming a diamond-shaped patch in middle of each segment.

Head (Fig. 4A)

Very similar to that of D. maculatus in shape, sclerotization and setation, but

lacking tubercles on frontal area. Six ocelli on each side arranged in more or less

two groups of three each, as illustrated. Frontal and coronal sutures distinct.

Head of most specimens examined with somewhat more setae than in D. macula-

tus (see Fig. 2E—G), particularly on frontal area. Antennae also very similar to

those of latter species, including six dorsal setae on first segment; the sensory

appendix, however, seems to be somewhat smaller.

Labrum and clypeus

Very similar to those of D. maculatus, including emargination in middle

of anterior margin of labrum. Epipharynx (Fig. 4E) also similar and with double

220 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 4. Dermestidae. A-J. Dermestes ater larva. A. Head, left lateral view. B. Pro-

mesothoracic spiracle, left lateral view. C. Left mandible, dorsal view. D. Right mandible,

dorsal view. E. Epipharynx. F. Left maxilla, dorsal view. G. Labium, ventral view.

H. Labium, dorsal view. I. Left maxilla, ventral view. J. Labium, left lateral view.

K-S. Dermestes peruvianus larva. K. Head, left lateral view. L. Left mandible, dorsal view.

M. Right mandible, dorsal view. N. Left maxilla, dorsal view. O. Labium, ventral view.

P. Labium, dorsal view. Q. Left maxilla, ventral view. R. Epipharynx. S. Labium, left

lateral view.

SOUTH AFRICAN ARTHROPODS DIA

row of proximal sensory papillae, the front six large, arranged in two groups of

three each in most specimens seen; posterior row with about fourteen minute

papillae. Epipharyngeal scleromes also similar to those of D. maculatus (see

Fig. 2H); broad median lobe setae about five in number on each side; lateral lobe

setae spear-shaped and covering each lateral margin to a point just beyond its

middle.

Mandibles (Fig. 4C-D)

Very similar to those of D. maculatus, including demi-scrobe and transverse

row of hairs; however, retinaculum much narrower and penicilli much broader in

most specimens examined than in that species.

Maxillae (Fig. 4F, I)

Similar to those of D. maculatus (see Fig. 2K) and without any setae on

dorsal side, except for hair brushes on galea and lacinia. Setal pattern of ventral

side similar to that of latter species. Demarcation line between galea and lacinia

fairly distinct, the spur as in D. maculatus. Maxillary palp with third and fourth

segments of about equal length and both slightly longer than first and second;

apical segment narrowest; setal pattern, subcardo and precardo exactly as in

D. maculatus, except no setae could be traced on subcardo.

Labium (Fig. 4G—-H, J)

Very similar to that of D. maculatus, including shape, setal pattern, labial

palpi and hypopharynx.

Pupa

Also very similar to that of D. maculatus in colour, body form, size and shape

of developing genital capsule, but differing from it by the somewhat smaller hair

tubercles, which are more or less confined to median area of tergites, particularly

the pronotum (Fig. 3C), even in older pupae. In D. ater the urogomphi (Fig. 31)

are also more curved anteriorly and are not straight. In most specimens examined

there seem to be slightly more setae on the body, and the setae are somewhat

finer and more slender.

Dermestes peruvianus Castelnau

DESCRIPTION

Adult

Previously described by Hinton (1945). Dark castaneous brown, clothed with

short decumbent fulvous hairs. Integument somewhat more shiny than in Der-

mestes maculatus. Lateral impressed line on first sternite straight and parallel to

lateral margin. Elytral apices entire, not serrate and also lacking sutural teeth.

Length 6,6-11,0 mm.

D2? ANNALS OF THE SOUTH AFRICAN MUSEUM

Probably a native of South America and widely distributed through the

major countries of the world (Hinton 1945). During surveys it was not found on

the islands along the Cape coasts; however, it is almost as common as D. macula-

tus in carcasses and fish-meal accumulations around Cape Town.

Larva

Very similar to that of D. maculatus (see Fig. 2) in size, body form and

setation, including the form of the spiracles; D. peruvianus, however, has pro-

mesothoracic spiracle with more setae (about twelve) of equal size as in D. ater,

and only first two abdominal spiracles are situated on a non-sclerotized mem-

brane below sclerotized plates of the two segments (Fig. SE); also transverse row

of strong tubercles present on anterior border of sclerotized tergal plates on

peas i

Fig. 5. Dermestidae. A. Dermestes maculatus \arva, left lateral view of spiracles and tergal

plates on segments 3 and 4. B. Dermestes maculatus larva, left lateral view of segments 9 and 10.

C. Dermestes ater larva, left lateral view of spiracle and tergal plate on segment 4. D. Dermestes

ater larva, left lateral view of segments 9 and 10. E. Dermestes peruvianus larva, left lateral view

of spiracles and tergal plates of segments 2 and 3. F. Dermestes peruvianus larva, left lateral view

of segments 9 and 10.

abdominal segments 3 to 10 (Fig. SE-F). Urogomphi in this species (Fig. 5F)

fairly straight, their apices curved caudally and downwards and, when seen from

above, slightly divergent as in D. ater. Tarsungulus (Fig. 3N) with broad base, on

each side with long slender ventral seta, tibiae with long slender posterodorsal

preapical seta as in D. maculatus.

SOUTH AFRICAN ARTHROPODS DDG

Head (Fig. 4K)

Very similar to that of D. maculatus (see Fig. 2E—F) in form, sclerotization

and setation, including two tubercles on frontal area and distinct frontal and

coronal sutures. Six ocelli on each side appearing somewhat smaller and arranged

in two groups of three each. Antennae also similar to those of latter species,

including sensory appendix, but in most specimens examined oval sensory spot

placed on posteroventral side of second segment (in D. maculatus it is more

ventral in position).

Labrum and clypeus

Very similar to those of D. maculatus including setation and labral emargina-

tion. Epipharynx (Fig. 4R) as in D. maculatus (see Fig. 2H) with proximal

sensory area consisting of an anterior row of six large and a posterior row of about

twelve minute papillae. Median setae about three to four in number in most speci-

mens seen and rather broad, their apices either truncate or rounded, the setae

somewhat broader than in D. ater and rather similar to those of D. maculatus.

Mandibles (Fig. 4L—M)

Very similar to those of D. maculatus, including broad retinaculum, rather

narrow penicillus and transverse row of hairs near mandibular base; each demi-

scrobe also with about four shorter setae. Apical tooth (ti) the largest.

Maxillae (Fig. 4N, Q)

Similar to those of both D. maculatus and D. ater in form and setation,

including maxillary palpi, forked spur on lacinia and hair brushes on both lacinia

and galea. In this species there are also about three setae on subcardo.

Labium (Fig. 40-P, S)

Very similar to that of D. maculatus including setal pattern and hypo-

pharynx. Labial palpi also similar, but differ from those of both D. maculatus and

D. ater by the first segment which bears at least two ventral setae.

Pupa

Similar to D. maculatus but the hair-bearing tubercles, particularly on the

pronotum (Fig. 3D), although of about the same size, are more or less confined to

the median area in D. peruvianus, and in the few specimens examined the hind

margin of the mesonotum is more rounded in middle (in D. maculatus it is more

pointed or rather acute). Urogomphi (Fig. 3H) more curved anteriorly as in

D. ater and not straight as in D. maculatus.

SERIES ELATERIFORMIA

Family Cantharidae

Small to medium-sized, elongate, somewhat flattened leather-winged

beetles, usually with reddish and bluish coloration. Head generally narrower than