ANNALS

OF THE

SOUTH AFRICAN MUSEUM

VOLUME XXIV.

1. The Montagu Cave: A full Report of the Investigation of the Montagu

Rock-shelter.—By A. J. H. Goopwin, B.A. (Cantab.), Research

Assistant in Ethnology, University of Cape Town.

(With Plates I-VII and two Text-figures.)

Montacu is in the Western Province of the Cape, some seven miles

from Ashton railway station. The cave lies another six miles to the

eastward of Montagu on Derde Heuvel Farm.

This cave stands at a height of some 200 feet above the bottom of

a narrow precipitous-sided valley which runs from the Langebergen

Range in a general north-easterly direction to join the Kingua River.

The valley is cut through folded Table Mountain sandstone, the dip

of which at the cave is almost vertical. From the stream-bed to the

cave entrance the slope rises steeply and is covered with thick bush

growing on a talus of large and small blocks of sandstone ; but above

the cave there is a 200-foot broken wall of rock. The cave is on the

western side of the valley, and faces nearly due east. The shelter and

the inner cave (“ tunnel ’’) above it thus extend in a direction roughly

parallel with the strike of the beds.

The cave consists of two chambers. The outer one is of the rock-

shelter type, being wide and high at the mouth and descending

towards the rear of the cave, and it is with this shelter that the present

paper will deal. The second chamber is, at the present day, entered

through the first by means of an iron ladder cemented to the back

(west) of the outer cave. This tunnel does not seem to have been used

at all by the Earlier Stone Age peoples, signs of whose occupation are

VOLSEXEXVA PART 1): il

2 Annals of the South African Museum.

restricted to the outer cave, but a “ kibi” or “!kwe”’ stone, con-

temporary with the Later Stone Age occupation, and also a wooden

club, are believed to have been found here.*

It should here be noted that throughout this paper terms such as

Palaeolithic, Neolithic, Chellean, Acheulean, Mousterian, and even

Bushman have been avoided, and the two terms “ Harlier Stone Age ”

and “ Later Stone Age ” have been substituted, for the reason that the

rejected terms were coined to apply to certain definite Huropean

prehistoric cultures in certain definite Huropean lands. It is still

doubtful whether the Neolithic phases of culture have spread into

Southern Africa, and it is obviously incorrect to refer to the Later

Stone Age of South Africa as Neolithic when it seems probable that

it is an offshoot of a European Palaeolithic phase of culture.

Again, if it were proved that Mousterian man or some such

man had migrated to South Africa complete with his Mousterian

culture, we would still be unable to refer to his period as ‘‘ Mousterian

times.”

For these reasons, then, and as no human remains were found, it

has been deemed best to confine ourselves to the terms Earlier and

Later Stone Ages. It will be shown in the course of the paper that

the Earlier Stone Age can be divided into two sub-periods in this

particular instance. This, however, only suggests a line of investiga-

tion in future studies, and is not a generalisation, as so far this sequence

has only been noted in this one cave, and until it has been noted in

many more instances no generalisation is possible.

THE CAVE.

The cave at the mouth is 42 feet across, and from front to back

measures 47 feet. Itis “ D ” shaped, the upright of the D lying to the

front of the cave mouth, and the semicircle to the rear of the cave,

from whence the tunnel or upper cave is entered (see text-fig. 1).

At the mouth of the outer cave, and at a height of a foot above the

modern floor-level on either side, are two rock-paintings of the usual

South African type. Their height above the floor-level, and the fact

that they are not covered by any of the floor-level, points to their being

drawn when the floor was little lower than at present. The paintings

are all small and not of sufficient interest to merit their reproduction

in this paper. The interior of the outer cave is dark, and the painters

seem to have confined their efforts to the cave-mouth for this reason.

* Péringuey, “‘ Stone Ages of South Africa,” p. 33, pl. xix, fig. 152.

The Montagu Cave. 3

The light was, however, sufficient for photographs of the lighter portions

of the cave to be taken (see Plate I).

The floor of the cave has been littered and strewn with debris

(marked X in the section) to a depth of some feet in places. This was

due to several causes. In the first place, there is a normal deposition

from the roof of the cave, and from the bats and swallows inhabiting

the cave. Secondly, the upper cave was excavated by guano diggers,

who, in throwing out guano for conveyance by aerial railway, disturbed

the surface and also added some of their own debris. Thirdly, previous

excavators of the cave had also added to the debris from the Later

occupational layer (B in the sections) and the upper Earlier occupa-

tional layer (D in the sections). This accumulation of debris was a

difficulty which was only surmounted on the last excavation.

History oF THE WoRrRK.

The cave was first visited in 1918 by a botanist of the South

African Museum and a mountaineering friend. They stayed some

four or five days, but failed to make any discoveries.

About a year later (see ‘Cape Times,” 3lst May 1919) Mr. E. J.

Jansen made three visits to the cave. On all occasions he was accom-

panied by Mr. van Alphen (Resident Magistrate). On the second

visit Major Jardine of Cape Town accompanied him, and on the third

visit Mr. Russell (Inspector of Schools). To them the Museum is

indebted for the discovery.

Mr. Jansen saw among the debris what he describes as “ an imple-

ment of true palaeolithic form,” and thereupon dug through the

debris (X), which was here undisturbed, to a depth of some 3 feet.

On digging through this he came to “‘a number of implements of

Mousterian times” (sic). He soon encountered a red sandy soil E,

but did not dig through it. This excavation is called Jansen’s pit 1.

On the second and third visits he attempted to get through this layer

(EK), but failed at a depth of 3 or 4 feet. He found, however, “a

boucher of pre-Mousterian times” (s7c}. They now turned their

attention to the western side of the cave (Jansen’s pit 1) and dug

through to the red sand again. Above the sand they found some

rejects, some with the “ chisel-edge,” and some smaller artefacts.

Following upon the publication of these finds, the South African

Museum sent up Messrs. Haughton, Barnard, and Tucker, who

arrived, together with two boys kindly lent by the New Cape Central

Railway authorities, on Ist October 1919.

Meanwhile some correspondence had passed between Mr. Jansen

4 Annals of the South African Museum.

and the late Dr. Péringuey, then director of the South African

Museum, and a plaster cast was taken of one of the implements

found. Unluckily, of the two casts associated with the Montagu

implements in the Museum collection, neither is marked, and it is

impossible to say which is the specimen sent by Mr. Jansen.

CoNCLUSIONS.

Mr. Jansen’s party, it will be noticed, did not discover either the

Later occupational layer (B) through which they passed, or the middle

Karlier occupational layer (F), unless the implement “found in the

hard-bake ” (7.e. the ferruginous sandy layer E) can be regarded as

from this layer (F). Had this been so Mr. Jansen would most certainly

have dug further and discovered the layer F some few inches lower.

STRATIFICATION OF THE Cave Deposits.

The several layers have been lettered A-H, from the top down-

wards, with X to denote the debris, and S the scarpage towards the

Grotto (south) side. The layers will be dealt with in order of deposi-

tion (see Plates II and III).

Floor.

What is apparently the floor of the cave consists of a whitish

rubble of sand and angular fragments of stone of all sizes, and in a more

or less decomposing condition. The depth of the deposit varies, and in

places passes down into bed-rock. It was difficult to investigate this

layer to any great extent owing tc the fallen rocks which covered the

deposit towards the front of the cave.

H—This is the earliest and smallest occupational layer. It

appears to occupy a comparatively small area in the centre of the cave,

not extending to the back or sides, but resting directly upon the floor

deposit. How far it extended towards the mouth of the cave could

not be ascertained on account of the large rocks under which it dis-

appears ; it would appear to peter out, and not to extend far in this

direction.

The layer varies in depth up to a foot, and is composed of a series

of irregular black, grey, and white bands. Only three artefacts were

obtained from this layer.

G—This layer contains no implements, and consists of a deposit of

red slightly ferruginous sand, with pieces of decomposing sandstone

intermixed. It covers the H occupational layer entirely, overlapping

it at the back and sides, where it rests directly on the floor deposit.

It is fairly even in thickness, varying from 2 to 3 feet in depth.

The Montagu Cave. 5

A specimen of sand from this layer was sent to the Government

Chemical Laboratory for analysis. It consists “ mostly of quartz sand

(partly water-worn) containing organic matter, and is phosphatic.”

Of three such specimens, sent from the guano cave and layer HE,

this specimen contains the least phosphate, and that from the guano

cave most. There is only a small amount of iron present. The sand

of the deposit has been formed, in the opinion of the chemist, “ in the

disintegration of sandstone, and has subsequently been mixed with

bat guano.”

F—This is the second occupational layer, and covers the whole

of G layer. This, too, is composed of black, grey, and brown bands

quite irregularly mixed with white bands (Plate I, fig. 2). On the

north side is a large amount of yellowish loam. No visible fragments

of charcoal were found in the black bands in either this layer or H.

A considerable number of artefacts were found; in all some three

hundred specimens in varying stages of manufacture were brought

back to the Museum. The greatest accumulation was found towards

the Grotto, on the south-south-west side, thinning out towards the

front of the cave and the opposite side.

The excavators remarked an aggregation into two layers, or

horizontal planes, the upper one lying immediately below layer H,

and the lower one lying about a foot above the bottom of this layer

(F).

The thickness of this layer varies from 3 to 4 feet.

E—This is an upper deposit of red sand, and overlies F,

covering the whole area of the cave. It is markedly thicker towards

the rear of the cave beneath the guano cave opening, and thins out

to the front and both sides. At the eastern end of the Grotto is an

isolated lenticular patch.

A specimen from this layer was tound on analysis to consist of

quartzitic sand, partly water-worn, containing less iron than the speci-

men from G but more phosphate. There are traces of carbonate

of lime.

As in layer G the sand is homogeneous, except for small whitish

fragments of stone. It is of loose, gritty texture, but at the base

of both layers it shades to a dark brown or red, leaving a greasy

streak when attacked with the pick. One implement was found in

this layer, a few inches below the next layer D, from whence it had

apparently been forced.

The scarpage S interferes with this layer at the Grotto end.

The thickness of this deposit varies from 3 to 4 feet.

6 Annals of the South African Museum.

D—This is the latest of the Earlier Stone Age deposits, to which

general period all the previous layers belong. Like F, this layer

covers the entire cave, and like both F and H consists of irregular

black, grey, white, and brown bands, with no visible fragments of

charcoal (Plate I).

A considerable number of artefacts were found, and over three

hundred specimens were brought back to the Museum. Here again

most of the implements were found towards the Grotto. No imple-

ments were found in the top 8 or 9 inches towards the north side. In

the centre of the cave, owing to the debris above, little was ascertain-

able. This layer is again interrupted towards the Grotto by the

scarpage, and varies from 2 to 3 feet in thickness.

C—This layer overlies D, and where it was possible to differentiate

(1.e. towards the back and sides of the cave) it consisted of

ordinary brown earth, capped with swallow-guano, which is very thin

in places.

The total thickness of the layerisfrom 6 to 9 inches. It contains no

implements. The scarpage (S) seems to consist largely of earth from

this layer, which thickens considerably towards the north and south

sides of the cave.

B—This overlies C and apparently covers the whole cave,

though this could only be recognised on the last expedition, towards

the south, north-west, and north sides, where it varied from 6 inches

to 18 inches in depth.

The stratum is formed on a basal black band; over this les a

white band, and above this is ordinary brown earth, with black fire-

zones and occasional thin white bands.

An entirely different type of implement was found here, and the

layer belongs definitely to the Later Stone Age. These implements

are of a non-local chert. Charred fragments of grass and reed were

found in the lowest and succeeding black bands.

A—This is the modern surface deposit, overlying B. Towards

the cave centre it consists of a brown earth a foot thick, giving

place towards the sides to swallow-guano 3 or 4 inches thick,

and still accumulating. In the centre, beneath the debris, are remains

of dead grass.

The Debris.

X—This consists largely of angular fragments of small stones,

with occasional larger fragments (up to 15 to 20 lbs. in weight). The

presence of fragments of an “agglomeration of small crystals of gyp-

The Montagu Cave. a

sum, darkened with organic matter and containing phosphates, some of

these being water-soluble,” points to part of the debris being thrown

down from the guano cave above, where the same type of material

is found.

Since then the Later Stone Age deposit (B) and the upper

Earlier deposit have been shovelled up and incorporated by previous

excavators. This seems much more likely than that the implements

of both periods were thrown down from the upper cave, as flakes

found in the debris seem to have come from identical cores with

flakes in situ in the B layer.

CONCLUSIONS.

We may so far conclude that there are four definite occupations

in this cave. The three earlier examples (layers H, F, D), divided one

from the other by some 3 feet of quartzitic sand, are homogeneous

and of one main cultural group. How the interjacent layers were

deposited cannot be ascertained. Slow accumulation owing to the

decomposition of the sandstone roof would imply a considerable

period between the occupational layers. On the other hand, a sudden

rush of water from or through the upper cave, if this were possible,

might reduce the intervening periods to a matter of one rainy season.

This is improbable, as no sufficient ingress is discernible in the upper

cave.

However, a definitely long period has elapsed between the

occupational layers B and D, as may be seen from the accumula-

tion of mould and swallow-guano to a depth of 9 inches. A distinct

period has necessarily elapsed between B and the present time.

Falls of rock appear to have taken place during each occupational

layer, which would point, as does the decomposing state of the frag-

ments which have fallen, to a comparatively fast deposition of sand

from the cave roof.

AN ACCOUNT OF THE CoURSE OF EXCAVATION.

(See text-fig. 1.)

Dr. Haughton, having travelled to Montagu for a preliminary

survey and to obtain leave from the owner of the farm, commenced

work with Messrs. Barnard and Tucker on Ist October 1919. This

expedition returned on 16th October.

Jansen’s pits i and ii were found, and excavation was started at the

point ili, and extended towards and took in Jansen’s pit i. This is

8 Annals of the South African Museum.

marked as Trench 1, and gives a continuous section half-way across

the cave of layers A, B, C, and D, and a part of F. Finds 1-12 were

found.*

A trench was then dug towards the back of the cave (Trench 2),

Find 13 coming from layer B, and 14 from D. Here they found that

the ferruginous sandy layer rose to the back of the cave, and 15 was’

discovered, layer D. Excavation, a foot at a time, was now begun

at the south side of Trench 2. Finds 16-19 resulted from D, also a

buck or sheep tooth, towards what is termed throughout “* the Grotto,”

70 Guano Cave

Fic. 1.—Plan of the Cave.

an overhanging rock, jutting out some 5 feet. Finds 21-71 were

found again in D. The red sand reaches to the mouth of the Grotto,

where it is overlaid by an ash bed. Finds 73-109 were made in D

again, Find 72 being picked up by the late Dr. Péringuey from the

debris in the course of a brief visit he paid to the cave.

Trench 3 was then dug from Jansen’s pit 1 towards the mouth of

the cave, but was abandoned owing to the debris. Finds 110 and 113

are from here; the red sand is petering out, and giving place to white

sand and implements. It is therefore possible that these two imple-

* Each implement as it was found was marked with a number, and the data

entered up in the field note-book.

The Montagu Cave. 9

ments are from layer F. Finds 114 and 115 are from D. Find 116

is again probably from F.

A trial pit was then sunk through Jansen’s pit i. This was then

trenched to Jansen’s pit 1 to a depth of some 4 feet, and the real

nature of the cave became apparent for the first time. Finds 117-120

were made in layer F, and the trench had to be abandoned as the walls

were in danger of subsiding.

The next deposit, G, was dug to a depth of 3 feet, but F was not

reached, and this too had to be abandoned.

Pit iv was then dug, and found to be of similar composition to

Jansen’s pitii. Trench 5 was then dug towards the cave mouth, but

this too had to be abandoned after Finds 121-123 had been found in

layer D. This ended the first expedition.

This expedition found layers D and F, and attempted to dig

through G. They did not differentiate A, B, and C, as they were

working to a great extent in disturbed ground.

The second expedition consisted of Messrs. Barnard and Tucker.

They started work on 5th November 1919. Mr. Tucker left on 19th

November, and the late Mr. Lightfoot took his place and stayed on

with Mr. Barnard until 27th November 1919.

They commenced operations on the north side at Trench 1. On

this occasion they carried all excavated earth outside the cave, thus

making for cleaner work. Layer B was recognised, and several flakes,

etc., were taken, also a cylindrical stone showing signs of rubbing.

Find 124 was obtained from D, 125 from F, and 126 from B.

Layer E of sand thins towards the north side, and towards the

mouth ; layer D also thins out, both being pressed up by F.

They now dug 2 feet into G, but abandoned the pit temporarily,

and cleared away the debris on the south-east side. Find 134 was

discovered in the debris and 135 in layer B. Find 136 was also from

the debris. Number 137 had evidently been trodden into the sand

(E) from D.

At the back of the cave, B was found undisturbed again, and here

again no cowps-de-poing were found.

A pit 15 feet square was now dug through to layer G. In this a

trial pit was dug, and this disclosed the presence of layer H. Several

artefacts were found. This layer peters out towards the back of the

cave. A third pit was opened up a few feet away from the second,

leaving a wall 2 feet thick separating the two. The upper layer D is

here 2 feet thick, and apparently Find 143 is from here.

10 Annals of the South African Museum.

The red sand decreases, then increases towards the Grotto. From

F were obtained 144-146. The layer is some 3 feet thick. The

coloured bands decrease in number towards the back of the Grotto,

and creamy bands make up the layer.

DESCRIPTION OF THE IMPLEMENTS.

Earlier Stone Age.

In all the three lower occupational layers the cowp-de-poing is

present in large numbers; they are fairly evenly distributed, but in

layers D and F there is a larger accumulation towards and in the

Grotto than in the more open parts of the cave. In Hit was impossible

from the few artefacts found there to ascertain whether this was true

here or not.

The material used throughout is Table Mountain sandstone. Itisa

quartzitic sandstone, coarse in texture, and of little use in small flakes.

The cave is at the base of a 200-foot broken wall of this same Table

Mountain sandstone, while the approach consists of a rising talus of

sandstone boulders. The stream in the valley below is also full of the

same material, and it is from here that it is presumed that the material

was obtained. The artefacts may therefore be considered to have

been made from local stone.

While the material throughout all these three Earlier Stone Age

layers remains the same, the shape has a distinct tendency to vary.

Two shapes of coup-de-poing are present, a regular almond-shape

(Plate V) and a regular pear-shape (Plate IV). It would appear that -

an attempt was made to produce finished implements of the same

size in both types. The size is roughly from 6 to 8 inches in

length, about 1 to 14 inches in depth, and from 3 to 4 inches

across.

In both types the butt is worked to an edge, and is not smoothed in

any way; this might point to the possibility of a haft or pad being used

when in use, but nothing of the kind is known. Both implements end

in a point, the pear shape is more acute in angle than the almond, and

is somewhat similar to the European ficron type.

The great majority of these finds are not finished implements,

though it is difficult to differentiate between finished implements and

rejects. Some possible finished implements are, however, present, and

it is safe to judge from these what the maker was aiming at, and what

an accepted implement was like. These specimens it is perhaps safer

to term “ better specimens.”

The Montaqu Cave. 1]

The greater number are, however, merely “ roughed-out,” and

these, judging by the standard attained by the better specimens, must

be regarded either as unfinished rejects or as unfinished implements,

known as blanks. From a study of the implements the conclusion

I have reached is that the method of manufacture was not the same,

but varied with the size of the flake struck off from the original boulder.

If the fragment dealt with was of sufficient size the process seems to

have been to make an object similar in shape to the desired cowp-de-

powng but very much larger, the process being to reduce the size

all round until it reached the desired proportions, but retaining the

original shape.

Several of these large unwieldy blocks are present. A particularly

large specimen measures 13 inches long, 4 inches deep, and 7 inches

across. Two other specimens are roughly shaped by some six or seven

heavy flakes from each side. These are almost identical in size and

shape, some 10 inches long, 3 inches thick, and 8 inches across.

A certain number of better specimens were chosen from each layer.

Layer H yielded no better specimens, only some three or four artefacts

being found. From layer F came seventy-seven better specimens,

of which fourteen were sufficiently advanced to be accepted (if they

had been found alone) as finished implements. From layer D came

eighty-nine better specimens, and of these some twenty-eight were

acceptable as being well into the last stages. It is not likely, from

the fact that they were found together with so many obviously

unfinished artefacts, that these should be really finished, though

some may be.

These better rejects show a desire for symmetry and an attempt

to produce a delicate implement. The edge is straight in the better

specimens, showing no “ alternate flaking” on the edges, nor is there

any screw effect attempted, so far as can be seen.

There is in some instances a secondary working along the edge,

producing the step effect known as “ resolved flaking.”

The disposition of the two shapes is worthy of note. Layer H could

not be judged, owing to the paucity of the deposit, though of the two

couwps-de-poing found both appear to be unfinished specimens of the

pear type.

In layer F, of the seventy-seven better specimens, sixty-two are

pear-shaped and fifteen are almond-shaped.

In layer D, of the eighty-nine better specimens, seventy are almond-

shaped and fifteen are pear-shaped, the remaining four being almost

round, or ovate.

12 Annals of the South African Museum.

This shows a decided tendency in the earlier stage towards the pear

shape, and in the later towards the almond shape, in this particular

instance only. No similar evidence is known from other' sites.

Amongst the specimens are found others of a different shape

(Plate VI). These consist of pieces of stone, a little larger than the

coup-de-poing (though not always so), in some cases worked slightly

round the butt end, but in most cases unworked except for a few large

shaping-flakes which have been struck off. The real character of

these is uncertain, and therefore it is best to leave it for the present

unnamed. It is possible that they are merely a stage in the manu-

facture of the coup-de-poing. Specimens of this type have been

found elsewhere, and were regarded by the late Dr. Péringuey and

others as implements.* None show use, nor do any appear to have

had any secondary working.

There were forty-eight present in layer D, and twenty-five in

layer F.

A final type is of interest ; it may be termed for the present a

polyhedral. This consists (Plate VII) of a piece of stone, usually some

4 inches in diameter and 2 inches in depth. The shape is generally

that of a fairly regular polygon, the sides of which are formed by

striking flakes off along the edge alternately, the flakes usually leaving

a small platform about half an inch to an inch across in the centre

of each side. The edge that results is highly irregular, and the result

is an ugly ill-finished artefact, bounded by the irregular edge, which

of course follows the line whence the alternate flakes were struck.

These are distributed evenly among the layers; H yielded one,

F yielded thirty-two, and D sixty-eight. Twenty-two were found in

the debris X. These also are of Table Mountain sandstone. It is a

puzzle to discover what they were. Other similar specimens are

mentioned from Spain, and elsewhere in South Africa. They are not

coups-de-poing, and could not be shaped into them; they are not

cores for small implements, as none were found, nor would they be of

much use in this material ; nor are they Levallois flake cores, as none

have been found with the final flake removed, and such a flake struck

from a core of this type would be of little use.

In Europe they have been termed fabricators, but this also must be

rejected, as none of the specimens show signs of having been used, the

edges being quite clear-cut and unweathered.

On the other hand. hammer stones have been found. These are

round river pebbles, the rounded edges, points, etc., of which have been

* Péringuey, loc. cit., pp. 25, 26, pl. viii, fig. 57; pl. x, fig. 66, ete.

The Montagu Cave. 13

pitted by the action of detaching flakes. About a dozen are present

from each of the layers Dand F. They are about 4 inches in diameter,

and of similar stone to that used in making the artefacts.

The accumulation of two layers of sand, G and EH, each some 3 feet

deep, and an upper layer C, of some 18 inches thick, besides the

accumulation of organic matter in the implementiferous layers, point

definitely to a very long period of occupation by a people of a single

culture, the typical implement of whom was the cowp-de-poing. (See

conclusions, stratification of cave deposits.)

The cave, however, still remains unexplained. We do not appear

to be dealing with a workshop site, as there are no flakes present in

any of the layers.

We do not seem to be dealing with a cache, as some specimens are

obviously rejected for some defect in the stone. :

A dwelling site is improbable, owing to the presence of both

obvious rejects in large numbers and possibly finished implements.

There are, however, evidences of fire; but, on the other hand, no

animal bones were present, nor any other signs of habitation. The

organic matter found in the occupational layers is phosphatic, and

probably consists partly of bat guano. .

We can, however, conclude that we are here in the presence of a

long series of occupations by a people of cowp-de-poing culture. The

coup-de-poing found here, being of the same type as that found in the

Belgian Congo, Tunis, and Spain, may be regarded as having a

common origin with the Chellean or Acheulean of Western Europe.

Later Stone Age Deposit.

(Text-fig. 2.)

Layer B, the latest of the deposits, must be regarded as a workshop

site. The type of implement aimed at is similar to that found in large

quantities at Smithfield and Cradock, and also at the Cape Peninsula.

These may be regarded as of the Later South African Stone Age.

The typical implement of this industry is an end scraper, which is

described from its shape as a “duckbill.” It consists of a flake,

trimmed by two or three long flakes before being struck off the core.

This gives a neat fluted effect to the back. The edge opposite the bulb

of percussion is then worked to give a chisel-edge. The angle of this

edge may be anything up to a right angle (fig. 2, No. 2).

The debris and layer B are regarded as separate for the sake of

accuracy.

Debris.—Three duckbills are present ; they are some 2 inches long,

Drawings, natural size, of implements of the Later Stone Age.

Fic.

Nos. 2, 4, 7, 8, and 9 from layer B.

5, and 6 from the debris ;

Nos. 1

~ The Montagu Cave. 15

and vary from } inch to 1 inch in width. These (as are all the imple-

ments in both deposits) are of a chert. This is not strictly a local

stone, and has been brought here by the workers, but from what

distance is not yet ascertained.

Four points are also present, two broken and two whole. Three

of these are worked along the edge only, on the face opposite the bulb

of percussion ; the fourth appears to be unworked.

Two or possibly three light points are also present ; they are on thin

flakes, and have a worked point on either one end or on both. A small

double-concave scraper was found, and also two ill-formed crescents

of a rough Pygmy shape and size.

Layer B—Undisturbed in this layer are several Pygmy cores

(fig. 2, No, 9) all without signs of secondary working. A small Pygmy-

like crescent is also present. One duckbill which has been under the

influence of fire was found, as also a slightly concave scraper, and a

large circular scraper, some 2 inches across (fig. 2, No. 8).

There are some other objects present ; these are not of the same

material as the other. A rough cylinder of stone, some 34 inches long

and 14 inches in diameter, is of a hard plum-coloured sandstone.

It appears to have been used as a “ cold chisel’ for removing flakes.

Of the same material are a small nodule (? chisel) about 2 inches long

and 4 inch in diameter, and also a lower grindstone.

A block of softer light-red sandstone has been used for sharpening

arrows, awls, etc.. Two faces are deeply cut by the action of

sharpening.

CoNCLUSIONS.

Layer B and the implements thrown up by earlier excavators in

the debris would appear to represent the same culture. It is a typical

workshop site of the most usual South African phase of the Later

Stone Age.

It is considerably later than D, F, and H, and extends to fairly

modern times. It may be regarded as associable with the paintings

at the cave mouth.

I wish to thank the Director of the South African Museum for the

opportunity of examining the material excavated, the field notes, and

Dr. Barnard’s Report on the excavations ; and Dr. Haughton for his

help in the geological statements. The particulars here given of the

cave itself, the course of the excavations, the stratification of the

deposits, and the positions in which the artefacts were found, are all

derived from Dr. Barnard’s report.

Annals of the South African Museum.

PLATE EXPLANATION OF PLATES.

Fig. 1. View of interior of the Montagu Cave from the eastern corner.

2. Close-up of section within the dotted line.

. Section of cave from front (east) to back (west).

. Section of cave from north side to south side.

. Front and side views of a pear-shaped coup-de-poing. (The scale applies only

to the left-hand figure.)

. Front and side views of an almond-shaped coup-de-poing. (The scale applies

only to the left-hand figure.)

. Implement with rounded butt and chisel edge.

. Polyhedral type of implement.

Plate I.

below.

Tae

Close-up of this section

3Uano Cav

4 peas to

~ ~~-*

pss

‘) pues pry

J 4ak

"1°97 2U09¢ “Heo,

— pue

S$ EOTG

vedo pas SUIS |

Ute

ea) J

g ~) Yisegq

eae agy 2u0}0 a9

V llos a2ezInG

xX NIGCIC

S. Afr. Mus., Vol. XXIV. late I.

yi fi

Via

Photos, K. H. Barnard,

Plate IL.

Vol. XXIV,

Ann. S. Afr. Mus.,

“199M 0) 7Sey] wos

OAr>y fo U01}day

Se ~ ood ~

eA) — Eats 5

Plate III.

Mus., Vol. XXIV.

Ann. 8, Afr.

HLAOG OD ALYO/A Wor

GAR) fo uolja¢

aS \ —! | |

—— (LILA

Fo OOU muip I

Ne ‘ . > ~~ 6 = —— |

: S . “fo = — (SE ea a |

4 .. Oe ONOT Ciena | . > . ~ 5 . D »

’ ie A ° MM 6 ze » Some cynic we eon US es ° we canes Tern : |

ars J 0 » os \\ ne Eos ie ei cene sone) 6 see » 2 Gert 6 5

< ‘ rY eas . ~ aR ~t » ~A » © x

N SS ws ‘ ‘ Be Se ee eat) Re et\aouian ‘ za, eames

\ = 4G Os < SUS eo See amare? 5 2 Se Oe >> : ae

Seas SN ———————

————

Ls |

Plate Y.

Ann. 8. Afr. Mus., Vol. XXIV.

Neill & Co., Lid.

Afr. Mus., Vol. XXIV Plate VL.

( 17 )

2. A Comparison between the Capsian and South African Stone Cultures.

—By A. J. H. Goopwin, B.A. (Cantab.), Research Assistant in

Ethnology, University of Cape Town.

(With sixteen Text-figures.)

THIS paper is primarily a detailed comparison between the Capsian

industry of North Africa as represented by a collection presented to

the South African Museum by M. Paul Pallary of Algeria, and the

South African stone industries as represented by material from

various sources in the Museum collection.

The necessity for such a comparison is evident. Owing to the

geographical form of Africa, it would be impossible for any culture to

enter Africa overland, unless it passed over one of the great land-bridges

existing across the Mediterranean and perhaps across the present

Straits of Gibraltar. On the other hand, it would be impossible for a

large influx of culture from the EKur.-Asiatic mainland to have thus

entered without leaving distinct traces of that immigration in North

Africa.

It would appear necessary in the first place to give an epitome of

our knowledge of the archaeology of both the areas to be compared.

It is already an accepted fact that the Lower Palaeolithic, Chellean, and

Acheulean phases of European culture are present in Africa ; though

whether both phases or only one appear is difficult to decide.

Examples of South African parallels of this industry are given in the

preceding paper on the Montagu Cave.

The Aurignacian industry proper is confined to an area in Kurope

which includes central and southern France and the southern coast of

the Bay of Biscay ; but in Africa the Aurignacian is known as the

Capsian culture, from Gafsa in Tunis ; it had a separate development

of its own, lasting on through the Solutrean and Magdalenian stages

in Europe, and forming, with these latter, the Azillean and

Tardenoisian of France.

Here we come across a difficulty : M. Pallary appears from his book *

to speak of the Getulian as the Early Capsian period, and of the Ibero-

* Paul Pallary, Instructions pour les recherches préhistoriques, Alger., 1909.

VOL. XXIV PART l. 2

18 Annals of the South African Museum.

Mauretanian as the late Capsian period. On the other hand, Ober-

maier, in his book,* appears to have a different use for the terms.

He divides the Capsian thus :

(a) Harly Capsian, corresponding to European Aurignacian.

(6) Late Capsian, representing an evolution independent of the

Solutrean and Magdalenian of Europe, but synchronous with these.

It is subdivided thus :

1. Getulian of Eastern Algeria and South Tunis.

2. Ibero-Mauretanian of Western Algeria.

This places us in a difficult position, but in this paper it has been

thought best to confine the terms used to M. Pallary’s meaning, as

he himself marked the implements in the collection he sent.

While the Aurignacian is represented in two distinct cultures,

one in Africa and one in Europe, the Solutrean industry is more con-

fined, and extends only through Southern France and North-West

Spain, with various stations across Central Europe,f and must,

therefore, be disregarded in searching for comparisons with the South

African Stone industries.

In South Africa we are in the presence of two Stone Ages, but these

are not to be confounded with the Palaeolithic and Neolithic of Kurope,

and for this reason the terms Later Stone Age and Harlier Stone Age

have been chosen. The Earlier Stone Age consists of a cowp-de-poing

industry, which is probably an offshoot of the Chellean of Europe, but

not contemporary with it, and not necessarily disseminated by the

same people.

The Later Stone Age consists of three industries :

1. The Eastern culture, confined to Swaziland and the Eastern

portion of the Union, and reaching as far West as Britstown and

Victoria West. The typical implements are a point worked over both

faces, and a large crescentic implement, some 4 inches long and 14

inchesin depth. The “ axe-edge,” as the writer has termed the second

implement from its shape, is without parallel, so far as he has been able

to discover.

2. An end-scraper industry, spread all over eentral South Africa,

with exceptionally good centres at Smithfield and Cradock, and

elsewhere in the Free State and Cape.

3. A Pygmy or microlithic industry, occurring in the Free State

and in the Cape. It appears to be pure, except at the Cape Peninsula,

* Obermaier, Fossil Man in Spain, p. 114, Yale. 1924.

+ Jaeques de Morgan, L’humanité Préhistorique, L’évolution de Phumanité

Series, p. 300, 1921. [Translated, 1924, “‘ Prehistoric Man.”’]

A Comparison between Capsian and South African Stone Cultures. 19

where it seems to have acquired the so-called “ Solutrean ”’ worked

point from the Eastern culture.

It is with these two last cultures that this paper endeavours to

deal, and it has been thought best to deal with them by detailed

comparison.

The specimens presented by M. Pallary consist of Ibero-Mauretanian

artefacts from La Mouillah, near Marniah; and Getulian artefacts

from Ain Kermah, South Tunis.

GENERAL CONSIDERATIONS.

All the cultures with which we have to deal are flake-cultures. In

fracturing stones for implements a flake of stone is struck off, and this

bears at the point of concussion a slight conchoidal mound. This is

known as the bulb of percussion. On the block from which the flake

was detached is a corresponding hollow; this is known as the negative

bulb of percussion. The parent block is known as the core. The

Chellean industry is a core industry, and the core is shaped, by the

removal of flakes, to a convenient shape. On the other hand, the

industries under consideration are based upon flakes struck off the

core and trimmed for use. Obviously in the manufacture of such

flakes many are merely trimming flakes, which shape the block before

the final flake is struck off. Again, many of the flakes are useless

owing to an ill-directed blow, or a flaw in the stone ; these are termed

refuse, and are left at the site of manufacture, which is termed a

workshop site. Many acceptable flakes are spoilt in the trimming ;

these also are found on workshop sites, and are termed rejects or

wasters.

It is obvious that where a small flake is used a finer-grained stone

is necessary than when the large core is used. Hence in Kurope and

North Aitrica flint or chert are used ; and in South Africa, where there

is no true flint, chert or indurated shale are used.

Secondary working consists of a number of fine flakes struck or

pressed off the implement along the edge to form a saw-edge or a

comfortable back. Trimming, a pressure technique, consists of

regular flakings over the whole of one or both faces of the implement

to give it the desired shape. This last is completely absent in the

Capsian cultures, and in South Africa except at the Cape Peninsula,

where the Eastern culture appears to have had an influence.

The obverse face of a flake is that showing the bulb of percussion,

the reverse is that showing the exterior of the boulder, or the marks

of previous flakes having been removed.

20 Annals of the South African Museum.

A CoMPARISON BETWEEN THE GETULIAN INDUSTRY oF NoRTH

AFRICA AND THE SMITHFIELD INDUSTRY oF SouTH AFRICA.

The name “ Smithfield type ”’ has been taken, as it is from here that

the greatest number of specimens at present at this Museum have been

obtained.

The most typical implement is an end-scraper (duckbill). These

vary in size and shape from a small rectangular implement about an

a b

Fic. 1.—a, from Ain Kermah ; 6, from Smithfield. Nat. size.

inch long to circular scrapers two inches or so across. The worked

edge is on the obverse side of the flake at the opposite end from the

bulb of percussion. Some few freaks are found, as for instance the

specimen depicted in the preceding paper, in which the bulb of per-

cussion is at the side of the implement.

Fig. 1 shows a scraper from Ain Kermah of the Getulian industry,

compared with a similar implement from Smithfield (O.F.S.). The

front end of the implement is worked to an edge, and the sides of the

implement are also shaped for hafting or for easier handling. This

A Comparison between Capsian and South African Stone Cultures. 21

implement is typical of all South African sites of this culture. Fig. 2

gives a similar comparison from the same sites.

Fic. 2.—a, from Ain Kermah ; 6, from Smithfield. Nat. size.

Of the implements depicted in fig. 3, Nos. a and d are from Ain

Kermah, 6 from Bloemfontein, ¢ and f from Cradock, and e and g from

Fic. 3.—a, d, from Ain Kermah ; 6, from Bloemfontein; c, f, from Cradock ;

e, g. from Smithfield. One-half nat. size.

Smithfield. These are of the same type, but they tend to be of a

better and cleaner shape than the Getulian examples. The sides are

only shaped in a few instances.

22 Annals of the South African Museum.

Fie. 4.—a, from Ain Kermah ; 0}, c, from Cradock. Nat. size.

Fig. 4 compares implements from Ain Kermah (a) and Cradock

(6 and c). Nos. a and 0 have a slight concavity on the worked edge,

possibly accidental, though sometimes made purposely in the South

African examples.*

Fic. 5.—a.irom Ain Kermah; 6,from Cradock; c,from Smithfield. Nat. size.

* Note.—Figs. 3f and 4¢ are reproduced in Péringuey, *“‘ The Stone Ages of

South Africa,” Ann. §. Afr. Mus., vol. viii, 1911, p. 128, fig. 8, pl. xvi, and

p. 123, fig. 5, pl. xvi, respectively. Other examples are also given.

A Comparison between Capsian and South African Stone Cultures. 23

Fig. 5 gives the same type of implement, but tending towards a

point at the working end. Fig. b from Cradock is an obvious reject,

but is worked at the point of the flake to form an edge.

Fic. 6.—a, from Ain Kermah ; 6, from Cradock; c, from Douglas. Nat. size.

Fig. 6 shows a similar end-scraper, the sides of which are uneven,

but worked the whole way down. None of the irregularities can be

regarded as a concave scraper.

Fic. 7.—a, d, f, from Ain Kermah ; 6, ¢, from Smithfield ; e, from Hanover

(Cape) ; g, from Beaufort West. One-half nat. size.

Fig. 7 (a, 6, and c) shows the same type of end-scraper, but in these

cases they are not shaped to any extent. In the same fig. Nos. d,

24 Annals of the South African Museum.

e, f, and g are thick side-scrapers, tending to be triangular in section.

This implement is also on a flake, and is comparable with those shown

in fig. 5 above. The flake is formed on the core before removal by

the detachment of some two or three previous flakes. Nos. d and f

are from Ain Kermah, while e is from Hanover (Cape) and g from

Beaufort West.

In fig. 8 we come to what is generally termed a point, but is more

likely to have been used as a pointed scraper. This type is not at all

Fic. 8.—a, from Ain Kermah ; 6, from the Cape Peninsula ; c, from East

London. Nat. size.

common in South Africa; several large pointed scrapers of this

type come from Smithfield, but the two specimens chosen are from the

Cape Peninsula and East London respectively (b and c) while a is

from Ain Kermah. These are flakes worked on one face only along

both edges, on the obverse side.

Fig. 9 depicts a peculiar scraper or perhaps a sort of borer. No. b

is from Ain Kermah. It is on a flake and is worked precipitously on

both edges on the obverse side, and is described by M. Pallary as a

“ Grattior.” Specimens a and ¢ are a little different. They are from

Smithfield ; the first is apparently a borer used in the manufacture of

the “‘ Bushman” stone rings. It is also on a flake. No. c, on the

other hand, is not made on a flake, but appears to consist of a long

A Comparison between Capsian and South African Stone Cultures. 25

core-implement, quadrilateral in section, and worked on two diagon-

ally opposite edges. Many similar implements come from Smithfield

and other sites, but no others have the butt which appears in the Ain

Kermah specimen.

Fic. 9.—a, c, from Smithfield ; 6, from Ain Kermah. One-half nat. size.

It is worth noting here that in similar precipitously worked imple-

ments from South Africa the technique is the same as in the Getulian

industry.

It is much more difficult to discover the “ Burin” blow in South

Africa. It seems to appear in the Pygmy industry (see fig. 14 below),

and is shown in Obermaier (op. cit., p. 204).

Fig. 10 gives examples. No. ais from Ain Kermah, while 6, c, and d

are from Zwartkops (South-West Africa), Pomona (South-West Africa),

and Grantberg (South-West Africa) respectively. Neville Jones,

in his article in the Journ. Roy. Anthr. Soc., vol. liv, December 1924,

states that he has found the “‘ Burin” blow (in this same industry)

at Sawmills, Rhodesia. But on looking up implements from a similar

site close by (Sipopoma), the writer fails to find it. It was then found

necessary to look further afield, and the doubtful specimens shown

were discovered. The industry of South-West Africa seems to be a

very deteriorated form of the Smithfield industry, if indeed it may

be regarded as belonging to this industry at all. Inc and d it will be

seen that the burin blow is struck off from the same end as the bulb

of percussion, thus casting a doubt on its validity, though even this

is possible, as one of the examples from Ain Kermah shows the same

peculiarity.

26 Annals of the South African Museum.

Example d has been retouched from the more usual end, but here

again the specimen is too wind-worn to make it possible to identify

Fie. 10.—a, from Ain Kermah ; 6, from Zwartkops ; c, from Pomona ;

d, from Grantberg. Nat. size.

the working across the top edge. It would be perhaps safer to reject

the occurrence of the burin blow in South Africa, at least for the

present.

A COMPARISON BETWEEN THE [BERO-MAURETANIAN BRANCH OF THE ~

CapstAN INDUSTRY AND THE MicRoOLITHIC oR Pyamy INDUSTRY

OF SoutH AFRICA.

It has so far been difficult to compare the South and North African

Industries owing to lack of data. But the discovery of M. Pallary’s

interesting presentation opens up the field considerably, as little

literature on the subject has hitherto been available in South Africa.

Fig. 11 shows small “ planing tools ” from North and South Africa.

Nos. a and 6 are from the Cape Peninsula, and Nos. ¢ and d from

La Mouillah. Both the Ibero-Mauretanian and the South African

examples consist of a flake, struck off the core, and worked to a

roughly circular shape. A small platiorm is usually left in the centre

en the obverse side. The implement is very similar to the Azilian

tools depicted in Obermaier (loc. cit., p. 346). A similar implement

A Comparison between Capsian and South African Stone Cultures. 27

is depicted in Dr. Péringuey’s paper (8. A. Mus. Annals, Vol. VIII),

from Queenstown.

Fic. 11.—a, 6, from the Cape Peninsula ; c, d, from La Mouillah. Nat. size.

Fig. 12 shows a number of “ crescents.”” Nos. a and } are from the

Cape Peninsula, No. c is from Smithfield Poort (O.F.S.), while Nos.

d, e, and f are Ibero-Mauretanian from La Mouillah. Here again the

Fic. 12.—a, b, 7, k, m, n, from Cape; c, from Smithfield Poort; d,e, f, g, h, 2; i,

from La Mouillah ; 0, from Caledon ; p, ostrich egg-shell beads from Caledon.

Nat. size.

28 Annals of the South African Museum.

similarity is extraordinary, the Cape specimens tending to be a little

rounder than the North African. Nos. g, h, 7, and | are from La

Mouillah, and are not sc symmetrical, tending more to the long borer

type of implement. Nos. 7, k, m, and » are parallels from the Cape

Peninsula, and o is from Caledon. These are more definitely of the

borer type, and the egg-shell beads in p are made with this type of

implement. The series is from Caledon. These are apparently

Pygmy parallels to the Chatelperran knife and La Gravette point of

the Aurignacian of Europe.

Fig. 13 shows examples of microliths from the Cape Peninsula

(Nos. ¢, d, and f) and from La Mouillah (Nos. a, 6, ande). The technique

Saran

<2 Z

d 2

Fira. 13.—a, 6, e, from La Mouillah ; c, d, f, from Cape Peninsula. Nat. size.

is the same throughout, consisting of small flakes never more than

4 inch thick, worked to form a smooth edge or back along the deepest -

edge, again similar to the Aurignacian of Europe.

Fic. 14.—Pygmy implements, and ostrich egg-shell beads from the Cape

Peninsula. Nat. size.

A Comparison between Capsian and South African Stone Cultures. 29

Fig. 14 is a series of Pygmy cherts from the Cape Peninsula. The

shape is generally triangular, and the implement appears to be the

borer used for perforating egg-shell beads. The accompanying beads

are from the Cape Peninsula as well, and are of ostrich egg-shell.

The delicacy of the maker is evident both from the extreme daintiness

of the borer and the careful manufacture of the beads; these too are

similar to the Gravette point.

Fig. 15 is a comparison between a peculiar scraper formed on a

flake, the one edge being sometimes worked with resolved flaking.

a b

Fic. 15.—a, from La Mouillah ; 6, from Bloembosch ; ¢, from the Cape

Peninsula. Nat. size.

Whether it was used as it stands or not is a difficult question, but the

extreme similarity between the La Mouillah specimen (a), the Bloem-

bosch (Darling District) specimen (0), and the Cape specimen (¢), is well

worth noting.

- d

SAAS

Fic. :6.—Implements from Britstown. Nat. size.

30 Annals of the South African Museum.

Mr. C. van Riet Lowe has in his possession a number of “ notched

flakes’ similar to European Aurignacian specimens, definitely asso-

ciated with the Smithfield industry, from various sites in the Orange

Free State.

Fig. 16 brings us to an interesting collection of arrow-heads from

Britstown. They are described as being found “in or near round

stone enclosures,’ but nothing more is known of them. a, 6, and ¢

are borers, while d, e, f are tanged arrow-heads.

In the South African Industry a small end-scraper (known as a

“ duck-bill ”’ from its shape) is also present. It is of the same type,

but smaller than the specimens in the preceding paper found at

Montagu. It appears to be a neater form of the Getulian end-scraper.

How far this end-scraper has been found with the Ibero-Mauretanian

industry the writer has been unable to discover; it is, however, present

in its larger form, and in the Aurignacian of Europe.

The evidence in favour of the South African industry being an off-

shoot of the North African has here been set out without any other

considerations, but these will be discussed, though not in detail,

later.

This method of proof is unsatisfactory. It is necessary to see the

implements and hold them, to look at the working, and compare the

technique. The accompanying drawings are as accurate as possible,

and in the examples taken the technique is similar in all examples

grouped together.

The writer has attempted to prove a double case, that the Ibero-

Mauretanian culture of Western Algeria is the same as the ““ Pygmy ”

industry of South Africa, and that the Getulian culture of Eastern -

Algeria and of Southern Tunis is the same as the Smithfield industry

of Southern Africa. More precisely, the writer regards the South

African industries (7.e. the Pygmy and Smithfield cultures) as being

modern representatives of a form of Capsian culture which broke

away from the North African group before the purely geometrical

flints came into fashion.*

It will have been noticed in the preceding paper that the paintings

at Montagu are associated with implements of the Smithfield type,

and this is so in every case where paintings have been noted in the

same cave as implements. Obermaier (op. cit., p. 257) states that

‘“ The realistic paintings of Eastern and South-Eastern Spain came to

* A further link between these industries appears in Man. June 1925, vol. xxv,

No. 6, p. 51. where Messrs. Dewey and Hobley announce the discovery of imple-

ments of the Smithfield and Pygmy types near Njoro, Kenya Colony.

A Comparison between Capsian and South African Stone Cultures. 31

an end no later . . . than the Magdalenian, and should be regarded

as the regional equivalent of the realistic art of Northern Spain and

Southern France. Their makers were people of Capsian culture.”

On the following page he remarks that “ surprising similarities are

found between these Capsian paintings and those in South Africa

which are commonly ascribed to the Bushmen, and which certainly,

for the most part, are very old.”’ MacAlister remarks similarly.*

The question, however, still remains as to who were the artists of

South Africa? The only definite answer is “ A people with a Smith-

field culture.” The art has been attributed vaguely to the Bushmen,

but no attempt has been made to define the term.

There are three possible explanations :

1. The present Bushmen are the same race that invented and

disseminated the Capsian culture.

2. The original Bushmen obtained both the art and the culture from

a Capsian source.

3. There is a racial element common to both the Capsian peoples

and the Bushmen.

The first possibility is opposed to the fact that the modern Bushmen

are a mixed race, the mixture varying in different parts of the country.

The second possibility might possibly be ruled out by the fact that

the stone culture of South Africa is purely Capsian, and has no elements

of pre-Capsian “‘ Bushmen” culture, which would most certainly be

retained.

The third possibility appears to the writer to be the only practical

solution to the problem. But whether the Bush-Capsian people left

North Africa as a mixed race, or whether they left as a pure race

and became mixed further South, is a problem for the physical

anthropologist.

CORRIGENDA AND ADDENDA.

These two papers on the Montagu Cave and the comparison

between South African and Capsian cultures respectively, were written

some three years ago, and were set up in print at that time. Owing to

both the accumulation of material and the additional cost of printing

brought about by the war, it has only been possible to publish them

now.

During these three years a very considerable amount of work has

been done on the subject of archaeology in this country, and as a

* MacAlister, Textbook of European Archaeology, Cambridge, 1921.

32 Annals of the South African Museum.

result some of the findings given in these two papers are of less value

to-day, while certain fairly basic changes in nomenclature have had

to be adopted.

I make no apology for the publication of these papers. The first

stands with but little need for alteration. The elements described

as of the “ Earlier Stone Age” are to be regarded as of the Stellen-

bosch industry of that main period. The Later Stone Age cave-

deposits are to be regarded as an atypical group representing a variety

of the Wilton industry (then termed “* Pygmy ”’).

Similarly in the second paper, the term Wilton should be sub-

stituted in every case for the word “ Pygmy”; though the term

Smithfield stands as in the text. As our conceptions of the inter-

relations of the Capsian and Smithfield industries have altered very

considerably, much of this part of the paper is of little use. It seems

more feasible to regard the Wilton (“ Pygmy ”’) as a fairly pure form

of the Late Capsian industry, and the Smithfield as an early, localised,

derived form, which apparently evolved convergently towards the

Wilton.

The similarities existing between the Later Stone Age of the South

and the Capsian industries of the North are still regarded as extreme,

save for three elements :

(a) The apparent complete absence of the burin in the South

African Later Stone Age industries.

(6) The complete absence of pottery in the Upper Capsian.

On the other hand, the burin has now been definitely discovered

by Mr. J. J. Hewitt in a cave in Howison’s Poort, Grahamstown,

among material showing strong Middle Stone Age affinities. In

North Africa pottery of typical South African Later Stone Age

facies and also bored stones have been found at a variety of non-

Capsian sites.

The “ Eastern Culture ” has now had to be regarded as constituting

a distinct “‘ Middle Stone Age” with strong Mousterian affinities,

and it has again been subdivided into a variety of industrial groups,

the finest of which (the Still Bay) is apparently confined to the

extreme southernmost littoral of the Cape.

A. J. H. Goopwiy,

May 1928.

3. The Osteology of a Bushman Tribe. By D. Stoms, B.A.

From the Anatomy Department of the University of Capetown.

(With Plates VILI-X1.)

THE material which forms the subject of this work is a collection of

fifty-three skulls, together with a number of skeletons, which were

recently acquired by the South African Museum, Capetown, from a

burial-ground near Colesberg in the Cape Province. I am greatly

indebted to Dr. L. Gill, Director of the Museum, for his courtesy in

permitting me to investigate these specimens.

SOURCE OF THE MATERIAL

The skeletons which have been studied were exhumed in November

1926, from graves situated near the town of Colesberg, by Mr. J.

Drury, Modeller to the South African Museum, who is an experienced

excavator. He has kindly supplied the following note in connection

with this part of the work :

“These skeletons are the remains of what are supposed to have

been Cape Bush people, who died during the smallpox epidemic of

1866. The graves were situated just below a small range of hills,

and within about one hundred yards of the present Colesberg wash-

houses, which are about a mile and a half from the village. The

graves were arranged in irregular lines of one or two and sometimes

three deep, and extended for about fifty yards. All the graves had

a heap of stones on the top. Large heaps indicated adults, and small

ones indicated children. — All the bodies were buried in the extended

position. In some cases two bodies were found in one grave, and in

two cases there was no trace of bones in what had certainly been

graves. In one grave (No. 29) there were traces of the remains of

a wooden box or coffin.

“A few iron ornaments (bangles), glass beads, brass buckles,

buttons, and pins were found in the graves, and there were traces of

VOL xexdV PART le: 3

34 Annals of the South African Museum.

blankets and leather belts. In one grave a George IV shilling piece,

dated 1826, was found.

“The depth of the graves ranged from two to four feet, but the

majority were about two or three feet deep. The preservation of

the skeletons was on the whole not too bad. In the majority, however,.

a dry rot had set in owing to the infiltration of plant roots, and these

are tending to crumble away. The skeletons from these fifty-five

graves seemed to belong to a somewhat mixed group of individuals.

Many of them are good Bushman types, quite like the Kalahari

Bushman and the Strandloper. A few are Bantu (Kaffir) in type,

and there may be Hottentot types. It is possible that No. 29 had

some European blood.”

RAR Tele

The Crania.

A casual inspection of the general features of this group of skulls

is sufficient to show that they do not all belong to one homogeneous

racial group; nevertheless it is also clear that the majority of the

skulls are very good representatives of the South African Bush race.

Accordingly the skulls were separated into a relatively pure group

of forty-five Bushman types, and a small group of eight alien, and for

the most part Bantu types. The extraneous skulls, which were

excluded from the main group, were numbers, 5, 20a, 21, 24, 28, 34,

37 and 45. Skull No. 29 has a somewhat long and narrow nasal

aperture, but it presents very many characteristic Bushman features,

so that it has been retained in the Bushman group. This subdivision -

of the skulls was substantiated by an examination of the remainder

of the skeletons where such existed.

The skulls have not been classified according to sex, owing to the

difficulty in doing this with any degree of certainty on cranial criteria

alone. The male Bushman skull presents many of the infantile

characters which one usually associates with the female sex. In

those cases, however, where the other bones of the skeleton were

present, it was relatively easy to determine the sex of the individual.

This has, therefore, been indicated in dealing with the limb bones,

and to this extent the results for these may be applied to the skulls if

required.

A number of the skulls have suffered from posthumous deformation,

so that only twenty-nine of the Bushman types were sufficiently free

from distortion to permit of accurate measurements.

The Osteology of a Bushman Tribe. 35

Two of the skulls of Bushman type must have belonged to sub-

adults, so that they are omitted in determining the averages.

The greater part of this work has been devoted to a consideration

of the characters of the Bushman representatives in the tribe, but the

details of the measurements of all suitable skulls are given in the accom-

panying tables.

THe BusHMAN SKULLS.

The Cranium.

On the whole the crania present the type of curvature character-

istic of the typical Bushman calvarium (1). The forehead is high

and remarkably vertical. The supraorbital ridges are only slightly

developed, and in some skulls are entirely absent. The vertex of the

skull is flattened and depressed, and it passes back almost horizontally

from the frontal region. The occiput protrudes in a characteristic

manner. The contour of the norma verticalis is roughly ovoid. The

frontal region is narrow, and the parietal eminences project prom-

inently, especially in the female and sub-adult skulls.

In 40 per cent. of these skulls there are distinct grooves on the

frontal squame for the supraorbital vessels and nerve. In some

cases these structures tunnel the bone for a short distance.

The outline of the norma occipitalis of these skulls is pentagonal

with the upper angle of the pentagon rounded off.

The sutures are very simple in character, presenting In many cases

a simple wavy line. Only four of these individuals show any trace

of sutural occlusion, and this affects portions of the coronal and

sagittal sutures.

A patent metopic suture was noted in one skull (No. 23), and in

another skull (No. 53) there was a trace of this suture.

The pterion is predominantly of the H type. Epipteric bones occur

in two of the skulls (Nos. 29 and 39). Seven skulls show Wormian

bones along the lambdoid suture, but in three of these there is only

one insignificant ossicle.

On the base of the skull the planum nuchale is singularly free from

rough muscle markings, and the external occipital protuberance is

scarcely perceptible. The mastoid processes are characteristically

diminutive, and the occipital condyles are small, short, and flattened.

The glenoid fossae vary slightly in depth, but on the whole they are

somewhat shallow.

36 Annals of the South African Museum.

Measurements and Indices of the Cranvum.

The individual measurements on the various crania are given in

Table I. The averages of the various dimensions and indices are set

out below in tabular form, together with certain data for the Bushman

and Hottentot, quoted from various papers by Shrubsall (2), Martin

(3), and Broom (4), and indicated by S. M. B.

TInnear Dimensions in Millimetres.

Maximum | Maximum Basion-Bregma

Length. | Breadth. Height.

Bushman tribe (23) 174-2 | 130-6 122-2

Bushmen(S.) (29) 0. | is-eae aie 132-7 eee

Hottentots (S.) (19) . 183-2 | 133-5 | 130-6

The mean auricular height is 104-8 mm., the mean minimum frontal

diameter is 88-0 mm., and the mean maximum bimastoid diameter

is 112°3 mm.

Cranial Indices.

| |

Length-breadth I. | Length-height I | Breadth-height I.

| Bushman tribe . | (23) 74:7 (23) 70-4 (23) 93-6

Bushmen . | -(16) 74-9 (B.) | (16) 68-5 (B:) | (16). 91-6 (B.)

| 5 ; | (18) 75-5 (M.) | (18) 71-2 (M.) | (26) 91-1 (S:)

Hottentots (6) 69-3 (B.) (6) 69-5 (B.) (6) 100-3 (B.)

ni . | (15) 72-5 (M.) | (15) 69-5 (M.) | (19) 97-9 (S.)

With regard to the cephalic index this tribe is dolicho-cephalic,

with regard to the altitudinal index it is chamaecephalic, and with

regard to the breadth-height ratio it is metriocephalic. As regards

absolute size these skulls are typically Bushman, but the proportions

show a tendency in the Hottentot direction.

Cranial Capacity.

The cranial capacity of seventeen skulls was measured by Turner’s

shot method (5), and the results are tabulated below together with

GEST | LLTL| O18 | 9-FIT| 9-90L| 8-8OT| L911 | O-STL| #96 | F-O3T | $-90T * | 0-SZL| Z-SIL| O-911T| SET] ¢o-ZIT| °° bi ho ' XOpPUL Le[OOALB-O[]IX eI,

6S | GOL 19:36 | 38 |0O-88 | °° | 1-86 | &-98 | 1:88 |8:8 | 9:82 |F-6L |9:66 | F-18 .| 8:28 \e:28 | “~ e929): : xOpUl [eILGIO,

0:29 |¢-€9 |2:9¢ |8-09 | #19 |G-8o |0-0G |8-8E |Z-39 |G-29 |9-89 |629 |%19 14:99 |6@L¢ 14:99 |PLg |erG | c° | ° ; : ' XOpuUr [Vs NT

we +e +e an vs an G-Ze . ws 6-0¢ | @eg "+ | 9.e% pic .. ain oi LT +. : . Xopul perovy soddy

F:Z0L| #86 |G-101|/ 6-16 | °° "* | 9-6 | 0-8L "* |€66 | 0-16 | €-S6 | 6:6 | 4:96 | 0-26 | 0-88 | 8-001 | 6-96 | 9-Z0L| ° * xopuL YypBorq-7y soy

G@:ZL | T-Sh | 6-SL | 8-SL as “* |€1L |9:99 | °° | ¥69 | 629 |G:L | L:0L | 9:29 |6:89 |9-29 |6-1L |¥-69 | 3-82 | ° : * xopur [eUIpnyy[y

6-02 | G08 | 8-22 | 6-64 ; " |@:GL | G-L "* |TGL |6°8L |0-9L |8SL |6-69 |6-FL |6-9L |8-0L | L:1L |O-1L | * ; * xopur oreyda;)

COE P|] OASIS 28 S : oe 22 2 SOME COME OME POO =o AW Osi 2 "* 10-966] ° : 2 * Aqtoedey

O-LSP| O-S8F} O-S8F| O-18F) °° Z = ae * | O-I8F| 0-00E | 0-F9F | 0-E8F | O-ELF | 0-BE | G-98F | 0-G0G | 0-10 | 0-997 | * ; "ou [LYWOZTAOT]

O-LLZ | O-L8Z | 0-S6S | 0:76 25 = us "* | 0-946 | 0-842 | 0-69Z | 0-FLZ | 0-994 | 086% | O-TLZ | 0-82 | O-GLS | O-LES | * g * Od OSTOASUA,,

0:09 | 0-19E| O-FEE| O-SFE} °° fs Pe 2 ‘* | 0-69E | 0-F9E | 0-TEE | 0-99E | 0-0FE | O-SLE| O-SFE| °° | 0-9E8] O-1FE | TPIOL

O-1IT| 0-201 | O-FOL| O-20T} *° : ae oS ‘* | 0-601] 0-G1L] 0-ZOT| O-LIT| 0-86 | G-O1T| O-SIT| 0-921 | 0-201 | 0-SOT | * * yeqidro00

0-911] O-FIT| O-T8L| O-SOT} ° °° ie : as O-SIT | 0-ZZT | O-O1T | O-STT | O-STT | O-L2T | 0-601 | O-TET | 0-SET| O-LIT| ° * yeqorred

O-SET | O-OFT| O-GIL| O-SZT} °° aie oa a ** | O-GEL| O-LET | 0-611 | O-FZL | O-FZT | O-FET | O-1ZL | O-6LL| O-FIT| 0-081 | ° * eyUoIJ—ore [egqqTHVY

0-6F |0:1¢ |0-9F | O:8F | 0-8 | 0-9 |0-TG | 0-09 |0-E¢ | 0-6F | ¢-g¢ "* |0-FF | 0-89 |0-0S |0-SF | 0-2¢ we Be aa las * Yyysuo] a =

G-09 |0:09 | G-09 | 0-S¢ | G-99 | 0-G¢ |¢-6G |0-6¢ |0-€¢ | 0-69 |0-6E | 0-9¢ |0-¢¢ | 0-6¢ |0-8¢ |0-09 |¢-8¢ |0-09 | °° | ° YFPVolg LV[OOATV-OT[LXV IV’

G-3e | 0-63 |G:SE | Ese |O-SE | “* |GLE |GEE | °° | 0-08 | G-0€ | 0-82 | 4-63 |0-FE |G-18 |0-7E |O-SE |O-SE | °° | ° ; wl | is

0-2 10-62 |¢o-G8 | Ese |O-eE | “* |G9E |O-SE | *° |G-1E | 0-1 | ¢-8¢ |0-0E |0-Fe | G-1E | 0-EE | 0-ZE | 0-sE | 0-08 * Gysti—yysroy [eyIqIO

O-8E 10-28 |0-6E |O-8E |GLE | “° | G68 |G6E | ~° |G-:9E |0-2E 10-98 |G-LE |O-2E 10-68 [0-28 |O-2€ |O-8E | 7° | ° ol eZ 2

O-8E | 0-68 |G-2E |0-8E |GLE | *° |0-0F |G-8E | °° | GLE |G-2E | 0-98 | G-2e |¢-98 |G-8E |O-L4E |O-LE | 0-8 10-68 | ° * 4ysti—yypeorq [eHGIO

GRGtG:GGe |) PRs | HES | 2° at cis "* 10-66 | 0-F% | 9-93 | 0-S% |0-S3 | 0-86 |S-2o 10-43 10-93 | —° | - " -YyypRerq [ezIqGIOL0Z UT

G:9Z |0-L6 | G-S3 | 0-FS | 0-26 | 0-6 | 0-E% |0-E% | 0-46 | 0-93 | 9-93 |G-46 |0-F% | 0:83 |G-L% 10-43 10-28 |¢-6¢ |-7° | ° : * -Yyprorg [ese Ny

Zr |S-cr | O-Gh | 0-68 |S-Fr |O-LP |0-0E |0-2F | 0-€% | 0-0F | G-8F | ¢-0F.|¢-68 |0-2F |0-8F |G-0F |O-LF |O-LF | °° | ° : VPI * es

O-eh |o-3F | 0-SF |0-0F |G-8h |G-9F |0-0¢ |0-EF |G-EF |0-0F |G-EF | G-0F | 0-68 |0-ZF |O-LF |G-0F |O-LF |O-LP | °° | ° 2 Vy sti—yysroy [ese Ny

0-e¢ 10-89 |0-L9 |0-99 | °° “" 10-69 [0-89 | ** |0-L¢ |0-€9 |0-Lh |0-8F [0-3 |0-89 | ** |0-69 | 0-19 | 0-G¢ s Woryysoad-WoTse |

0:26 | 0:06 |0-16 |0:88 | ~° * |0-68 | 0-cOT| °° | 0-16 | 0-66 | 0-06 |0-€8 | 0-86 |0-16 | °° |0-#6 |0-SOT| °° zs woryysoid-woIseg

0-46 |0:66 | 0-76 |0-16 | ~° "> 10:06 | 0-S0T ' 10-68 | 0-76 | 0-76 | 0-68 | 0-96 | 0-L6 | 0-F6 | 0-LOT| 0-F0T | 0-9 fs UoIsvU-UOIseg

O-SIL| 0-631] O-L0T| °° a 2 “* | O-LOL] °° | O-80L| O-FIT| O-TIL| O-LTT| O-TIT| O-8IT | 0-911 | O-SIT | 0-901 | 0-GOT ce proqseuniq ss

ire oe oe o on : 0-021 oe OO O-ZIT | O-FIT oe 0-011 o's o O0 Od O-RII elle IoyoueIp oye uLosAZIq WNUIIXP Jy

0:16 |0:16 |0-76 |0-88 | °° = A028" 0226 * |0-68 | 0-96 | 0-68 | 0:26 | 0-16 | 0-021] 0-06 | 0-26 | 0-68 |0-L8 | ° doyourerp [eyUoIZ UINUATOT AL

0-LOL| &-S0L | 0-ZIL| 0-801 ’ "* 1 0-GOT| 0-011 O-FOL| 0-80 | 0-101 | 0-GOT | 0-66 | 0-011] 0-66 | 0-L0T| 0-FOT| 0-66 | * : * 4ysroy aepnonmy

0-081] 0-221 | 0-SET| 0-931] °° 0-FZI | 0-€01 ' | 0-061 | 0-11 | 0-121 | 0-E31 | O-LIT| 0-931 | O-LTT| O-ZET | 0-SZT | O-3T | * * gystoy vuIsorq -UOIseg

0-LZ1| 0-981 | O-E81| O-SET| °° " |O-181| 0-8} °° | 0-O€1| O-SET | O-LZT | O-TET | 0-121 | O-LET | O-SET | O-TET | 0-62T | 0-031 | ° ; qypeoeaq 4

0-6L1| 0-691 | 0-ILL | 0-691 : * | 0O-PLT| 0-381 * | O-ELT| 0-08T | 0-L9T | O-FLT | O-ELT | O-8T | 0-41 | 0-S8T | 0-08T | O-89T | ° : * YQsuey WNUWIXe

ty NOP RS | OB AB I SUS as a SAB) STE ar P| SIE I SSF Si Li € We ‘ * [pyg Fo toquinyy

‘OqIry, WeULYysSng

‘Saou yy Us DiUdIO fo SJuaWaWnsHa ;T—'] ATAV,

** |P-O1L| T1SL| 9-18] P-L1T| °° 98 PoPGIE | RSI) Zevaitit | °° ** |G-PIL| #-08T| °° oe 20. Gari | °° W.° * XOpUL IRpOOATe-OT[IEXe

8:18 | 4:68 "* 16:98 | 8:F8 ci PEGSSS “* 19:86 | 4:06 |G:8L |6-S8 | ¥-6L |6-IL | 9-L8 “> 10:76 | €:98. | ° : : XOpuUl [VqIGIO

6:39 | 9:09 | LFS “+ 1 8-Bg ore "* 18-29 |P-€9 16:69 | 679 | 2:99 |9-G¢ | T-LE |G-29 | @FE ** 19-89 Reel : : * xopul [eseNy

oe 6:8F se ae Z-6P ee ae oa se oe oe F£-0¢ oe oe oe oe 8:9F GPP oe . . xoput yerory reddy,

“+ | 1-001 | 8:66 ae "* 18:46 | 4:86 | L-L6 9 OO 22 ARO 2° || OIG. || aoe, "* 16:68 | 0-16 | $-88 * Xopur Yypvorg-yystopy

"* |9-FL | PL ce "* 16:32 | 6:32 | 9:92 * 10-PL eso. e7 “* 1@-Th | 3-9 * |€:0L | 0-22 |0:¢9 | ° : * Xopur [eUrIpnyypy

Sen GG ely: 2, "* 10-62 | 9-64 |0-FL | $-8L | 9:92 | O-FL 2 |} Bale, 28 A, |) eine, SS NPS, WROD, || AB * xoput oreydap

"+ | 0-31¢]0-21¢| °° si e "* | 0-99F | 0-FLF | 0-08F 0-28F| °° |0-S6F] °° ; "* | 0-9LF | O-L8F | : "OLB [RYUOZTIOF

“+ | 0-208/ 0-018) °° oe ap ee kO.GOCIKO-G9GIO:08Gila ta lOrcSGl) sane O:Giz |p aece 28 ** | 0-182 | 0-692 | ° : * OAR OSTOASTRAT,

"* 10-898] 0-948} °° Se Be "* | 0-9&8 | 0-6F§| 0:0G¢] 0-196] 0-998] °° |O-6FE] °° |O0-LFrE * | O-ZFE | 0-6EE | [eIOL,

"* |0-FIT|/ 0-901] °° te ae * | 0-801] 0-L0L| O-SOL| 0-OTT}O-SIL) °° |0O:-FOT] °° se "* | O-LIT] 0-901} ° * yeqidtoo0

2 | @PTESTE | ORES) = 25 ee 2 "* | Q-L01| 0-231] O-LIT| 0-131) O-GTT| °° |0-0ZT| °° | 0:8TT * | O-FOT | 0-221 * peqgotred

"* 10-G¢ |O-L¢ |0-6F | G-Lg se "* 1G-6p | ¢-Ch | 0-8P "* |O-LP |0-G¢ | 0-9F | 0-TE 3 2 | Opi 28 yysue, =“ és

"+ 1 0-F9.|¢-69 | G-F9 | G-L9 iS "* 10-19 | 0-9¢ | ¢-9¢ ne "* 10-89 | 0:09 | 0-€¢ | 0-E¢ * |¢-8¢ | 0-8¢ YApBVorq IV[OOATL-O][IXV Py]

G-6Z 10:98 |¢:08 |0-GE | G-E€ 2 ‘* 10-Fe | 0-8 |G-FE 10-66 |G-1E |¢-FE 10:82 | 0:82 | O-1E "* 16-48 | 0-28 ; 47°] si *

0-62 | 0-88 | ¢-08 |0-GE | ¢-EE oe "* |0-¢¢ | 0-8 |¢-9¢ | ¢-6% |¢-08 |¢-¢e¢ | 0-08 | 0-82 | &TE ' | 0-4 2 "4 SLI—4ystoy [e4IGICO

G-Gg | 0-1F "* |0-1F | 0-68 oes ** 1o-8€ | 0-68 |0-G8 10-38 |0-68 |0-1TF |0-98 |¢-68 | 0-98 : 0-88 | 0-8¢ eS) ns

0-98 | G-IP "+ 1G-0F | 0-0F nie "" 10-88 |0-88 | 0-98 |¢-2E |0-0F | ¢-0F | O-2E | G-8E | 0-98 "* 10-28 | 0-88 * QYSti—yyprorq [eIIqIG

0:23 a ema (KGS Gan. is 22 OBES a "* 10:62 | 0-8% | ¢-¢% | 0-22 | 0:9% | 0-61 ; O07 kee * -YApRoIg [e4F1q.10.109U |

0-E% | 0-08 | 0:92 20 || QGP ve "* 10-42 | 0:92 | 0-62 | 0-42 | 0-62 |0-0€ | 0-F% | 0:82 | 0-12 “* | 0-9z go ; : Ypproaq [ese

0-0F | 0-6 | 0-8F NOL 6 "* |0O-EF |O-1F |G-IF |0-cF | O-€F |0-F¢ |0-2F | O-1F | 0-8 0-PP ok : VT = ch

G-68 | 0-0¢ | 0-8F "+ | 0-9F ee "* |0-eF |O-1F |S-1F | 0-2F | 0-FF |0-F¢ | 0-ZF | 0-ZF | 0-68 ** | O-pP ere be : qysti—yyoroy [ese Ny

“* 10-89 | 0-99 ‘- 10-79. |0:99 10-89 |0-8¢ | 0-E¢ | O-E¢ 28° Oa “+ 10-8¢ | 0:0¢ -> 10:-T¢ | 0-¢¢ sr Me woryysord-WOIse Ny

"* | 0-FOT| 0:Z0OL| °° “* 10:46 | 0-06 | 0-G6 "+ | 0-68 “* 10-68 "* 10-68 | 0-66 0-88 | 0-88 vee eS uoryysord-woIseg

“* |0:001|0-c0T| °° “* 10-6 | 0-901 | 0:96 ee OzcO * 10°26 “* 10-86 | 0-86 * |0-88 | 0-16 | 0:88 ss UWOISVU-LOIStg

"+ |0-81TT|0-0¢L| °° ‘* 10-FIT| 0-611] 0-601 | 0-S0T| 0-O1L] ° 0-LOT OHA || °° ORR 2° | OAT 5 proyseurrg *

0 0-681 On site 0-081 she Bae gO ite ois : O-EI1 OO 00 OG =i 0-601 | 0-611 | 0-0ZTL |foqyoTUe Ip oreuros Az1q TUNUITX BTA]

nO 16) 10°86 "* 10:36 | 0-76 |0-L8 | 0-98 | 0:88 | 0:06 * | 0-16 * 10:16 | 0-46 0-68 | 0-86 | 0-6 | ° oyoureIp [eyUOAy CANUATUTY

“+ /Q-GIT| O-LIT| °° | O-TIT| 0-611} 0-TIT| 0-ZOL| 0-001} 0-90T| ° O-LOT! ** | 0-€0T| $-26 "+ | 0-001 | O-FOT | 0-ZOT | ° 5 " gystoy wepnoLiny

ZO Sele O:Sel| «a: “* 10:38L| 0-381] 0-421} °° | 0-831} °° |O-86T} °° |0-22T| 0-9T1 O-9IL| O-TZ1 | O-STI * qystoy eusorq -uoIseg

"> |O-LET|O-6ESL| °° | 0-881] O-SET | O-FEL| 0-08T| O-TET| O-82T| ° O-FEL| ** |O-FEL}O-S8T| °° | 0-631] 0-EET| 0-08T} ° Yypvorq pe

** 10-281] 0-98T| °° | 0-F8L] 0-181] 0-I8T] 0-991] O-TLT}O-SLT) °° |O-FLT] °° |O-TLT] 0-821] °° | 0-G91]0-89T] O-LLT] ° z * Yyysuoy vanwIxe yy

= ne on : os.. z peace i

ip | WS | SAB AB | Sts Mies as ep ail TSM AS eh WON ape Sep] SRP I eh | ap |) Seip | © i " [[MYG Jo coquinyy

sod Ay, ngueg Faee ‘oqtiy, UeULysSngy

“panurjuoo—saipouMupipy We nuns fo syuawmaimspa pf[—'] AIAV J,

The Osteology of a Bushman Tribe. 39

certain data bearing on them. The capacity is the lowest which has

so far been recorded for the Bushman. One of the skulls (No. 2) has

the remarkably low capacity of 995 c.c.

Males. | Females. | Both Sexes. |

Bushman tribe . : | (2) 1165" | (6) 1098) | (15) 1126

Bushmen (26), (Shrubsall, 1898) . 1330 1255 | Al

Strandlopers (7), (Shrubsall, 1907) oe ve | 1345

Bushmen (24), (Shrubsall, 1907) . | Hs: iN. 1285

Kalahari Bushmen (Shrubsall, | |

LOI) ; 1500 1100

Cape Bushmen (Shrubsall, 1911) 1260 1200

Strandlopers (Shrubsall, 1911). 1500 1350

The Face.

The face is broad, short, and retracted beneath the anterior part

of the cranium. The upper part of the face is always flattened, but

several of the skulls show a considerable degree of subnasal prog-

nathism. In one skull (No. 11) there is a subdivision of the zygomatic

bone into two (Os Japonicum). The upper facial index was deter-

mined in seven skulls and gave a mean value of 49-0, which is chamae-

prosopic. This is almost identical with Ried’s figure (3) of 49-2 for

eighteen Bushmen, and differs from his figure of 51-5 for fifteen

Hottentots.

The orbits are rectangular in outline, with the transverse axis

directed laterally and downwards, and they are widely separated.

This appearance of a wide interorbital interval is heightened by the

flat nasal bridge.

The average orbital index for this group is 85-2, so that the orbits

are mesosemic. This index varies greatly in the different crania,

but the tendency is in the microsemic direction. Ried’s figures for

eighteen Bushmen and fifteen Hottentots are 81-4 and 79-6 respec-

tively.

The nasal bridge is flattened and in line with the forehead. The

lower margin of the nasal aperture is sharp in 50 per cent. of the skulls,

in 11 per cent. there are well-marked subnasal fossae, and in the

remainder there is a moderate degree of simian guttering. The

anterior nasal spine is diminutive or absent.

The mean nasal index of 61-2 is markedly platyrhinic ; the maxi-

40 Annals of the South African. Museum.

mum index is 69-9, and the minimum 53-5. Ried gives an index

of 59-9 for eighteen Bushmen, and 57-9 for fifteen Hottentots.

The average breadth of the upper alveolar arch, measured from the

outer alveolar border opposite the second molar teeth, is 60-2 mm.

The average length of the arch, measured from the prosthion to the

middle of a line joining the posterior extremities of the alveolar

border, is 49-8. The absence of maxillary tuberosities is a character-

istic feature of these and of Bushman skulls in general.

The average maxillo-alveolar index is 117-5, which is brachyuranic.

The dentition of this tribe has been made the subject of a special

study by Professor M. R. Drennan (6).

The degree of prognathism is often expressed by the alveolar index,

which is the ratio between the basion-prosthion length and the basion-

nasion length. This does not, however, take into consideration the

vertical displacement of the prosthion, which is a very important

factor as regards the shape of the face. It is advisable therefore to

take into consideration the three sides of the facial triangle, namely

the basion-prosthion, nasion-prosthion, and basion-nasion lengths.

In the following table the basion-prosthion length and the nasion-

prosthion length are expressed as percentages of the basion-nasion

length. This gives the degree of forward displacement or prognathism,

and of the downward displacement or “ subgnathism.” The corre-

sponding values for the Bushman, Hottentot, and Kaffir, calculated

from Shrubsall’s data, are added for comparison.

| Basion-prosthion. | Basion-nasion. | Nasion-prosthion.

Bushmen ‘ idl

| 98-5 100 63-4

Bushman tribe (20) | 99-8 100 59-1

Hottentots . ey 101:3 100 66-6

Kaffirs (40) 100-9 100 65:8

The Mandible.

In all the skulls of this tribe the lower jaws are small. The rami

are low and almost square in appearance, and the mandibular notch

is shallow. The condyles are flat, and roughly oval in outline. The

inner surface of the symphysis slopes downwards and backwards,

and the chin is poorly developed. Accordingly if one looks vertically

down on the symphysis menti, there is more of the bone to be seen

Number of

Skull

The Osteology of a Bushman Tribe. 4]

Tasie I1.—Measurements of Mandibles in Millimetres.

a ire = A é

H = on Oar |e ok eS =) =

Ssol|/sd/2£mIlS a/R S/R S| Sol ton] Boalss

esiSselaoeP|/HeiAs As l(aa|oel|seel|sey

Seles |s8| se) e8 | | se | On| ed | sh

ge /Ra gt SF es 2/5 eMaE

= = B

87:0 | 72:5 | 95:5 | 46:5 | 32:3 | 36-5 | 33:0 | 22-5 | 120-0 | 109-8

98-0 | 43°5 | 33°8 | 38-5 | 31-0 | 26-0 | 125-0 ig

101-0 | 91-0 | 100-0 | 46-0 | 32:0 | 35:5 | 29:0 | 28°5 | 130-0 | 99-1

88-0 : 53-0 | 33°8 | 44-0 | 34:0 | 31-5 | 121-0

73°5 | 93-5] 51-5 | 37-0 | 44:0 | 25:0 | 27-0 | 112-5 0.6

105-0 | 84-0 | 95:5 | 41:5 | 33-5 | 39-5 | 24:0 | 25-0 | 1280 | 91:0

108-0 | 92-5 | 102°5 | 46°0 | 37-5 | 45:0 | 25:0 his 121-0 | 94-9

88-0 | 80-0 | 95:5 | 40-0 | 32:0 | 35-0 | 29-0 | 28-0 | 1195 | 108-5

86°5 | 100-0 | 45-0 | 37-5 | 40-0 | 31-0 | 28-5 | 114-0

80°5 | 95:5 | 42-0 | 31-8 | 37-5 60 24°5 | 127-0 a

87-0 | 76-5 | 94:0] 45:0 | 36:0 | 40-8 | 30-0 | 26-0 | 119-0 | 108-0

105°5 | 84:5 | 100-0 | 48:5 | 36:0 | 39-0 | 32°0 | 26°5 | 121-5 | 94:8

101°5 | 78:0 | 94:0} 44:0 | 33-5 | 37-3 | 26-0 | 25:0 | 117-0 | 92-6

97-5 a 31-0 Be 31-0 oh 117-0 ws

101°5 | 86-5 | 93-5 | 43:0 | 32-0 | 36-5 | 30-0 | 26:5 | 127-0 | 92:1

103-5 | 86-5 | 96-0} 48:5 | 33:5 | 40:5 | 34-0 | 28-0 | 115-0 | 92:8

98-5 | 91-0 | 103-0 | 49°5 | 28°5 | 35-0 | 30-0 | 27-0 | 129-0 | 104-6

96-5 | 75-0 | 101-5 | 45-5 | 39-5 | 42°8 | 31-0 | 30-0 | 116-5 | 105-2

109-0 | 86-0 | 105-0 | 55°5 | 40-5 | 46-8 ahs ue 110°5 96:3

99:0 | 81:5 | 94:0} 40-0 | 33°8 | 39-5 | 33:0 | 30-0 | 126°5 | 94-9

105-5 | 93-0 | 99-0| 39-0 | 35:0 | .. ae |) 26:0) |) 131-0 | 693-8

95-0 | 75:0 | 94:0} 48-0 | 36:0 | 40-0 | 31-0 | 30-0 | 113-0 | 98-9

101-0 | 84:0 | 101-0 | 46-0 | 31-3 | 40-5 | 30-0 Be 138-5 | 100-0

104-0 | 86:0 | 93-5 | 45:0 | 34:0 | 38-0 | 24-0 | 27-0 | 123-0 | 89-9

84:0 | .. | 43-5.) 32-5 | 37-0 | 32:0) .. He

90:0 | 79:0 | 95:0) 44-5 | 29-3 |/ 33-5 | 30-0 | 26-0 | 126-5

92-5 |-79-0' | 93-0)| 38:0 | 31-3 | 35-5 | 31:0 | 26:0) 125-5

111-0} 80-0 | 99-5} 49-0 ihe 23-0 | 26-0 | 120-0

108-0 | 83-0 | 93-5 | 43-5 | 33-3 | 34:8 | 32-0 | 27-0 | 117-0

112-5 | 91-0 | 102-0 | 57-3 | 38-0 | 43-0 | 34:0 | 31-0 | 115-5

79-0 a6 48-5 | 31-5 | 36-5 sie 28-0 | 123-0

107-5 | 81-5 | 96-0 | 40-0 | 33-0 | 38-5 | 26-0 | 28-0 | 131-5

91:5 | 94:0} 47-0 | 30-0 | 37-0 | 27-0 | 26-0 | 116-0

87:5 | 74:0 | 94-0] 42-5 | 32-0 | 35-0 | 28-0 | 20-0 | 123-5 | 107-4

94:5] 71:0 | 95-0! 47-0 | 33-3 | 36-0 | 30-0 | 27-5 | 118-0 fe

99:0 | 90-5-| 98-0] 49-0 | 38-5 | 42-0 | 32-0 | 31-5 | 120-0 | 99-0

87:0 | 70-0 | 83-5] 38-5 | 28-0 | 31-8 | 22-0 es 123-0 | 96-0

91-0} 81-0 | 95-0} 43-0 | 33-8 | 37-0 | 28-0 | 26-0 | 121-0 | 104-4

ae a 98-0 | 51-5 | 33-0 36 30-0 | 27-0 | 117-0

112-0 | 80-0 Ke ats 36:5 | 42-0 ails 34:0 | 117-0

49-5 | 36-0 | 41-0 | 32-0 | 27-0 | 119-0

oe oe oe 55:5 | 33:0 | 39:0 | 30:0 ) 27-0 | 110-0

112-0 | 85-0 ah 53: 36:0 | 48:5 | 37-0 | 32-0 | 125-0

42 Annals of the South African. Museum.

behind the teeth than in front. This is in contrast with the condition

of the Huropean, where there is more bone seen in front of the teeth

than behind. Measurements of the individual mandibles are given

in Table IT.

Average Measurements of the Mandibles.

Adults. Subadults.

Bicondylar breadth (24) 100-8 (5) oes

Bigonial breadth. (32) 80-5 (}) Rawics

Mandibular length (29) 97-2 (5) 93-1

Symphyseal height (29) 29-6 (5) 28-0

Length of ramus ys (32) 45-7 (5) 44:0

Minimum breadth of ramus . | (32) 33-8 (@) sisell

Maximum _,, | (30) 38-9 (5) 36-4

Body height a (27) 27-2 (4) 26-3

Mandibular angle sul (32) 121-6 | (5) 121-1

index | (18) 98:2 (4) 101-7

THE Limp Bones.

In addition to the skulls there were portions of the rest of the skeleton

of thirty individuals. Some of the skeletons are practically complete,

whilst others comprise merely a few fragmentary long bones. Of the

available material three skeletons are subadults, so that the measure-

ments of the bones of these and of four skeletons, which with their

skulls show Bantu characters, are excluded from the averages. All

the bones of the remaining group of individuals are small and delicate

in character. This feature, together with the various proportions

which are set out in what follows, confirms the view, already arrived

at from an examination of the skulls, that these individuals must be

regarded as Bushmen.

The Femur.

The femora are diminutive, and they appear to display a marked

degree of torsion. The shaft has a pronounced forward curve, and

there is a well-marked linea aspera. Only in one case was there a

trace of a third trochanter. The measurements of the individual

bones are given in Table III. The average length of paired bones

irrespective of sex, and the results for the two sexes irrespective of

side, are tabulated on p. 44.

The Osteology of a Bushman Tribe.

O-9ZI | O-NZT | 0-93 | O-8S | GTE | G-1E | 9-OT | €-€6 | 0-82 | 0-08 | 0-62 | 0-82 | o-8LF | 0-08F | G-18h| ¢-I8F| LP 08

6G-SOT | 2-901 | 23s | GBs | GFZ | OFS | FL | EES | GES | O-FZ | 0-08 | 0-04 | 0-088 | G-84e | G-u8E| Gose] & 1%

O-8ZI | O-8ZI | 0-83 | OSS | O-ZE | O-ZE | 4:26 | LOT | GLB | O-LS | G-Ss | 0-86 | G-0GF | O-8FF | O-ZEF | G-ocF| L ¢

€-90T | 0-001 | 0-FZ | O-FS | 9-93 | O-FZ | 6-GL | 6-1L | 0-28 | G-8B | G-0Z | GOs] °° zi e } G

L-OIT| GGIT | 0-13 | 0-33 | 94S | 9-96 | F-68 | 0-98 | G-€S | 0-3 | 0-12 | G1 | O-8LE| O-8LE| O-T8E] O-sse| 3} Zl

6-LIL | G-ZIT| G61 | 0:08 | 0-3 | GZS | €Z8 | 0-08 | G-Es | 0-E% | 0-1z | 0-08 | 0-18€ | 0-E8E | 0-F8E | 0-¢ i I

O-LIL | T-61T | GES | GES | GLB | 0-83 | L:9L | L:08 | 0-0€ | G-8zs | 0-€% | 0-€% | 0-SOF | 0-LOF | 0-90F | O-6OF| L gg

S:LE1 | G-08T | 9-33 | 0-63 | O-1E | 0-0€ | &-FOT | 8-86 | 0-3 | 0-%Z | 0-F% | G-Bs | 0-08E| O-18E | 0-Z8E| G-I8g| @ Ig

SCIT | SOIL] O-€3 | GIS | 9-92 | O-G3 | PSL | 9-6L | GSS | GFZ | 0-02 | GEL | 0-90F| O-GIF| O-ZIF| O-91F| 3 SF

O-LIT | T-61L | 0-13 | G13 | GFZ | G-F% | 9-46 | 0-00T| 0-12 | G-1% | G-0B | G-1z | G-ILe| G-sLe] G-sLze|G-rLE| S Lt

G91 | GLE | G13 | G-1Z | OG | 6S | SL8 | 6-8 | GES | GES | GOS | GEL | O-T8E| 0-618] O-F8E| O-zsE| 3d OF

G-6IT | S-6IT | 2:03 | G03 | GFZ | GFZ | F-68 | LG6 | GES | 0-3 | 0-18 | 0-ZS | 0-26E | 0-F6E | G-968 | 0-968] S bP

T-61T | 9-011 | 8% | G-€S | 0-86 | O-L% | 9-FL | LEL | 26S | 8S | 0-Bs | 0-18 | O-LOF | 0-LOF | G-LOF | 0-80F| LP OF

0-0ZI | F-STL | 0-Es | 0-92 | 0-0€ | 0-0€ | 1-96 | T-68 | G-8% | G-Lz | FS | G-FS | 0-60F | 0-60F | 0-60F | 0-60F| L €€

6-96 | S-8ZI | 0:93 | 0-9 | O-GE | GEES | 4:98 | LFS | 0-08 | G-6% | 0-98 | 0-9 | G-9zF | G-O8F | O-TeF | O-EeF| LP LG

PPS | O-TST | GS | O-1Z | 0-83 | GLz | 4-28 | F-08 | 0-96 | GS | GS-18 | G-0B | 0-€6E | 0-F6E | 0-86E | 0-868] LP 9%

6-901 | G-FOL | GS-13 | 0-3 | 0-3 | O-€S | 9-LL | E18 | G-#s | O-FS | 0-61 | GGL | G-698 | O-T9¢ | 0-198] 0-99¢] 3 61

G-901 | €-FOT | 0-3 | 0-€% | G-FS | O-FZ | 8-9L | GFS | 0-83 | G-9B | GIs | GZS | 0-69] °~* | 0-028] 0-:99E| & 81

T-Z01 | 6-26 | 0-F% | 0-F% | GS | GES | 8-8 | F-6L | 0-6% | 0-62 | O-FZ | 0-8 | 0-G6E | 0-F6E | 0-968] O-S6E} 3S 91

G-LZI | O-8SL | SSS | O-GS | G-ZE | O-ZE | G-LOT | 6-TOT | 0-28 | 0:93 | GLB | G-9B | O-OFF | O-ZEF | O-1FF | GOFF] 4 a

GFIT | 6-32 | GFZ | O-FS | 0-83 | G-6S | G28 | 9-78 | 8S | 8S | Ges | GES | O-SIF] O-FIF| O-GIF| O-FIF] 3 él

E801 | 8-021 | 0-73 | 0-FZ | 0-93 | 0-63 | 8-08 | FSS | 0:92 | O-L3 | 0-12 | 0-83 | O-LIF| O-ZIF| O-GIF| O-IF] L& II

0-611 | 6-01 | 0-12 | GIS | 0-GS | 0-96 | 0-48 | 96 | 0-E% | G-1d | 0-02 | G-0% | G-9G8 | 0-8Ge | 0-E98] O-e9e] 3} Ol

9-O€1 | FSET | GFZ | O-FS | O-ZE | G-ZE | 8-LOT | E-LZT | GSS | 0-43 | GLB | 0-88 | O-1ZF| O-TIF| 0-2] O-GIF| L 6

6-801 | L801 | S-3s | 3S | GFS | 0-93 | 6F8 | G18 | G-9B | O-LZ | G-Zs | 0-GB | O-G8E| O-E8e | O-E8e] O-sE] 3} 8

0-901 | GEOL | 0-93 | Ge | G-96 | G-9Z | 8-18 | 9-LL | GLB | 06% | GBs | 9-3s | 0-9GF | O-LZF | O-6ZF | O-eeF| 3S L

L-TIL| 1-111] $-8% | $s | 0-96 | 0-86 | L-0L | BIL | 0-62 | G-8% | G-08 | G-08 | G-EsF | O-SZF | G-98F] O-1eF]| S €

1-601 | 1-601 | 0-3% | 0-32 | 0-F% | O-FS | 9-Z8 | L:98 | 0-ES | GZS | O-GT | G-6T | F-Z6E | 0-988 | 0-268] G-98E] 3S z

HOT PITS | WOT PFISEL | WOT [FIST] HET [TUSRL]) WET [ISL] WET |IUsRT| WET PSL] “ET [yt

“TOPOULB I(T “LOPOUVICT “LOPOUTCI(T “LOPOULVICT “Yygsuory “yyouory eae as

Fas aSIOASUBIT, | LoLteysod-quy “xopuy asmoasuviy, | Lortoysod-quy onbyqo TUNUUIEX BJT a 2 t

. xopuy onroursqey Taq wan Ny

SULLOYSBI LE tortod ng == —

“‘DULLOYSELT ‘errouAyetg todd, “yyouory

‘nwa ~— [I] LTAV YL,

44 Annals of the South African Museum.

Right. Left. | Males. Females,

Maximum length

(22) 401-0 | (22) 400-9 | (14) 412-8 | (26) 392-8

Oblique 3 (21

) 400-0 | (21) 400-0 | (14) 410-6 | (25) 391-0

The average oblique length of forty-three bones irrespective of side

and sex is 399°3 mm. This figure may be compared with Vermooten’s

results, quoted by Martin (3) :—

Strandloper. | Inland Bushman. Hottentot.

Bicondylar length . | (8) 407 (36) 419 (6) 433

| |

The superior platymeric index in this group is 80-1, and the pilaster-

ing index is 115. This constitutes a marked degree of platymeria,

and a moderately pronounced degree of pilastering. In one individual,

No. 40, the upper half of the femur was so markedly flattened that a

distinct flange-like border was produced on the medial side of the bone.

The Tibia.

The standard length, that is to say the maximum length without

including the spine, was measured, and the results for the individual

bones are given in Table IV. The average length of paired bones

irrespective of sex, and the results for the two sexes irrespective of

side, are tabulated below.

Right. Left. | Males. | Females. |

| Standard length | (21) 328-0 | (21) 328-0 | (14) 337-0 | (25) 320-8

The average standard length of forty-three tibiae irrespective of

sex or side is 327-7 mm. These measurements are remarkably low,

and in great contrast to Vermooten’s results, which were as follows :

Strandlopers. Inland Bushmen. | Hottentots.

8) 345 mm. 35) 353 mm. | 8) 373 mm.

! |

The Osteology of a Bushman Tribe. 45)

TaBLE IV.—Tibia.

Num- Length. Platyenemia.

ber Platy-

of |Sex. Sia cnemic

Skele- Maximum | Standard | Ant.-post. | Transverse; Index.

ton. Length. Length. Diameter. | Diameter.

2 Q | 335-0 | 333-0 | 330-0 | 329-0 | 25-5 | 25-5 | 20-5 | 19-5} 80-4 | 76-5

3 Q | 355-0 | 355-5 | 345-5 | 347-0 | 32-0 | 32-0] 21-5 | 21-0} 67-2 | 65-6

7 Q | 361-0 | 360-0 | 355-0 | 355-0 | 29-5 | 31-0 | 21-5 | 21-0 | 72-8 | 67-7

8 Q | 309-0 | 303-0 | 302-0 | 297-0 | 28-0 | 27-0 | 19-5 | 19-5 | 69-6 | 72-2

9 36 | 357-0 | 356-0 | 349-0 | 350-0 | 34-0 | 36-5 | 24-0 | 24-5] 70-6 | 67-1

10 Q | 306-0 | 320-0 | 305-0 | 314-0 | 25-0 | 27-0 | 17-0 | 18-5] 60-7 | 64-8

11 6 | 351-0 | 346-0 | 344-0 | 336-5 | 32-0 | 31-5] 20-0 | 20-0} 62-5 | 63-5

13 Q | 345-0 | 342-0 | 335-5 | 337-0 | 35-0 | 33-0 | 22-0 | 23-0 | 62-9 | 69-7

15 2? |373-0 | 373-5 | 367-5 | 366-5 | 32-0 | 33-0 | 25-0 | 23-5 | 78-1 | 71-2

16 Q | 329-5 | 331-0 | 324-5 | 325-0 | 31-5 | 33-5 | 24-0 | 24:0] 76-1 | 71-6

18 Q .. |319-0) .. |313-5| 27-5 | 28-5 | 23-0 | 23-0 | 83-6 | 80-7

19 Q |311-0 | 307-5 | 306-5 | 303-0 | 26-0 | 23-5 | 19-0 | 19-0} 73-1 | 80-9

26 3g | 329-0 | 328-0 | 313-0 | 313-0 | 35-0 | 34-5 | 21-0 | 20-5 | 60-0 | 59-4

27 S | 369-0 | 374-5 | 362-0 | 366-5 | 34-0 | 34-5] 22-0 | 21:0 | 64:7 | 60-8

33 5 | 346-0 | 348-0 | 335-0 | 338-0 | 33-5 | 35-0 | 21-0 | 21-5| 62-7 | 61-4

40 3g | 343-0 | 343-0 | 339-5 | 341-0 | 33-0 | 35-0 | 21-0 | 22-0 | 63-3 | 62-9

44 Q |316-0 | 316-0 | 310-0 | 308-5 | 25-0 | 26-5] 19-0 | 18-0 | 76-0 | 67-9

46 Q | 322-0 | 321-5 | 311-0 | 314-0 | 26-5 | 26-5 | 21-0 | 20-0 | 79-2 | 75-5

47 Q | 302-5 | 305-0 | 297-0 | 299-5 | 26-5 | 26-0 | 20-5 | 19-5] 94-3 | 75-0

48 Q | 337-0 | 336-0 | 328-5 | 328-0 | 27-5 | 27-5 | 20-5 |19-0| 74-5 | 69-1

51 ? | 315-5 | 316-0 | 310-5 | 307-0 | 30-5 | 30-0 | 23-5 | 22-0 | 77-0 | 73-3

53 6G | 325-5 | 319-5 | 317-0 | 313-0 | 32-0 | 30-5 | 25-5 | 20-0] 69-7 | 65-6

1 ? |317-5 | 319-0 | 310-0 | 312-0 | 27-5 | 27-5] 19-5 |19-5| 70-9 | 70-9

12 Q |312-5 | 312-0 | 308-0 | 306-0 | 28-0 | 28-0 | 22-5 | 22-0 | 80-4 | 78-6

2 Seen o4-o20)) tere a 28-0 | 27-5 | 20-0 | 20-5 | 71-4 | 74.5

5 & | 382-0 | 381-5 | 370-0 | 369-0 | 36-5 | 34-5 | 25-0 | 25-0 | 68-5 | 72-5

24 Q | 320-0 | 317-0 | 315-5 | 310-5 | 29-0 | 28-0 | 22-0 | 22-0 | 78-9 | 78-6

30 & | 401-0 | 402-5 | 394-0 | 395-0 | 38-0 | 39-0 | 27-0 | 27-5 | 71-1 | 70-5

46 Annals of the South African Museum.

The diameters of the shaft at the level of the nutrient foramen were

measured and compared, giving a platycnemic index of 71-9 for the

right side and 69-4 for the left side. This constitutes a less pronounced

degree of platycnemia than is usually given for the Bushman, and the

figures may be compared with the index of 68-7 given by Shrubsall (2)

for the Bushman, and with Vermooten’s results, which were 64:6 for

thirty-seven Bushmen, and 70-1 for eight Hottentots.

The tibiae of this tribe exhibit a marked degree of torsion. They

also have facets on the anterior aspect of the distal end of the bone,

which articulate with corresponding facets on the neck of the talus in

the extreme dorsi-flexion of the ankle joint accompanying the habitual

squatting posture.

The Fibula.

The fibula in this tribe is always very slender and delicate, so that

in many cases it is quite fragmentary. The measurements of the indi-

vidual bones are given in Table IX, and the results for the two sides

and the two sexes are tabulated below.

| |

| Right. | Left. Males. Females.

| |

Maximum length | (4) 321-8 | (4) 321-0 | (8) 333-6 | (8) 303-4

| |

The average length of fifteen fibulae is 319-5, which may be compared

with Vermooten’s results :—

Strandlopers. | Inland Bushmen. Hottentots.

| ()336mm. - | (24)340mm._ i (8) 360 mm.

The Patella.

The shape of the patellae shows the usual modification found in

primitive races as the result of the habitual squatting posture. Thus

there is a well-marked perpendicular facet. An interesting feature

is the occurrence of a “ vastus notch ” for the attachment of vastus

lateralis, which was present in nine cases. The mean maximum height

of the patellae is 36-3 mm., the mean maximum breadth is 36-1, and

the mean maximum thickness is 18-3 mm. The individual measure-

The Osteology of a Bushman Tribe.

TaBLE V.—Patella.

Max. Length. Max. Breadth. Max. Thickness.

Number of Sex = |

Skeleton, Te MRichta selelts i Rishte\jeelWett al Richt Il) Bete

2 Q 34-0 is 32-0 x 15-0

3 2 38-0 38-0 38-0 39-0 19-0 19-0

7 Q 35-0 Ns 34-5 Be 18-0 as

8 Q 33-0 af 31-5 Pe: 16-0 BD

9 3 a: 38-0 oe 37-0 oe 16-0

11 3 38-0 39-0 41-0 41-0 18-5 20-0

15 ? 38-0 40-0 34:0 16 18-0 19-5

26 3 38-5 38-0 37-0 37-0 20-0 20-0

27 3 42-0 a 42-5 20-0 ie

33 3 31-5 Be 35-0 AP 17-5

40 3 43-5 S oe Se 23-0 ae

44 © 32-0 35-0 a6 33-0 17-5 19-0

47 Q 32-0 31-5 32-0 32:0 18-0 18-0

48 Q 38-5 39-0 39-0 39-0 18-0 19-0

51 ? 35-5 36-0 39-5 39-0 18-0 18-0

53 3 40-0 eid 18 oe 22-0 :

1 ? 31-5 31-0 30-5 31-0 14-0 15-0

12 Q oe 34:0 ee 35-0 Hs 17-0

30 3 49-0 47:0 | 45-0 42-5 24-0 25-0

TaBLeE VI.—Calcaneus.

INambernot Length Breadth. Height.

Skeleton, | “= |

‘ Right. Left. Right. Left. Right. Left.

2 $ 68-0 68-5 21-0 21-0 30-5 30-5

7 Q 75-0 76-0 23-0 22-0 31-5 32-0

8 @ 66-5 65-5 22-0 iin 32-5 33-0

9 3 i 72-5 ke 24-0 i &

10 Q 66-0 68-0 ie 19-0 30-0 30-5

11 3 69-0 69-0 22:0 22-5 34-0 34:5

15 @ 74-0 73-0 ou 65 33-0 33-0

16 2 71-0 70-5 24:0 24-0 30-5 31-0

18 Q 66-5 re a fe De a

26 3 65:5 66-5 a0 25-0 ane 32:5

27 3 71-0 73-0 22-0 24-0 31:0 32-0

33 3} 70-0 71:5 25-0 25-0 35-0 34:5

40 3 72:5 70-5 23-0 24-0 32:5 34-0

44 2 66-5 62-0 Ae 22-0 33-0 32-0

46 Q 5.0 64-5 50 20-0 oO 30-5

47 2) 516 66-0 20-0 20-0 28-0 28-0

48 Q Be 67-0 a 22-0 Be 32-0

51 ? 68-5 72-0 22-0 23-0 31-0 31-0

53 3S 69-0 68-0 22-0 22-0 34-0 34:0

1 ? a : bd ae

12 Q 64-0 es 20-5 29-5

25 Q 70-5 a 21-5 : 28-0 Bi

5 3 mis a 26:5 28-0 36-5 37-0

24 Q 68-0 68-0 Me ee S: 30-5

30 3 83-5 84:5 30-5 30-0 43-0 43-0

48 Annals of the South African Museum.

ments are given in Table V, and the results for the two sides and the

two sexes are tabulated below.

Length. Breadth.

Right. Left. Right. Left.

Male. | (6) 38-9 | (3) 38:3 | (4) 38-9 | (3) 38:3

Female . | (6) 34-9 | (5) 35-5 | (5) 35:0 | (5) 35-0

The Calcaneus.

The proportions of thirty-two paired and seven single calcanei were

determined. The average length of thirty-three is 69-3 mm., the

average breadth of twenty-seven is 22-4 mm., and the average height

of thirty-one is 32-0 mm. The smallness and relative narrowness is

the feature of these dimensions. The measurements for the individual

bones are given in Table VI, and the results for the two sexes and the

two sides are set out below.

| Length. Breadth, Height.

Right. Left. Right. Left. Right. Left.

Male . | (6) 69-5 | (7) 67-8 | (5) 22:8 | (7) 23-8 | (5) 33-3 | (6) 33-7 [

| Female . | (8) 69-3 | (8) 70-1 | (6) 22-0 | (7) 21-1 | (8) 31-0 | (8) 30-9

Observations were made on the number and shape of the facets on

the calcaneus, which articulate with the talus. In all the calcanei

belonging to this Bushman group only two facets were found in place

ot the more usual number of three found in Europeans. According

to Hrdlitka (7) the percentage occurrence of two facets in Huropeans

is 25 per cent. in males, and 40 per cent..in females. The presence

of only two facets in 100 per cent. of the Bushman calcanei is probably

associated with postural habits and greater mobility of the foot. In

two of the non-Bushman individuals there were three facets, the

anterior and median facets being separated in one case by a wide gap

and in the other case by a slight ridge. .

The Osteology of a Bushman Tribe. 49

The Talus.

These bones are very small. The mean length of thirty-four bones

is 49-4 mm., the mean breadth of thirty-five is 35-7 mm., and the

mean height of thirty-seven is 26-7 mm. The measurements of the

individual bones are given in Table VII, and the average dimensions

Taste VII.—Talus.

Nees Length. Breadth. Height.

(o) pex.

peclton: aah, | Tews | Sea | ees, | TRF | ibe

2 Q 48°5 50-0 35:0 35°5 26-0 Nes

3 Q 52-5 | 51-0 38:5 38-0 27°5 27-0

7 Q 49-0 ae 34:5 me 26-0

8 @ | 485 RO) oly de ill EON bee

9 3 ae 42-0 ne 31-0 i 23°5

10 fe) AT-5 47-0 33°0 33°0 24-5 24-0

Wl é 49-0 52-0 31°5 37-0 26:5 28-0

15 Q 50-5 52-5 aie 36-0 27:0 27-0

16 Q 49-0 48-0 34:5 35:0 26-0 26-0

18 fe) ae 44-5 39:0 35:0 23:5 23-5

26 3 50-0 ID) 36°5 38-0 27°5 27-0

20 3 5D:d 58-0 39°5 40-5 30:0 31-0

33 So 52-5 52-0 38:5 ad) 28-0 28-0

40 é 49-0 49-0 35:0 36-0 26-5 27-0

44 Q 51-5 48°5 34-0 34:5 28-0 27°5

46 Q 43-0 42-0 32-0 32:0 24-5 25-0

| 47 Q 45-5 44-5 33°5 35-0 25-0 25-0

| 48 Q 49-0 48-0 3D:5 34°5 27-0 28:0

| yl Q 54-0 52-0 36-0 37:0 27-0 28-0

| 53 3 37-0 29-0 | 28-5

il 2 48-5 50-0 36:0 37:0 24-0 24-0

12 Q 44-() 45-0 30° 34:0 24-0 24-0

25 je) Ae

5 3S 53:0 53°5 40-5 41-5 30:5 31°5

| 24 Q 49-0 49-5 35°0 35:0 28:0 27-0

30 3 62-0 62°5 46-0 46:5 30°D 35:0

according to the side of the body and the sex of the individual are

given on p. 50.

Vi Ole XOX Vin, PAVE Ae 4

50 Annals of the South African Museum.

Length. Breadth, Height.

Right. Left. Right. Left. Right. Left.

Males. | (5

) 51-2 ) 6) 37-3 | (6) 3

Females | (9) 48-3 |(10) 3

(

(10) 34-7 |(10)

Both the talus and the calcaneus show deep grooves, which, when

these bones are articulated, bound a well-marked sinus tarsi. In

one case the sinus is shut off laterally by the two articular facets coming

in contact.

The Clavicle.

The clavicles are specially delicate bones, and the mean length of

twenty-five of them is 133-5. The results for the two sides and the

two sexes are tabulated below.

| Right. | Left. | Males. Females.

Maximum length | (9) 134-8 | (9) 137-1 a2 141-0 | (10) 125-2

} |

| |

The measurements of the individual bones are given in Table IX.

The Scapula.

The scapulae are in a fragmentary condition.

The Humerus.

The measurements of the individual bones are given in Table VIII,

and the average results for the two sides and the two sexes are tabu-

lated below.

| Right. Left. Males. Females.

Maximum length | (15) 282-9 | (15) 278-8 | (12) 285-8 | (16) 271-5

The Osteology of a Bushman Tribe. 51

TaBLe VIII.—Humerus.

| Maximum Maximum Minimum Indes

Number Length. Diameter. Diameter.

of Sex.

Skeleton. |

2 Q | 268-0 | 263-0 | 16-0 | 15-0 | 12-5 | 12-0 | 78-1 | 80-0

3 OF 303: 0) 299-0) 8-09 S05 a0 22a 2225 OO

7 2 | 303-0 | 299-5 | 20-0 | 19-0 | 13-5 | 14-0 | 67-5 | 73-7

8 OM e203. 00 209: Only -O. 9 le - On Mic OM 15-0" OOO 83:3

9 3 22-0) \20:0))) 18-0 E7-O) PS 1-8N785-0

10 Q Ss ae 19-0 | 20-5 | 14:0 | 15-5 | 73-7 | 75-6

Tha 3 | 283-5 | 283-5 | 20-0 | 21-0 | 14-0 | 14-0} 70-0 | 66-7

13 2 am 2s 6 ae) es 17-0 se OS

15 S08 Oni 293-00" 205) 20-08) 16-0) 1d-o) aS: O tio

16 OP 2 OF 266-0) 9-5) 19-0) lS oils on OA On ROT

18 Q a 247-5 asf 19-0 be 14-0 S65 |) Cave

1g) Q | 256-0 | 250-0 | 17-5 | 18-5 | 13-0 | 13-0 | 74-3 | 70:3

26 3 | 284-5 | 283-0 | 21-0 | 18-5 | 17-0 | 15-5 | 81-0 | 83-8

27 @ | BlOw) | SOC || 2a | A804) MYO) ee lO) |) Wee)

33 6 | 285-0) 282-0 721-5) 21-07) 17-0 | 17-5 | 79-1 | 83-3

40 CHPZMOsOnle2 MOrOn tel Oe hoe on aloo ilo ONM3:05 | 0-9

44 Q ae i Be, a Ne “ie ae

46 @ | 249-0 ae ISRO AGOGO) Pea ries Clk

47 7 259-0) |) 264-5: |) 16-0) 16:0) | 13:0" 13-5 | 8l-3 | 64-4

48 Q | 280-0 | 270-5 | 20-5 | 19-0) 14:0 | 14:0} 68-3 | 73-7

51 % | 253-0 see 18-5 ae 16-5 a 89-2 Me

53 3 | 276:°5 | 275-5 | 20:0 | 18:0 | 16:5 | 15:0 | 77-5 | 83-3

il Q 263:5 | 18:5 | 18-0 | 13-0 | 13-0 | 70:3 | 72-2

12 2 oe se ut a a BS Se a

25 O28: On e2o-on te lo- Os sO ne5- OF | NG-OF 8:9 183-9

3) 3 | 311-0 “3 22 2 O ON LS. | UEOR Sl Za t3

24 Q | 269-5 | 262-5 | 17-0 | 16-5 | 14:0 | 14:0] 82-4 | 84:8

30 Caos tou oon Onlm22°08 202) 2020) 18-0.) 9O0-9i 67-8

The average length of thirty-three bones is 278-1 mm.

They are

thus much shorter than those measured by Vermooten, whose results

were as follows :—

Strandlopers.

Inland Bushmen.

Hottentots.

(7) 289 mm.

(33) 288 mm.

(7) 308 mm.

520 Annals of the South African Museum.

The transverse and antero-posterior diameters of the shaft of the

humerus are also recorded in Table VIII, and the relation between

these two diameters is expressed as the “ index of robustness.” The

value of this index is 79-4 on the right side and 78-3 on the left side,

which may be compared with the index of eighty-three for the more

robust European humerus.

The lower half of the humerus has a very definite forward concavity.

A definite supratrochlear foramen occurred in fifteen of the fifty-two

humeri examined with regard to this feature. This incidence of 28-8

per cent. is less than the incidence of 52-8 per cent. found by Vermooten

in seventy-two Bushman humeri. The number of foramina depends,

however, to a certain extent on the degree of weathering to which

the bones have been subjected. There was no trace of a supracondylar

foramen.

The Radius.

The radius 1s short and slender and well curved. The measurements

of the individual bones are given in Table LX, and a comparison between

the bones of the two sides and of the two sexes is submitted below.

Maximum length

(7) 223-5

| Righ ay Left. Males. Females.

| |

| | (7) 219-6 | (12) 223-7 | (12) 202-8

The average length of twenty-six bones, irrespective of side or sex,

is213-8mm. This figure is again much smaller than the corresponding

figures arrived at by Vermooten, which were as follows :—

| |

| Strandlopers. | Inland Bushmen. | Hottentots.

| | |

| (6) 224 mm. | (35) 222 mm. | (6) 242 mm.

| |

The Ulna.

The maximum length of the individual bones is given in Table IX,

and a comparison between the averages of the bones from the two

sides and the two sexes is given below.

| ———_|- Right. Left. Males. Females.

aT ELSE ie ea |

Maximum fengthy ) 239-3 ) 236-5 (6) 246-2 (6) 221-5

The Osteology of a Bushman Tribe. 53

TABLE IX.

Clavicle. Radius. Ulna. Fibula. |

Num- Maximum Maximum Maximum Maximum

ber of Length. Length. Length. Length.

Sex g

Skele- ;

ton

2 | 2 | 125-0 | 123-0 | 205-0 He) |) PALER) |) PAletos) se | Bele)"

By | 2 ASR) 56 | 245) a lezoo 0 a a See a|

7 | 2 | 133-0 | 137-5 | 241-0 5 AE Lai gy Nea

Ss [2 ae Dee ORO Ny, Oars a seein Wu aes

Da LO, .. | 224-5 | 2217-5 ue .. | 339-0 | 340-0 |

ORR Onl 32-0 sis “3 ue oS ae So OHO)

ll | 3 | 136-5 | 136-5 ne a Bs .. | 331°0 | 329-0

13 | @ - Ba) Alle) a i oe ne

15 | ? | 136-0 | 135-5 | 232-0 ae

ies} ay ay a oO)

LO ee oe wor) | tio Ase Ae a4 Ss Ns

26 | 3 | 132-0 | 132-0 | 228-0 | 221-5 qe 23020 .. | 320-0

27 | g | 136-0 | 142-0 | 247-0 | 245-0 | 273-0 | 269-5 oe alee)

33 | go | 165-0 | 167-0 | 223-0 | 218-0 | 239-0 ad te a

40 | 3 126-5 | 217-0 | 218-0 | 228-5 .. | 326-0 | 324-0

44 | 9 tes Tee. 0) ae sys see | CULO) oe

46 | @ | 116-5 | 116-0 ba) oi alka) ae ifs a3 P2080

47 | 2 | 115-5 “i So) Wists) 205-08 291-0") 291-0

48 | 2 | 125-0 .. | 211-0 | 206-0 | 228-0 | 225-5 aie ue

lk se ea 4) 4500) || re) aN a ale

53 | 5 | 133-5 | 144-0 | 214-0 | 201-0 | 231-0

IA i ee Be a a 5a) |) ASKS a5)

ZO OR 20: oa 26-00N one si ai Se) | Cay

Oe 139-0 | 245-0] .. ma) CLOUD Fatorles)

30 | 3 D815

Aare MIT Oeos Onlet | Wo46-0) (ale 379-5

The mean length of twelve ulnae irrespective of side or sex 1s 233-9

mm., which is again shorter than the corresponding results ascertained

by Vermooten.

Strandlopers. Inland Bushmen. Hottentots. |

(7) 237 mm. (32) 237 mm. (5) 260 mm.

Annals of the South African Museum.

Proportions of the Limbs and Limb Segments.

In the following table the results for the intermembral, radio-

humeral, tibio-femoral, and humero-femoral indices are given for

the Bushman tribe, together with other data from Shrubsall (2),

and Vermooten (3) for comparison.

Inter- Radio- Tibio- Humero-

membral | humeral femoral femoral

Index. Index. Index. Index.

Bushmen (Vermooten) (22) 66-7 | (30) 77-6 | (83) 83-8 | (32) 68-2

ae (Shrubsall) 68-1 78:3 84-1 70:0

Bushman tribe 67°4 (ell 82:5 70-0

Hottentots (Vermooten) . | (4) 68-2) (5) 79-3] (6) 87-2} (5) 72-4

ie (Shrubsall) 68-8 ar 79°3 72-0

Kaffirs (Shrubsall) . 69-4 78:7 81-5 70-6

Europeans (Shrubsall) 69:7 | 72°3 80:8 72-5

The Sacrum.

The maximum height, maximum breadth, the curved length, and

the maximum height of the sacral curve are recorded in Table X.

The average sacral index in five females is 103-4, and the corresponding

TABLE X.—Sacrum.

index in three males is 101°8.

| | |

Number | | | Teeth Maximum Seyi Sacral

of Sex. | Length. | Breadth. | Gi, Height of| 7, dex Curve

Skeleton. | ‘urved. | Curve. ndex. Index.

3 Q 89-0 99-0 100-0 22-5 111-2 89-0

ih Q 108-0 112-0 117-0 21:0 103-7 92:3

10 Q 100-0 94:0 102-0 11-0 94-0 98-0

13 je) 99-0 98-0 107-0 16-5 99-0 92:5

18 Q 87-0 95-0 97-0 17:5 109-2 89-7

PAE 3 85:0 93-5 95-0 20-5 107-5 89-5

40 3 97-0 91-0 102-0 13-0 93:8 95-1

53 6 88-5 92-0 95-0 18-0 104-0 93-2

25 Q | 96-0 4-5 102-0 18-0 98-4 94-1

The sacral curvature index, which

expresses the relation of the breadth to the curved length, was also

determined. This index is 92:6 for three male sacra, and 92:3 for

five females.

The Osteology of a Bushman Tribe. 55

A striking feature of these sacra was the extent to which abnormal-

ities occurred in them. Thus in four sacra the first coccygeal segment

was incorporated in the sacrum, whilst in an equal number of instances

there were only four pieces composing the sacrum. In no case was

ossification quite complete, and gaps could therefore be seen between

the bodies of the sacral segments. In one case the first piece of the

sacrum was free, and articulated with the rest of the sacrum at the

body and on the lateral masses. A male sacrum showed a unilateral

alar articulation with the last lumbar vertebra. In another case

the sacro-iliac joint was synostosed. These marked variations are

probably associated with the low morphological position of the Bush-

man, and with a high degree of flexibility of the vertebral column.

Highteen of the twenty-five sacra showed a distinct “ sacral notch,”

that is to say, a very definite narrowing of the width of the bone in

the region of the second segment. This is a simian feature, which

is rarely met with in Europeans. The posterior third of the ala is

slightly raised, giving a distinctness to the transverse process element,

which is not found in other races asa rule. It is asif the first sacral

segment had not altogether undergone its transformation from an

isolated vertebra into a sacral element.

The Innominate Bones and the Pelvis.

Although relatively few measurements could be obtained from the

innominate bones, yet they were useful in determining the sex of the

individuals. The average maximum length of the innominate bones

of nine males is 186-8 mm., and of the corresponding bones of five

females is 164-3.

The average pelvic index is 85-5.

The individual measurements of the bones are given in Table XI,

and the mean dimensions of the pelvic brim and the pelvic brim

indices in the two sexes are set out below.

Antero-posterior Transverse :

eee Diameter. Rslees

Males. (5) 86-6 (5) 97-2 | 89-2

Females. (2) 104-8 (2) 105-2 99-7

Comparison between the length of the long bones of the two sides

of the body :—

56 Annals of the South African. Museum.

TABLE XI.

Innominate Bone. Pelvis.

Ee M | wt 4 | S

Oye aximum | Maximum = ais supr. &

=e SBE Height. | Breadth. a= aS | Aperture. ax

ss | gs | Oe | we)

Pe : | HS | Se | =

| | c= aS 26

Right.| Left. “Right. Left. | a | aaS. | Benth ‘Breadth, x

| | | |

2 | Q | 162-0 | 161-0 | edie det hele Gee ee Ry

3 | 2 | 180-0 | 229-0 | 78-6 | 1080 | 110-5 | 97-7

|| & 120) ee |194:0| 99:0] .. | ©

10 | © | 168-0 | 166-0 | 187-0 | 89:3 | 1085 | 102-0 | 106-4

11_| g | 174-0 | 180-6 | 199-0} 889 | 840 | 880 | 95-5

26 | go | 176-5 | 176-5 | | 2080) 84:9 | 78:0 | 93-5 | 83-4

27 | 3 | 198-0 | 198-0 | 137-5 | 137-5 | 234-0] .. | 89-0 | 101-0 | 881

40 | S | 184-0 | 184-0 .. | 214-0) 86-0 | 880 | 97-0 | -90-7

47 | Q | 165-0|1640 .. .. | 207-0] 79-5 | 98-0 | 103-0 | 95-1

1 ONE EO | Re 55 [GEO se 86-5 | 93-5 | 92-5

25 | 9 171-0 | 182-0 | 130-0 | 189-0; .. | 106-5 | 91-0 | 117-0

5 | g | 185-0 94-0 | 106-5 | 88-2)

30 | & | 208-0 be Sibel ie oc)! all aera

DAN Ob nae: | 131-0 | 131-0

As regards the femur there is no difference between the average

lengths of the two sides, which is in agreement with Vermooten’s

findings. In this group the right exceeded the left in 62 per cent., as

compared with Vermooten’s results, which showed that 47 per cent.

were longer on the right side in thirty pairs of Bushman femora.

There is no difference in this group between the lengths of the right

and left tibia. The average length of twenty-nine right tibiae,

measured by Vermooten, exceeded the length of the left by 1 mm.

Right exceeded length of left in 43 per cent. of instances, and

according to Vermooten in 52 per cent.

Vermooten found no difference in the lengths of the right and left

fibulae of nineteen Bushmen, but in this group there is an excess of

the right over the left of 8 mm. The right exceeded the left in 50

per cent. of instances, and according to Vermooten’s results in 32

per cent.

In this group and in Vermooten’s group of twenty-six paired humeri

the right exceeded the left in length by 4 mm. The right exceeded

the left in 86 per cent. and 81 per cent. of instances respectively.

The right radius exceeded the left in this group by 4 mm., and in

The Osteology of a Bushman Tribe. 57

Vermooten’s group of twenty-seven pairs of radii by 2 mm. The

right radius is longer than the left in 71 per cent. of instances in this

group, and in 63 per cent. of instances in Vermooten’s group.

The right ulna exceeds the left by 3 mm. in this group, and by 2

mm. in Vermooten’s group of twenty-six pairs of ulnae. The right

exceeded the left in 100 per cent. and 73 per cent. of instances

respectively.

The average length of the left clavicle exceeded that of the right

by 3 mm., and in 33 per cent. of instances the length of the right clavicle

exceeded that of the left.

On the whole, therefore, where there is any difference between the

two sides, the right bone exceeds the left, with the exception of the

clavicle. The bones of the right and left legs are approximately equal

in length, but the bones of the right arm, except the clavicle, are

decidedly longer than those of the left arm.

STATURE.

The stature has been estimated by Manouvrier’s tables from the

average lengths of the various long bones, and the following results

were arrived at :— :

Males. Females.

Femur . : ; (14) 1577 (24) 1480

inbiage = : (16) 1582 (24) 1506

Humerus (12) 1481 (17) 1446

Radius (12) 1589 (12) 1469

Average stature. 1557 1475

The stature was also determined by the use of Pearson’s formulae (8).

Fe-

Males. iales

Femur . | (S=81-306-+1-880 F) | 1589 | (S=72-844+1-945 F) | 1492

Tibia . .| (S=78-664+2-376T) | 1588 | (S=74-774+2-352 T) | 1503

Humerus - | (S=70-641+2-894 H)) 1584 | (S=71-475-+2-754 H)| 1467

Radius 5 9. (S=85:925+3-271 R) | 1591 (S=81-2

2443-343 R)| 1493 |

| Average stature oe 1575 ols 1489

58 Annals of the South African Museum.

Pearson has shown that in the dwarf races the above methods tend

to overestimate the stature. The stature was therefore calculated

according to his method for dwarf races.

Males. Females.

Femur . (14) 1525 (14) 1447

Tibia (16) 1525 (16) 1447

Humerus (12) 1448 (17) 13899

| Radius (12) 1525 (12) 1459

Average stature 1506 1438

It is difficult to say, with our present knowledge, which of the above

figures is the best approximation. In the meantime the results may

be compared with Shrubsall’s findings.

| |

Strandlopers. | ae Bee eee | Hottentots.

Males 1544 1565 1561 1611

Females 1512 1422 1431 1491

SKELETONS OF THE EXTRANEOUS INDIVIDUALS.

The skeletons of only two of the non-Bushman individuals were

available. The average dimensions and indices of the long bones of

these, No. 5 and No. 30, are tabulated below.

Femur. (Oblique length) 464-3 mm.

Tibia. (Standard length) 382-0 mm.

Fibula. (Maximum length) 370-9 mm.

Humerus. (Maximum length) 327-5 mm.

Radius. (Maximum length) 249-5 mm.

Ulna. (Maximum length) . 249-0 mm.

Clavicle. (Maximum length) 143-5 mm.

Superior platymeric index . 98-3

Pilastering index 127-0

Platycnemic index 70-7

Intermembral index 80-0

Radio-humeral index . 17-2

Tibio-femoral index 82:3

Humero-femoral index 69-8

The Osteology of a Bushman Tribe. 59

SUMMARY AND CONCLUSIONS.

(1) This work is an anthropological survey of the skeletons of a

tribe of aborigines exhumed at Colesberg by the South African

Museum, Capetown.

(2) Fifty-three skulls were available for study, and these were first

of all separated into a group of forty-five relatively pure Bushman

types, and into a remainder of definitely mixed types. A number

of these skulls were fragmentary and deformed, so that the Bushman

types which were suitable for study numbered twenty-nine.

(8) The limb bones and the sacrum were also studied, and details

of the measurements are given in separate tables. The average

results for the two sides of the body and for the two sexes are tabu-

lated for each bone.

(4) Indices expressing the proportions of the limbs and limb

segments to each other were also determined, and compared with

appropriate figures from other workers.

(5) The stature was estimated by various formulae.

(6) The conclusion is arrived at that the larger group is composed

of relatively pure Bushmen, more typically Bushman, especially as

regards the smallness of their bones, than many other groups which

have been described as Bushmen. Nevertheless there appears to

be. a strain of what might be termed Hottentot blood in them, as

evidenced by the narrowing of the skull, and the apparent subnasal

prognathism.

(7) The smaller group presents a mixture of types, two being

Hottentot in type, two are quite negroid, one may have European

blood, whilst the others are fragmentary.

(8) The measurements used are those which have been sanctioned

by international agreement.

In conclusion, I wish to express my thanks to Professor M. R.

Drennan of the Anatomy Department, University of Capetown, under

whose supervision this investigation was carried out, and to Mr. A. A.

Lamb for the photographs used in the Plates.

60 Annals of the South African Museum.

REFERENCES.

(1) Stome, D.—The Curvature of the Bushman Calvarium. Amer. Journ. Phys.

Anthrop., vol. x, No. 3, July-Sept.

(2) SHruBSALL, F. C.—Notes on Some Bushman Crania and Bones from the South

African Museum, Capetown. Annals of the 8. A. Museum, vol. v, part 5, 1907.

(3) Martixy, Rupotr.—Zur Anthropologie der Buschmanner. Dietrich Reimer

(Ernst Vohsen), Berlin, 1926.

(4) Broom, R.—A Contribution to the Craniology of the Yellow-skinned Races of

South Africa. Proc. Roy. Anthrop. Inst., vol. liii, 1923.

(5) Drennan, M. R.—A Short Course on Physical Anthropology.

(6) Drunnan, M. R.—The Dentition of a Bushman Tribe. Annals of the 8. A.

Museum, vol. xxiv, part 1, p. 61, 1929.

(7) Hrpuiéka, ALes.—Physical Anthropology of the Lenape or Delawares, and of

the Eastern Indiansin general. Smithsonian Institution, Bureau of American

Ethnology Bulletin, 62.

(8) Pearson, K.—On the Reconstruction of the Stature of Prehistoric Races.

Phil. Trans. Roy. Soc., London, 1899.

TITLES OF PLATES VIII-XI.

Prats VIII.

1. Female skull (No. 44). 2. Female skull (No. 19).

3. Male skull (No. 53). 4. Male skull (No. 40).

PLATE TX.

1. Female skull (No. 44). 2. Female skull (No. 19).

3. Male skull (No. 53). 4, Male skull (No. 40).

PLATE X.

1. Female skull (No. 44). 2. Female skull (No. 19).

3. Male skull (No. 53). 4. Male skull (No. 40).

PLATE XI.

Limb-bones from the Bushman tribe compared with corresponding

bones from Europeans.

Ann. 8. Afr. Mus., Vol. XXIV, Plate VIII,

1, Female skull (No. 44); 2, female skull (No, 19); 3, male skull (No. 53); 4, male skull (No. 40).

Neill &: Co., Lid.

1a. oh

Ann. 8. Afr. Mus., Vol. XXIV. Plate LX.

1, Female skull (No. 44); 2, female skull (No. 19); 3, male skull (No. 53); 4, male skull (No. 40).

Neili & Co., Lid.

Ann. S. Afr. Mus., Vol. XXIV. Plate X.

1, Female skull (No. 44); 2, female skull (No. 19); 3, male skull (No. 53); 4, male skull (No. 40).

Neili & Co., Lid.

Ann. S. Afr. Mus., Vol. XXIV. Plate XI.

Shows limb-bones from the Bushman tribe, selected as being of about average length. and compared

with corresponding bones from Europeans also of approximately average length.

Newl & Co., Lid.

4. The Dentition of a Bushman Tribe.—By Professor M. R. DRENNAN,

Anatomy Department, University of Capetown.

THRouGH the courtesy of Dr. Leonard Gill, Director of the South

African Museum, Capetown, I have been permitted to study the

skeletal remains of a tribe of South African Bushmen, which have

recently been acquired by that Institution. The general results of

this research on the skulls and other bones are being incorporated in

another paper, but it seemed to me to be advisable to report the

special results of the study of the teeth of this group of individuals

in a separate paper, where they could be available to those who are

specially interested in these organs.

The work follows the lines adopted by T. D. Campbell, D.D.Sc.,

in his monograph on “‘The Dentition and Palate of the Australian

Aboriginal,” although, owing to the small amount of material avail-

able, it has not been considered advisable to aim at approaching the

comprehensiveness of this excellent monograph (1). I wish to take

this opportunity of thanking Dr. Gill for the material, and Dr. Campbell

for the plan of the work and for his comparative data.

As the work proceeded it became clear to me that it would help to

elucidate the results if data from the uncivilised Bushman and from

the Bantu tribes of South Africa could be used also for comparison.

Unfortunately I could not find any reference to any special work on

the teeth of the Bushman or of the South African Bantu groups, and

I doubt if anything at all has been done onthem. I have attempted,

therefore, to remedy this defect to a certain extent by making a pre-

liminary examination of the dentition of fifty Bushman skulls in the

Anatomy Department of the University of Capetown. Unfortunately,

however, many of the teeth of these skulls have dropped out, and many

of the skulls are fragmentary and without mandibles. I have also

examined ten fairly typical Kaffir dentitions, and also four baboons

and two gorillas, in order to get approximate figures for comparison.

The main work, however, has been directed towards making as

thorough an examination as possible of the dentition of the Bushman

tribe.

62 Annals of the South African Museum.

SOURCE OF THE MATERIAL.

The skeletons which have been studied were exhumed in November

1926, from graves situated near the town of Colesberg, by Mr. J.

Drury, modeller to the South African Museum, who is an experienced

excavator. Mr. Drury’s account of the circumstances of the burial

and exhumation is quoted by Mr. Slome in the preceding paper.

PRELIMINARY DISCUSSION.

The first poimt which has to be decided is how far one is justified

in calling this group of individuals a Bushman tribe. So far no history

of the group has been ascertained, but this is not any drawback in

connection with the settlement of the question of the race to which

they belonged. The only reliable criterion of race is physical features,

and all the observations and measurements which have been made so

far point to the conclusion that the great majority of these individuals

were relatively pure representatives of the Bushman race. It is

generally conceded nowadays that there is no such thing as a pure

race, and it is probable that many of the variations from type, which

occur in connection with every character under consideration in this

and in other groups of skulls, are an expression of a mixed origin in it

may be a remote past.

This does not account for all the variations in the present group.

Several of the skulls are practically pure Bantu in type, and there are

others which are so far removed from the Bushman type and in the

Bantu direction that it has been considered advisable to remove them

from the tables of measurements. The numbers of the skulls which’

have been excluded are 5, 20 (a), 21, 24, 28, 30, 34, 37, and 45. Skull

No. 29 does show traces of European blood, especially in its nasal

aperture, but it has many other Bushman characters, and its teeth

present the usual Bushman dimensions, so that it has been retained

in the calculations.

By this process of selection it is hoped that the figures ascertained

will be of more value in connection with future studies on the Bushman

teeth. The results for the teeth corroborate the view ascertained

otherwise that this was a Bushman tribe. The figures differ only

slightly from those derived from the small collection of Bushmen in

the Anatomy Department. In this connection, however, it must be

noted that this tribe had obviously been living under at least semi-

civilised conditions, so that their teeth are therefore not reduced in

The Dentition of a Bushman Tribe. 63

size by attrition, and, on the other hand, the measurements on them are

far more numerous than those on the uncivilised Bushmen, so that

they may therefore be a more accurate representation of the Bushman

dentition. In the investigation of the physiological and pathological

conditions, such as the degree of attrition and the incidence of caries,

the tribe was treated as a unit, and all the skulls were included in the

survey of the dentition.

No attempt has been made in dealing with the dentition to dis-

tinguish between the Bushman and Hottentot types in the present

group. No agreement has been arrived at as to just what are the

critical physical differences (if any) between the Bushman and

Hottentot tribes. This difficult question has become complicated by

the frequent use of the term Hottentot in reference to almost any kind

of crossing amongst the native races. There is evidence that there

was a constant intermingling of the Hottentots and Bushmen in the

early days of the European occupation. It is going to be difficult,

therefore, to get sufficiently pure representatives of either type from

which to dogmatise as regards criteria of difference. In an important

paper published recently by the late Professor Rudolf Martin, entitled

“Zur Anthropologie der Buschmanner”’ (2), the literature on this

_vexed problem is exhaustively quoted, and the opinions of the various

authorities on the subject are summarised. Shrubsall (3), who de-

voted considerable attention to this subject, concluded that ‘‘The

Hottentots are intermediate in character between the Strandlopers

or primitive Bushmen and the Bantu tribes. It appears that if the

Hottentots are a cross between these two races, the intermingling

must have taken place a long time ago, as the Hottentots’ skulls form

a very homogeneous series.’ This homogeneity, however, might also

simply mean that cross-breeds breed true, without necessitating the

intermingling to have taken place a long time ago. At any rate, if

approaching the Bantu in a slight degree is a criterion of being a

Hottentot, some of this tribe may well be regarded as Hottentots.

In a recent paper dealing with this question by Dr. R. Broom,

F.R.S., on “The Yellow-Skinned Races of South Africa” (4), he comes

to the conclusion that it is relatively easy to distinguish between a

Bushman and a Hottentot. He states that “One of the most striking

differences in the cranial measurements of the Bushmen from those

of the Hottentots and Bantus is that the height is invariably less

than the breadth (in the Bushman), and generally much less.”

According to this criterion 75 per cent. of this tribe were pure Bushmen,

as in this percentage of the skulls the height is less than the breadth,

64 Annals of the South African Museum.

and in another 4 per cent. the height was just equal to the breadth. It

is not possible, however, to define a race by one single character.

The sexing of Bushman skulls with any degree of accuracy is a matter

of very great difficulty, so that following the precedent of Dr. Campbell,

who found an equal difficulty in sexing the Australian aboriginal,

no attempt has been made to give results for separate sexes.

Nor has any effort been made to classify the skulls according to

age, except that the average measurements of the permanent teeth

of all young skulls, that is to say, of all whose third molars had

not completely erupted, was calculated separately. The difference

between the size of these larger young teeth and that of the teeth as

a whole is regarded later on as being an indication of the degree of

interproximal attrition, which takes place as the individuals grow

older.

MEASUREMENTS ADOPTED.

Following on the same lines as Dr. Campbell, the maximal mesio-

distal and the maximum labio-lingual diameter of the crown of each

tooth was measured with fine calipers. The root length and total

length of those teeth which were loose in their sockets were also

measured, and the material was specially favourable from this point

of view, so that the results for these quantities are specially valuable.

In the case of the incisors, canines, and premolars there was little

dubiety as to how these root and total measurements should be taken,

the root length being measured from the enamel edge of the crown

at the cervix on the labial aspect to the apex of the root, and the

total length being the maximum one. The crown length was cal-

culated by subtraction. In the case of the molar teeth, where the

roots varied in length and divergence, not knowing of any standard

procedure, I took the average length of the roots. The total length of

the molars was measured by placing the occlusal surface of the crowns

flat on the blade of the calipers and sliding the movable blade along

to touch the longest root. This gives a projected or physiological total

length, and the crown length calculated by subtraction can therefore

only be approximate for the molars. (See table at end.)

All the measurements were read to the nearest quarter of a milli-

metre, and the calculations are taken to the nearest decimal point in

the tables. The terms mesial, distal, labial, and lingual are expressed

in the tables by the capital letters M., D., L., and L.

The averages of the various measurements of the individual teeth

are set out in tabular form. The figure in brackets refers in every

The Dentition of a Bushman Tribe. 65

instance to the number of observations made. In addition to the

six sets of measurements which I have made on the Bushman tribe,

the South African Bushman, the Bantu types in the tribe, the Kaffir,

the gorilla, and the baboon (Papio porcarius), I have availed myself

of the data collected by Campbell for the European, Japanese, and

Australian aboriginal for purposes of comparison. Valuable com-

parative data from Australian and prehistoric molars can be found

in a paper by Nicholls (9).

DISCUSSION OF THE MEASUREMENTS.

An analysis of the figures obtained for the various dimensions of

the teeth of these different types compared, as set out in tabular form

for the individual teeth, presents many interesting problems for dis-

cussion. Owing, however, to the lack of suitable standards for com-

parison, especially in connection with the South African native races,

it has not been considered advisable to submit a discussion on each

individual type of tooth. For the present preliminary survey the

figures must speak for themselves, and they can be used at a later

date, if necessary, for further discussion.

As regards the general size of the teeth, the gorilla has easily the

largest teeth, being followed at a considerable distance by the baboon.

In the human series the Australian aboriginal comes next, having

wonderfully large teeth, exceeding even the baboon in thirteen out

of the thirty-two measurements. Next in order of size stand the

Japanese and Karffirs, the teeth of these two races being on the whole

about equal in size. It is an interesting fact that, of the sixteen

measurements taken on the upper teeth, six are greater in the Japanese

than in the Kaffir, whereas, of the sixteen measurements on the lower

teeth, ten are greater in the Japanese than in the Kaffir. The Bantu

types of the Bushman tribe have teeth which differ very little from

the EKuropean’s teeth as regards size, and these two groups therefore

stand next in order of size. Next in size are the teeth of the Bushman

tribe, which are just a little larger than those of the uncivilised South

African Bushman. The teeth of these latter are the smallest in the

series, but a certain amount of this inferiority in size of the teeth of

the uncivilised Bushman may be due to the greater degree of attrition

in the uncivilised than in the civilised Bushman. Although the general

order as regards size is as stated, it is found that as regards the

individual teeth there are many exceptions to this sequence.

The question arises as to how far the size of the teeth can be taken

VOL. XXIV, PART 1. a

Annals of the South African Museum.

Average Measurements of the Teeth in the Upper Jaw.

Upper Central Incisor.

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffirs .

Europeans

Japanese

Australians

Gorillas

Baboons

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffrs .

Europeans

Japanese

Australians

Gorillas

Baboons

S.A. Bushmen

Bushman tribe

Bantus in tribe |

Kaffirs .

Europeans

Japanese

Australians

Gorillas

Baboons

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kafirs .

Europeans

Japanese

Australians

Gorillas

Baboons

Root

M.D L.L. Index. length.

VvVr oo O——— —M

(4) 7-8 (4) 6-4] 82-5 se

(25) 8-3] (27) 65] 78-4] (19) 12-5

(3) 9-1 (3) FO 7-1 (3) et3-5

(5) 8-7 (5) 7:0} 80-4 as

9-0 Us Onadides: 12-0

8-4 74) 88-1 11-3

(56) 9-4) (93) 7-9} 84-0 | (18) 14-6

(2) 13-8 (2) 10-3| 74-6 8s

(4) 10-1 (4) 9-4 |. 93-2

Upper Lateral Incisor.

(4) 6-9 (4) 61) 89-1 oie

(24) 6-7} (26) 6-0] 89-1 | (18) 12-5

(4) 6-8 (4) 6:0} 88-1] (2) 13-5

(2) eal (7) 6-6} 92-4 3

6-4 6-0 | 93:8 13-0

7:0 6-5 | 92-9 IISef

(78) 7-7 | (126) 6-9| 89-6 | (15) 15-6 |

(2) 9-0 (2) 9-5 | 105-6 oh

(4) 7-8 (4) 8-6} 109-6

Upper Canine.

(S)P1G:9) | (8) 7 LOmeT is

(26) 7-5] (26) 7-8} 104-1 | (20) 16-3

(5) 7-6 (5) 8-1) 107-3} (5) 16-1

(Ae ON 070) teas |) Ish By:

7-6 8-0 | 105-3 17-3

7-9 8-6 | 108-9 15-4

(116) 8-4 | (159) 9-0 | 107-1 | (41) 19-9

(2) 16-3 (2) 13-0 | 80-0 |

(4) 16-1 (4) 11-1] 68-9

Upper First Premolar.

(8) 66] (8) 8-7 | 1382-5 Ss

(27) 6-8] (25) 8-6 | 126-9 | (17) 14:5

(5) 6-9 (5) 9-1) 131-2} (3) 14-9

(S) yall (8) 9-5 | 134-4 oie

7-2 9-1 | 126-4 12-4

7:3 9-5 | 130-1 12-2

(124), 7-8 (163) 10-3 | 132-1 | (35) 16-3

(2) 11-0 (2) 16-0 | 145-5 ale

(4) 7:8 (4) 7-4] 95:3 |

Total

length.

(19) 21:8

(3) 23:3

99-5

22-9

(4) 24-5

9-2

11-0

7:6

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffirs .

Europeans

Japanese

Australians

Gorillas

Baboons

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffirs .

Europeans

Japanese

Australians

Gorillas

Baboons

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffirs .

Europeans

Japanese

Australians

Gorillas

Baboons

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffirs .

Europeans

Japanese

Australians

Gorillas

The Dentition of a Bushman Tribe. 67

Average Measurements of the Teeth in the Upper Jaw—contd.

Upper Second Premolar.

3 Root Total |Crown |

MolD le BEORRS length. length. | depth. |

ese —_—_—_ 8 ——

(8) 6:3 (8) 8-7 | 138-2 me ie a

(28) 65} (26) 8-5 | 131-1 | (15) 14:8 | (15) 21-1} 6-3

(5) 6-7 (5) 9-1 | 1386-8) (2) 14-6] (2) 21-1} 6-5

(8) 6-7 (8) 9-8 | 146-4 | ae bs ae

6:8 8:8 | 129-4 14-0 21-5 | 7-5

7-0 9-3 | 132-9 12-9 20:7| 7:8

(89) 7-2 | (168) 10-1 | 140-3 | (36) 16-9 | (13) 22:3| 5-4

(2) 10-5 (2) 15-0 | 142-9 alt fs

(4) 7-9! (4) 9111141

Upper First Molar.

(18) 9-6 | (18) 10-7 | 111-9 i ei ae

(33) 9:9 | (33) 10-6 | 107-2 | (15) 11-6 | (15) 17-6 | 6-0

(7) 10-2 (7) 11-0 | 107-7 | (5) 12:5 | (5) 18-9] 6:4

(8) 10-1 (8) 11-4 | 113-1 i, (3 one

10-7 11-8 | 110-3 13-2 20:8 | 7-6

10-7 11-3 | 105-6 13-2 20-2 | 7-0

(198) 11-4 | (255) 12-8 | 112-3 | (5) 13-5 | (1)19:5| 6-0

(2) 14-1 (2) 15-1 | 107-1 a ae

(4) 11-0 (4) 9-9 | 89-8

Upper Second Molar.

(18) 9-6] (18) 10-8 | 111-6 : a ae

(30) 9-7 | (30) 10-6 | 110-2 | (14) 12-6 | (14) 18-4| 5-8

(7) 10-0 (z/))s Unk s) a) USS) PCD AKO XO) | (Gil) Tess G83

(9) 10-2 (9) 11-7 | 114-6 He yy a

9-2 11-5 | 125-0 13-0 20:0) 59 27-0

9-9 TSIEN ee 13-1 20:7| 6:9

(168) 10-9 | (241) 13-1 | 120-2 | (15) 13-6 | (8) 19-6| 6-0

(2) 15-0 (2) 15-5 | 103-3 is as

(4) 13-9 (4) 11-6! 83-4

Upper Third Molar.

(13) 8-0) (13) 10-4 | 130-0 sie ie ‘

(26) 8-2| (26) 10-3 | 125-7 | (10) 12-6 | (10) 18-0] 5-4

(7) 9:0] (7) 11-0 | 122-7] (2) 126| (2)17-6| 5-0

(6) 9-5 (6) 11-5 | 120-6 a 2s 5

8-6 10-6 | 123-3 11-4 7 out

8-9 10-6 | 119-1 10-5 16- 6-4

(142) 10-0 | (193) 12-3 | 123-0 | (22) 13-9 | (10) 19:7 | 5:8

(2) 14:3 (2) 14:3 | 100-0 ce ae

(4) 15-1 (4) 11-5 | 76-0

Baboons

Annals of the South African Museum.

Average Measurements of the Teeth in the Lower Jaw.

Lower Central Incisor.

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffirs .

Europeans

Japanese

Australians

Gorillas

Baboons

1S.A. Bushmen

Bushman tribe

{Bantus in tribe

Kaffirs .

Europeans

Japanese

Australians

Gorillas

*Baboons

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffirs .

Europeans

Japanese

Australians

Gorillas

Baboons

Kaffirs .

Europeans

Japanese

Australians

Gorillas

Baboons

L.L. Index.

S.A. Bushmen |

Bushman tribe |

Bantus in tribe |

HOAROEHS

Lower Lateral Incisor.

(4) 5-6) 101-3

(23) 5-6| 99-6

(5) 6-0 | 102-6

(8) 6-0 | 106-4

6-4 | 108-5

6-3 | 105-0

(92) 66] 98-5

(2) 10-1 | 112-5

(4) 8-2] 116-0 |

Lower Canine.

(5) 7-0 | 107-8

7-0 | 103-7

7-9 | 116-2

(120) 83) 109-2 |

(2) 14-5 | 138-1 |

(4) 11-5 | 153-3 |

Lower First Premolar.

(6) 7-3 | 106-0 |

6) 7-6 | 110-0

5) 7-9 | 109-0 |

8) 81} 114-4

9 | 114-5

| 74-1

) 5-4) 104-9

) 52) 104-6

(4) 5:6 | 108-4

) 5-9 | 108-7

6-0 | 111-1

5:8 | 107-4

(77) 6-3 | 105-0

(2) 8-9 | 131-6

(4) 8-9! 107-6

)

(6) 7-4} 102-9 |

) 7-8 | 112-0 |

7-9 | 114:5 |

3 | 109-2

8 | 115-8

24-9

(26) 23-6

(5) 25-5

"95-6

24-5

(4) 28-0

8-3

9-4

10-3

10:9

9-9

The Dentition of a Bushman Tribe.

Average Measurements of the Teeth in the Lower Jaw—contd.

Lower Second Premolar.

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kafiirs .

Europeans

Japanese

Australians

Gorillas

Baboons

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffirs .

Huropeans

Japanese

Australians

Gorillas

Baboons

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffirs .

Europeans

Japanese

Australians

' Gorillas

Baboons

S.A. Bushmen

Bushman tribe

Bantus in tribe

Kaffirs .

Kuropeans

Japanese

Australians

Gorillas

Baboons

M.D L.L. Index. ;

—— —_——__

(3) 6-8 (3) 7:8 | 114-8

(30) 7:0) (30) 7-8 | 110-8 |

(5) Zell (5) 8:2] 115-6

(9) 7:2 (9) 8:3 | 116-0 |

71 8:0 | 112-7

7:3 8-2 | 112-3

(79) 7-7 | (109) 8-9 | 115-6

(2) 11-0| (2) 12-6 | 114-7 |

(4) 9:6 (4) 7:31 75:9:

Lower First Molar.

(10) 10-3 | (10) 9:9| 95-6

(33) 10-9 | (33) 10-2 | 93-7

(6) 11-0 (6) 10-3 | 93-9

(7) UW-3 (7) 10-6 | 93-6

11-2 10:3) 91-9

11:5 10:7 | 93-0

(139) 12-3 | (186) 11-9} 96-7

(2) 15:0} (2) 13-3 88-3

(4) 10-4 (4) 8:5 | 81-4 |

Lower Second Molar.

(11) 10-4) (11) 9:8, 95-0 |

(32) 10-6 | (32) 10-1| 94-8

(Zi) ikea (7) 10:3 | 93-0

(7) 11-0 (7) 10-6 | 96-4

10:7 10:1 | 94-4

11-1 10:2 | 91-9

(152) 12-5 | (184) 11-7] 93-6

(2) 16-8 (2) 14:9 | 88-8

(4) 13-5 (4) 10-0 | 74-1

Lower Third Molar.

(10) 9:2) (10) 9:2) 99-5

(33) -9-9'| (33) 9:6] 97-0

(8) 11-2| (8) 10-4] 92-5

(5) 11-4 (5) 10-7 | 93-9

10-7 9-8 | 91-6

10:5 9:8 | 93:3

(136) 11-0 | (152) 11-1 | 93-3

(2) 16-4 (2) 14:0 | 85:5

(4) 17-6 (4) 10:7 | 60-6

12-8

(6) 16.5

(27) 12-4

(4) 13-0

"13-2

12-1

(7) 14-2

(21) 12-5

(7) 13-3

“118

10-5

(6) 14-0

Total

length.

70 Annals of the South African Museum.

as a criterion of the morphological position of the types under con-

sideration. Whilst it is probably true that large teeth indicate a

simian tendency, yet mere size of teeth has to be discounted in several

directions. The gorilla has much larger teeth than the baboon, yet

the latter is undoubtedly a lower form than the former. A great deal

of the size of the gorilla’s teeth must be due to the fact that the gorilla

is a larger animal than the baboon, and the size of the teeth is neces-

sarily strongly correlated with bodily size. Just as the gorilla is

handicapped morphologically on the score of absolute size of teeth,

so the Bushman is enhanced in status by the smallness of his teeth,

although a considerable degree of this absolute smallness of the

Bushman’s teeth must be due to the correlation of his small teeth

with his small stature. Nevertheless it is quite conceivable that the

Bushman’s teeth may be both absolutely and relatively small, and

that in respect of his teeth the Bushman presents a much less simian

condition than even the European (see later under Flower’s dental

index). There are quite a number of features, notably the length of

the limbs, in which the Bushman is less simian than the so-called

higher races.

In order to get rid of the difficulty as regards absolute size, anthro-

pologists are accustomed to form an index or ratio between two

measurements. I have therefore used as an index for each tooth

the relationship between the labio-lingual and the mesio-distal dimen-

sions. The former is expressed as a percentage of the latter thus :—

Labio-lingual diameter « 100

Dental index= ae ;

Mesio-distal diameter

This gives a figure by means of which it is possible to compare

corresponding teeth of different sizes. It has the advantage in this

case of giving an idea of the shape of the tooth, which obviously has

a morphological significance as well as size in any comparison. This

index has accordingly been calculated for each type of tooth in each

of the groups. Although in itself it might not be a sufficient criterion

of morphological sequence, nevertheless taken in association with

size and other features it has considerable value. It should be noted

that owing to the different dimensions of the different types of teeth,

the index sometimes rises and sometimes falls to indicate morpho-

logical ascent. As was the case in connection with absolute size, there

is no constant morphological order followed by the individual teeth

according to thiscriterion. Sometimes an individual tooth in an other-

wise high or low series assumes an altogether anomalous position.

The Dentition of a Bushman Tribe. ia

This seems to indicate that, morphologically speaking, the various

teeth are somewhat plastic, and assume a shape which best suits

the physiological needs. These variations in shape will repay further

investigation. In the meantime it is only necessary to note that, as

determined by this index, the teeth of the Bushman are on the whole

of a high morphological type.

DEGREE OF REDUCTION IN THE BusHMAN Mo.Lars.

In the lower primates the three molars increase in mesio-distal

diameter from before backwards. In the baboon the ratio is

11: 14:15 in the upper jaw, and 10:13:17 in the lower jaw.

In the gorilla the ratio is approximately 14:15:14 in the upper

jaw, and 15:16:17 in the lower jaw, showing that reduction has

commenced in the upper wisdom tooth. In higher races there is

a diminution in size of the molars from before back in both jaws,

but the reduction is more marked in the upper jaw.

In the Australian the relative size of the molar teeth is 11-4 mm. :

10-9 mm.: 10 mm. in the upper jaw, and 12-3 mm. : 12:5 mm. : 11-9

mm. in the lower jaw. ‘That is to say, the second lower molar is still

larger than the first, and both wisdom teeth are still actively function-

ing members of the dentition. In the Bushman tribe, excluding the

Bantu types, the relative size of the molars is 9-9 mm. : 9-7 mm. : 8-2

mm. in the upper jaw, and 10-9 mm. : 10-6 mm. : 9-9 mm. in the lower

jaw. ‘That is to say, there is a reduction from before backwards as in

the higher races, and as in them it is more marked in the upper than

in the lower jaw.

There is also an absolute reduction in the breadth of the molars

from before backwards, and this reduction is greater in the lower than

in the upper jaw. Relative to the length, however, the breadth, as

indicated by the dental index, increases from before backwards, and

much more so in the upper than in the lower jaw.

On comparing the individual teeth it was found that in the upper

jaw the first molar was larger than the second in 43 per cent., equal

to it in 50 per cent., and less than it in only 7 per cent. of the thirty

dentitions compared. The first and second upper molars were both

larger than the third in 96 per cent., and smaller than it in 4 per cent.

of the twenty-five dentitions compared in this respect.

In the lower jaw the first molar was larger than the second in 51

per cent., equal to it in 26 per cent., and less than it in 23 per cent. of

the thirty-one dentitions compared. The first molar was larger than

72 Annals of the South African Museum.

the third in 93 per cent., and equal to it in 7 per cent. of the thirty

dentitions compared. The second molar was larger than the third in

90 per cent., and equal to it in 10 per cent. of the thirty instances

compared.

Morar Cusp PATTERNS.

An investigation of the cusp pattern of the teeth of the Bushman

tribe, excluding the Bantu types, was also undertaken. As regards

the number of cusps on the occlusal surface of the molar teeth, the

upper molars were classified into those having four well-developed

cusps, those having threewell-developed cusps anda cusplet or tubercle,

and those having three cusps or less. The lower molars were classified

into those having five well-developed cusps, those having four and a

cusplet, and those having four or less cusps. The results are tabu-

lated below, the number of teeth showing the different types of cusp

formation being expressed as a percentage, and the number of obser-

vations is given, as usual, in brackets. Campbell's figures for the

Australian, slightly adapted, are given for comparison.

4 cusps. | 33 cusps. | 3 cusps.

Per cent.|Per cent.|Per cent.

lst upper molar | 100-0 Bushman tribe (32).

+ m OOO .. | Australians (88).

2nd upper molar | 100-0 | .. iP Bushman tribe (31).

e - | 100-0 | as | rt Australians (208).

3rd upper molar | 12:0 | 38-0 50-0 | Bushman tribe (26).

a Me Leo AOR He) | Australians (156).

5 cusps. | 43 cusps. | 4 cusps. |

| | |

| |

|Per cent.|Per cent.) Per cent.

lst lower molar | 100-0 Bushman tribe (28).

SOM NEe | 940 | 25 | 3-7 | Australians (88).

2nd lower molar 85-0 75 | 7-5 | Bushman tribe (27).

A i 32:0 4-8 63-4 | Australians (125).

3rd lower molar | 76-0 10-0 14:0 | Bushman tribe (29).

in Z | 727 | 27-3

| Australians (110).

The results show that the first upper molar invariably shows the

four cusps characteristic of the rest of the hominidae. The second

The Dentition of a Bushman Tribe. 73

upper molar in the Bushman tribe also invariably presents four cusps,

which is in marked contrast to the fact that four cusps are only

present to the extent of 58 per cent. on this tooth in civilised races

(Duckworth, 5).

The measurements have already shown that the third upper molar

is strikingly reduced in size in this Bushman tribe and in Bushmen in

general. The results as regards cusp reduction bear this out, and are

in marked contrast to the results for the corresponding tooth in the

Australian. According to Duckworth, 37 per cent. of all races still

have four cusps on their upper wisdom tooth, so that it would appear

that this tooth is more degenerate in the Bushman than in most other

races.

Five cusps are invariably present on the first lower molar of the

Bushman tribe. This is in contrast to the presence of five in 94 per

cent. of Australians (Campbell), and in only 77 per cent. of highly

civilised races (Duckworth).

The second lower molar of the Bushman tribe has five cusps in

88 per cent. of cases, which is also in marked contrast to the 32 per

cent. occurrence in Australians and other primitive races.

Equally striking is the presence of five cusps on the lower wisdom

tooth in this tribe to the extent of 76 per cent. of instances, and

in this respect there is agreement with the Australian aboriginal.

According to Duckworth, even “four cusps are rare in the lower

wisdom, and three are found in as many as 64 per cent. of highly

civilised races, so that three cusps may thus be considered the

normal number in these races. Even in primitive races, such as the

Melanesians, three cusps occur in 28 per cent. of cases.”

One seems justified, therefore, in drawing the general conclusion

that in the upper jaw there is a very considerable and very sudden

reduction in size of the last molar and in the number of its cusps,

whereas in the mandibular teeth the reduction is much more gradual

and much less extensive both in regard to size and to the number of

cusps.

Root Form Anp SIZE.

The average dimensions of the roots of the various teeth of the

Bushman tribe, excluding the Bantu types, are set out in tabular

form together with the other dimensions of the teeth, and the root

lengths of other types are given for comparison. In the table at the

end the maximum and minimum length of the root of each type of

tooth is also given.

74 Annals of the South African Museum.

It may be noted that the roots of the upper teeth are on the whole

about the same length as those of the European, but they are very

much shorter than those of the Australian. In the lower jaw, however,

the roots of the Bushman teeth are considerably longer than those

of the European, in keeping with the better development of the

mandibular teeth in the former than in the latter, but they are

still much shorter than the corresponding roots of the Australian’s

teeth.

Anything which might be classed as abnormal divergence of the

roots was noticeably absent. The maximum divergence of the upper

roots was 14-7 mm., measured between the apices of the two roots

widest apart, which is far from being extreme, and there was no

abnormal divergence of the lower roots.

The roots of the whole tribe, including the Bantu types, were classi-

fied into those which were normal and into those which showed a

more or less pronounced degree of fusion. The percentages of teeth

belonging to each class are tabulated below.

Normal. Fused

Per cent. | Per cent.

lst upper molar roots. 5 94 6

2nd BA 3 : tl 50 50

3rd 5 . : 33 67

1st lower molar roots 100 ss

2nd 5 ms : . 65 | 39

3rd . x , : 27 | 73

FLOWER’S DENTAL INDEX.

Attempts have been made to use the size of the teeth as one of the

criteria of race, and for this purpose anthropologists frequently employ

a dental index, devised by Flower. This index expresses the ratio

or relationship between the length of the teeth measured “‘from the

most anterior point on the first upper premolar to the most posterior

point on the third molar,” and the length of the base of the skull from

the basion to the nasion. It is usually expressed as a percentage :

Premolar-molar length « 100

Dental index=

Basion-nasion length

The Dentition of a Bushman Tribe. 75

According to the value of this index skulls are classified as follows :-—

Microdont . ; . up to 41-9, e.g. Europeans.

Mesodont . . 42-0 to 44-0, e.g. Chinese.

Megadont . : . above 44-0, e.g. Australians.

As the basion-nasion length varies with the length of the skull, it

follows that the relative value of teeth of the same size to it must vary

inversely with this quantity. Thus the Polynesians, who have large

teeth, but also a great length of skull, are classed as microdont, and

the Andamans, although they have small teeth, are classed as mega-

dont, because of the shortness of the base of the skull. The index,

however, is useful in the majority of races if used with caution. The

Bushman has a short base to his skull, which must tend to increase

the relative size of his teeth according to this index; but in spite of

this the index is fairly low, showing the smallness of his teeth. The

data derived from this investigation are tabulated below.

Average length of premolars + molars: Bushman. Australian.

Upper arches : . (24) 41-0 mm. (159) 45°38 mm.

Lower arches : . (30) 45-1 mm. (105) 49-6 mm.

Range. Upper maximum . : 44-0 mm. 50-5 mm.

Upper minimum . 38-0 mim. 40-0 mm.

Lower maximum . 47-5 mm. 55:0 mm.

Lower minimum 41-0 mm. 41-0 mm.

Dental Index.

pilcgadont: Baboons 4) 58-9 (Drennan).

Gorillas

Australians

Kalahari Bushmen .

Mesodont. Kaffirs

Bushman tribe

Cape Colony and

( (

. (2) 53-9 (Drennan).

. (86) 46-1 (Campbell).

(4) 44-5 (Broom).

. (4) 43-6 (Drennan).

. (15) 43-5 (Drennan).

Griqualand W. Bushmen (5) 42-1 (Broom)

Microdont. S.A. Bushmen ; (7) 41:3 (Drennan).

Europeans (99) 41-2 (Flower).

Hottentots (6) 40-3 (Broom).

Bushmen (7) 39-6 (Shrubsall).

Koranas (7) 38-5 (Broom)

76 Annals of the South African Museum.

The following table sets out a comparison between the average

mesio-distal and linguo-labial diameters of the teeth of young skulls

and the corresponding measurements of the teeth of the whole tribe,

including the young individuals but excluding the Bantu types. A

young skull is defined as one in which the third molars have not com-

pletely erupted into their adult positions. The teeth of the young

individuals are almost invariably larger than the corresponding teeth

in adults, probably because they have not been worn away to the same

extent as the latter.

Upper jaw. | Lower jaw.

Young | . Adult Young Adult

| skulls. | skulls. | skulls. skulls.

Central incisor, M-D. | (4) 8:8 | (25) 8:3 | (38) 5:3) "(2255-0

a *: 4) 077) (i) 6:51) (3) 25-22) ae

Lateral incisor, M.D. | (4) 7-1 | (24) 6:7 | (8) 5:8 | (24) 5-6

- i te) (4) 6-0") (26), 6:0" | (8) 56's (23) aioe

Canine, MD. | 4) a7 | (26) 7-5) (4) 6-95 5 (29) eas

as Ti.) (4) °8:0) | (26) S78 | (4)) 7-328) eee

Tstipremolar;, MEDI) (5) 7) |) (27); 96-8.) (4) 6:90 2rd

* TSO) 48:9) (25) 826 (4) base 2G) eco

2nd premolar, M.D. | (5) 6-8 | (28) 6-5 | (4) 7-2 | (80) 7-0

i Tush.) | (6) 8°83) | (26) 8-55) (4) 7-6 (SO) aes

Ist molar, M.D. | () 10-4) (83) 9:9) 7), (7) DiS (6 3) anos

oe LL. | (7) 10:8 | (83)-10°6 |. (7) 10:27 (SS)o2

2nd molar, M.D. | (4) 9-8 | (80) 9:7 | (4) 11-0 | (82) M0%6

oe L.L. | (4) 10-9 | (30) 10-6 | (4) 10-0 | (82) 10-1

3rd molar, MED. | (8) 8-7 | (26) 8:2: | (1) 10:0 I@3)magcS

“ ie.) (8) 10°38 |- (26) 10-3" a) 9-3 GS) eees

PHYSIOLOGICAL CONSIDERATIONS.

Attrition.

One of the most characteristic features of the dentition of the pre-

historic and of the primitive modern races is the wearing away of the

crowns of the teeth as the result of mastication. This attrition is

most noticeable on the surface of the crowns, where it 1s referred to as

“occlusal attrition,” but 1t also occurs between the teeth, when it is

called “‘interproximal attrition.”” The former is due to the friction

between the opposing tooth surfaces, and it is generally regarded as

The Dentition of a Bushman Tribe. 77

being associated with the use of hard, gritty foods by these people,

but it is probably largely due to the use of soft foods contaminated

with sand. It is also associated with the characteristic edge-to-edge

bite of these races. The latter interproximal attrition is due to the

friction between adjacent teeth as the result of the slight lateral

movements of the teeth in mastication.

With a view to estimating and comparing the amount of occlusal

attrition in different races, Broca has suggested classifying such denti-

tions into four groups according to the degree of wear :

1. Those with enamel worn without cusp obliteration or exposure

of dentine.

2. Those with cusps worn down and dentine exposed.

3. Those in which quite an appreciable amount of the crown is

worn away.

4, Those in which most of the crown has disappeared and the wear

has extended to the neck of the tooth.

Campbell has adopted this classification in his monograph on the

dentition of the Australian aboriginal, and his results in an abridged

form are given below, together with the results for this Bushman tribe

as a whole, and for the 8.A. Bushmen in the Anatomy Department of

the University of Capetown.

If a TTT, TINTS

‘Per cent.|/Per cent./Per cent.|Per cent.

Australian aboriginals (168) 23 39 26 12

Bushman tribe (51) . 27 51 22 oe

S.A. Bushmen (28) . ; 11 32 25 32

From the table it can be seen that the great majority of the denti-

tions of the Bushman tribe are only moderately worn away. There

are no instances of extreme attrition, such as seems to be present in

the Australian group, and such as is shown by about a third of the

Bushman skulls in the Anatomy Department. These latter Bushmen

have come from rock-shelters and caves on the sea-coast and in the

interior, and there is no doubt that for the most part they were quite

uncivilised in their habits of diet, etc. On the other hand, the Bush-

man tribe had certainly reached a semi-civilised condition of life,

with more modern habits of diet. We know that, in Europe, people

who were identical in features to the present-day races presented

78 Annals of the South African Museum.

extreme degrees of attrition, but that with the gradual so-called

progress of civilisation the attrition has become less and less, until it

has practically disappeared in the course of centuries. The present

results, however, point to the fact that the degree of attrition can be

very much reduced in a few generations.

This is not altogether unexpected, because, whilst in Europe the

progress of civilisation has been very gradual, it has been relatively

sudden in its impact on the South African aboriginals. Inasmuch

also as attrition is to a considerable extent due to functional causes,

such as the use of hard, gritty foods, one cannot expect it to persist

to the same extent once the chief exciting cause has been removed.

The acquired results of use alone are not likely to be perpetuated

indefinitely. “ If use had any effect here,” states Sir Arthur Keith in

an interesting paper on the adaptational machinery concerned in the

evolution of man’s body (6), “ it would be to produce teeth with eroded

crowns’; that is to say, the teeth would tend to erupt with such crowns!

Nevertheless there may be hereditary factors sufficient to account

for the special tendency towards attrition found in some races and

in some individuals, which have been brought about as the result of

some evolutionary mechanism modifying the relative proportions of

the upper and lower teeth and jaws, and this will be referred to again

in connection with the bite.

In functioning. the great majority of these incisor and canine teeth

must have met edge to edge, as they mostly show quite considerable

evidence of wear. The nature of the wear on the molars is interesting.

The first and second molars have become worn down on the lingual

aspect in the case of the upper teeth, and on the labial aspect in the

lower jaw. ‘The third molar does not show anything like the same

degree of attrition as the first and second in either of the jaws, but

what little wear there is affects the lingual aspect in the upper wisdom

tooth and the labial aspect of the lower one. This seems to be a

different state of affairs from that found by Campbell in the Australian

aborigines.

Most of these skulls, however, show in a very definite manner the

twisted occlusal plane between the opposing molars, as described by

Campbell in the Australian. This plane slopes downwards and

outwards between the first molars, and downwards and inwards

between the third molars, instead of there being a gentle downward

and outward slope between all three molars, as in Europeans.

The plane of attrition coincides with this plane between the first

two molars, but not in the case of the third. The downward and inward

The Dentition of a Bushman Tribe. i

slope behind is due to a most pronounced inward tilt on the part of

the third lower molars, and to an opposite outward and upward tilt

on the part of the third upper molars.

The few temporary teeth, which were still present in the dentitions

of some of the young individuals of this tribe, all showed quite a

definite degree of attrition. In these same young skulls it was also

observed that the permanent incisors showed the characteristic

crenated edge as they erupted through the alveolus. This feature

seems to be absent in the Australian, simce Campbell states that “‘ From

the material available, it would seem that the three developmental

tubercles on the unworn incisal border, which are often seen in white

children’s teeth, are either absent or not well accentuated.”

It is worth noting in connection with attrition that, although it

often seems to involve serious destruction of the teeth, it does not seem

to predispose to caries. In fact, races showing the most extreme

degrees of attrition are particularly immune to this disease. There is

hardly an example in this tribe of definite caries affecting the worn-

down surfaces. It would seem therefore that, given teeth which are

sound in structure and composition, mechanical injuries are not

sufficient to give rise to caries. In fact, the smoothing out of the

fissures between the cusps may be a factor in the prevention of caries.

On the other hand, it will be pointed out in a later section how

frequently the wearing down of the teeth far enough to expose the pulp

cavity leads in these races to apical abscesses and periodontitis.

The Bute.

The bite in the great majority of the individuals of this tribe is

on the whole of an edge-to-edge type. There is, however, a slight

tendency towards overlapping, when compared with the more accurate

edge-to-edge bite of the uncivilised Bushman, but in no case is there

anything like the degree of overlapping which is so prevalent in

European dentitions.

The association of this type of bite with the attrition introduces

morphological conceptions into these questions, because the nature

of the bite must be determined to a large extent by the relative sizes

of the teeth in the two jaws, and by the relative sizes of the teeth and

jaws themselves. |

It has been already shown how much less degenerate the mandibular

teeth of this tribe appear to be when compared with the condition

of the maxillary teeth. This points to a relative diminution of the

80 Annals of the South African Museum.

maxillae relative to the mandible. The presence of relatively well-

developed mandibular teeth would seem, therefore, to be sufficient to

account for the edge-to-edge bite manifested by this tribe.

One has still to explain the slightly greater overlap in this group

when compared with the more accurate edge-to-edge bite of uncivilised

Bushmen. I think this can be accounted for to a large extent by the

fact that the attrition in this tribe has not advanced to the same

degree as in the uncivilised Bushmen. Sim Wallace, in a recent

monograph on “The Variations in the Form of the Jaws” (7), states:

“Until growth has stopped (say about the age of twenty-one), the

amount of wear of the cusps never permits of edge-to-edge bite. We

have seen that in a well-developed jaw the forward growth tends

towards this when there has been attrition of the teeth. We know,

too, that in prehistoric times our ancestors had generally edge-to-edge

bite. But they did not have it till adult life and until the wearing

of the cusps of the teeth permitted it. The overlap of the incisors

was normal with Neolithic man, just as it still is with savages until

they wear the cusps of the teeth down.”’

It is possible, however, that in presenting the slight overlap as

opposed to the more definite edge-to-edge bite of the uncivilised

Bushman, this tribe has begun to show a variation of a more inherent

character. It is still a somewhat moot point how far these modern

tendencies in the dentition are variations which are the result of

definite modifications of the germ plasm, and are therefore truly

hereditary, and how far they have simply been acquired as the result

of environmental influences which have been imposed perhaps very

early in the development of the teeth. The majority of authorities

are of opinion that for some unexplained reason and by some as yet

obscure mechanism the size of the teeth and the size of the jaws are

diminishing, but to what extent, if any, the reduction in teeth and

jaws is correlated is not yet clear. Humphreys in a paper on

“Function in the Evolution of Man’s Dentition” (8) states: “The

plastic character of the teeth from the evolutionary standpoint is

well illustrated by the extreme rapidity with which the decline of the

third molar . . . has become established in modern races.”

Sim Wallace (7), on the other hand, whilst admitting that hereditary

variation as regards size of teeth, etc., cannot be doubted, is of opinion

that “‘hereditary degeneracy in the civilised cannot be admitted until

it is proved that lack of the trophic stimuli of mastication and normal

healthy development is not sufficient to account for the abnormally

small jaws and crowded dental arches generally seen among those in

The Dentition of a Bushman Tribe. 81

whom these stimuli have been in abeyance.” He also states that “‘we

have thus evidence that the germ plasm has not been permanently

injured with regard to the possibility of the civilised having as regular

teeth as prehistoric man.”

Until we are able to distinguish the great modifications of the teeth

and jaws which are undoubtedly acquired as the result of defective

diet, imperfect function, and other environmental causes, from the

normal hereditary variations ; and until we have more knowledge as

to whether or not these acquired variations are in any way trans-

mitted, and as to the cause and mechanism of ordinary variations, it

cannot be said that the slightly greater overlap in the bite, and the

slightly greater degree of irregularity in the dentition of this tribe,

when compared with that of the uncivilised Bushman, constitute

evidence of any morphological change in the “ Bushman”’ dentition

in the short time under consideration.

Interproximal attrition is also a very noticeable feature in the

majority of the dentitions of this Bushman tribe, and as it is rarely

met with in the highly civilised races, it is worthy of some consideration.

There can be little doubt but that it is brought about by a slight

lateral movement taking place in the vigorous mastication of very

hard foods. It does not appear to be at all pathological, nor does it

seem to be the result of the slight alteration of the shape of the teeth

due to occlusal attrition.

The loss of tooth substance due to this mutual attrition does not

leave open spaces between the worn teeth, but these are brought into

close contact and almost interlocked. As a consequence the measure-

ment from premolar to molar as taken for Flower’s dental index is

less than the sum of the ordinary measurements of the individual teeth

taken, where the attrition is less marked. Campbell used this differ-

ence to denote the extent of the interproximal attrition in the pre-

molars and molars. He found the difference to be 1-6 mm. in the upper

jaw and 2-4 mm. in the lower jaw. In the present group the follow-

ing is the result :—

Upper. Lower.

Sum of the average M.D. diameters of the

individual premolars and molars . 41-1 45-3

Average premolar-molar length 41-0 45-1

Difference (7.c. wear) 0-1 0-2

VOL. XXIV, PART 1. 6

82 Annals of the South African Museum.

The result is a very insignificant difference, and I feel confident that

it does not express anything like the amount of attrition which takes

place. The method is further open to the objection that the premolar-

molar length is the measurement of the chord of a flat arc, and that

the individual measurements of the teeth as taken may be compared

to measuring five small segments of the arc and adding them together.

The result is bound always to exceed the single measurement of the

chord, apart from the presence or absence of attrition. I have there-

fore decided to take the difference between the sum of the averages

of premolars and molars of all the Bushmen in the tribe, and the sum

of the averages of those of all the young Bushmen in the tribe, as a

better indication of the premolar and molar attrition. This gives

1-7 mm. for the upper jaw and 1-1 mm. for the lower jaw.

PATHOLOGICAL CONDITIONS.

Caries.

The incidence of caries in this little group of skulls, including all the

types, and the incidence in a small group of uncivilised Bushmen, 1s

given in the subjoied tables, together with a selection of data from

Campbell’s work for comparison.

Rae: Percentage of persons or skulls

showing carious teeth.

Hsquimaux : : 1-4

Maoris ‘ : 3-0

Indians of N.W. American !

coast ‘ 39

Fiji Islanders 5-2

North American Indians 9-5)

Australians : , (583) 13-5 (Campbell).

Zulus : 14-2

Bushmen . ' : 20-6

Negroes (slaves) . ; 20-8

Bushman tribe. : (53) 22-6 (Drennan).

S.A. Bushmen. (34) 32-4 (Drennan).

English and Scottish school- (10,500) 86-0 (British Dental

boys and girls . Association).

North German school children (19,725) 95-0 (Schleswig-Holstein

Dental Association).

The Dentition of a Bushman Tribe. 83

Race epier a ne | Perens

Australian aborigines , ; 10,561 1-6

Bushman tribe ; : : 1,211 2°6

S.A. Bushmen . : : oom 55

Huropeans : : ; 3,422 40-4

The incidence of caries in the teeth, which were present in the skulls

of this tribe, amounted to 2-6 per cent. ; and there was a total dental

defect of 12 per cent. in the dentition due to pathological and senile

losses and to caries. As the group of uncivilised Bushmen examined

showed an incidence of 5-5 per cent. of caries, and a total dental defect

of 16 per cent., the influence of civilisation does not appear to have

damaged their dentitions.

Of the skulls of the Bushman tribe 22-6 per cent. showed caries,

usually in a very few teeth in each case, whilst 50 per cent. of the skulls

had dental deficiency with pathological or senile loss of teeth or caries.

This is a better state of affairs than in the case of the uncivilised

Bushman skulls, 32-4 per cent. of which showed a degree of caries,

and 70 per cent. of which showed dental deficiency.

From such a small amount of material it is impossible to draw any

dogmatic conclusions as to the significance of the site of incidence of

the caries in this tribe. The back teeth were definitely more affected

than the front teeth, the wisdom teeth, especially the lower ones,

presenting about half the instances of caries.

The caries occurs twice as often in the interproximal situation as

it occurs on the occlusal surface, and it is usually low down, affecting

crown and neck.

About half the dentitions of the Bushman tribe showed evidence

of “ food-packing,”’ a depression of the alveolus between and around

the teeth, without being necessarily associated with alveolar absorp-

tion, but only about one-third of the uncivilised Bushman dentitions

showed evidence of this.

Erosion.

If we define “ erosion,” as Campbell does, as “ a condition character-

ised by tooth cavities situated near the gingival margin, and typified

by a wedge or saucer-shaped shallowness and a dense polished surface,”

this condition was absent from the Bushman tribe. It was present,

84 Annals of the South African Museum.

however, to a considerable extent in the uncivilised Bushman skulls.

There was a fair amount of evidence of an unhealthy state of many

of the dentitions in this tribe in the gingival region. Compared with

Kuropean dentitions there was a remarkable absence of tartar deposit.

The tissue of the necks of many of these teeth was soft and friable,

and even the enamel was in many instances soft and brittle, suggesting

a defective composition of the teeth or an unduly inimical medium

in which they were buried.

Periodontitis.

About a quarter of the mouths showed a more definite pathological

condition, with alveolar absorption, loosening of the teeth, and con-

siderable evidence of ante-mortem losses. This generalised condition

is in marked contrast to the condition met with in the uncivilised

Bushmen, 50 per cent. of whose skulls show one or more examples

of isolated typical apical abscess. These abscesses are the result of

infections spreading from the exposed pulp cavity or root canal up to

the apex. In many of these cases the teeth are worn down to the

roots, and the pin-hole opening into the root canal is patent. The

deepest part of the socket for such a tooth shows a nest-like cavity

surrounded by condensed alveolar tissue, and in many of such cases

there is a sinus opening through the alveolus.

Crowding.

The alignment, although rarely perfect, was good in thirty of the

dentitions examined, slightly irregular in twenty-one skulls, and -

in one skull there was considerable displacement of the teeth. The

alignment of the teeth is much better in the uncivilised Bushman

dentition, twenty out of twenty-five arcades being reeular, four being

slightly irregular, and one with several of the teeth zigzagging.

Variations.

The general form of the crowns and roots of all the teeth in this |

Bushman tribe shows no variation from what is characteristic for the

hominidae. The characteristic feature is the general smallness in

size, which is brought out in the tables. Nor does there appear to

be anything unusual in the range of variation.

Three of the skulls in this tribe, namely, skulls numbered 18, 19, and

29, show a reduction in the number of the teeth, in that they lack the

two upper third molars. These skulls seem to have definitely com-

The Dentition of a Bushman Tribe. 85

pleted their growth, and Mr. Leslie G. Gray, L.D.S., has kindly verified

their absence by radiography. ‘This reduction in the number of teeth

in the upper jaw is in harmony with the general tendency towards

lack of full development of the upper teeth already noted.

One skull, No. 34, shows signs of having had three definite upper left

incisor teeth. This skull also shows a remarkable mandibular torus,

and it has a peculiar opening in the supero-medial angle of the left

orbit communicating with the anterior cranial fossa.

In skull No. 5 there is a small enamel nodule on the labial surface

of the crown of the upper right wisdom tooth.

Skull No. 11 has a deficiency in the floor of the left orbit, so that

this cavity communicates with the maxillary antrum.

SUMMARY.

1. This paper contains an anthropometric survey of the teeth of a

Bushman tribe, the skeletons of which were kindly loaned to the author

for this work by Dr. L. Gill, Director of the South African Museum,

Capetown.

2. Mr. J. Drury, modeller to the South African Museum, has

supplied the data regarding the excavations.

3. The collection comprises the remains of fifty-three individuals,

reputed to be Bushmen, and reasons are given in the text for classify-

ing them into a group of approximately forty-four Bushman and nine

Bantu types for purposes of study.

4. With a view to providing the necessary comparative data, which,

with the exception of Campbell’s work on the Australian aboriginal,

seems to be entirely wanting, the author also includes a preliminary

examination of the dentitions of the uncivilised South African Bush-

man, of the Kaffir, and of a few gorillas and baboons.

5. The author follows to a great extent the plan of work adopted

by Campbell. The various dimensions of the crowns and roots, etc.,

are set out in a tabular manner for comparison. The morphological

conclusions arising from this data are then dealt with. This is followed

by notes on: .

6. The size and cusp pattern of the molars, and the condition of

the roots.

7. Flower’s dental index.

8. A comparison between the size of the permanent teeth in young

and older skulls.

9. Physiological considerations, such as attrition and the bite.

86 Annals of the South African Museum.

10. Pathological considerations, such as caries, erosion, periodontitis,

apical abscesses, and crowding.

11. Variations.

In conclusion I wish to thank Mr. A. A. Lamb, my technical

assistant, for checking the various calculations. I am also indebted

to Sir J. F. Colyer, who examined these dentitions, for valuable

criticism and suggestions regarding the pathological section.

REFERENCES,

(1) CampBett, T. D.—Dentition and Palate of the Australian Aboriginal, Keith

Sheridan Foundation, University of Adelaide, The Hassell Press, 1925.

(2) Martin, Rupotr.—Zur Anthropologie der Buschmanner, Berlin, Dietrich

Reimer (Ernst Vohsen), A.G., 1926.

(3) SHRUBSALL, F. C.—‘‘ Notes on some Bushman Crania and Bones,” Annals of

the South African Museum, vol. v, pt. 5.

(4) Broom, R.—‘‘ A Contribution to the Craniology of the Yellow-Skinned Races

of South Africa,” Journ. Anthrop. Inst., London, vol. liii, 1923.

(5) Duckwortse, W. L. H.—Morphology and Anthropology, Cambridge: At

the University Press, 1904.

(6) Kerra, Str A.—‘* The Adaptational Machinery concerned in the Evolution of

Man’s Body,” Nature, vol. exii, August 1923.

(7) Watuacr, J. Stu.—Variations in the Form of the Jaws, Bailliere, Tindall

& Cox, London, 1927.

(8) Humpureys, H. F.—“ Function in the Evolution of Man’s Dentition,”’ British

Dental Journal, October 1921.

(9) NicHotis, Brookr.—‘‘ Transactions of the Sixth International Dental Con-

gress,” published by the Committee of Organisation, Hanover Square,

London, W.

Tooth Measurements.

The Dentition of a Bushman Tribe.

Mesio-distal diameter:

Observa-| Aver

“| Max.

onl

—

i—|

He oS

~ —

an i=

& ie

i ® o

S > on

} aS

—4 ane

ie)

Re]

< s

a0 —

[=]

x H :

— o

@ Ae

eee

rey

—

Pa al

5 =

= Jf

s =

2 =

3 =

ioral .

3 ae

tenn) an

q qs

2 rw

3 o 8

4 Sage

jo)

—_

Upper Arch.

SONSSSOS

Cm MNMOO®

fe ee ce oe ee

19 S19 S 2919 29

ID DAH HD eHo

Sess esest See ee

19 19 49 19 DOOD

ANNODHAANAN

rssaesseesoeesteee

DOOMrwM1O HS

SeNe ae

—

DO OWOO19N 1 WO

DAHADMOOH

SoaNe aaa

19 O00 S199

19190 M1010 Om Ss

NANNANAAN

DHAAAOHS

sot row

ANNAA TRS

Ne eee

DMOrMMO HO

Nee

YS YS ES rrSrwTSErr TET

[EE

G8 OD 09 Te) One

Dorr Kradad

1d DH W119 69 09 OD

LODADHAN

DODD IDOO

SSR DDSOOS

att

Sooo MDMNDS

HOoroconor

—o

1919 69 GO 6A 69 GO

AKNOrMAISH

oO P19 DID QIN

DOrDOAaARO

a nm

1D sH CO P= CO 3 OO

ANNAN HOON

.s he Sat

Oo ey

Ge el See. 2

Cy er TS! ck

5) oo °

Beane Sues

a= colina

aay aS) pes Ss

anoaSmaond &é

ANORNANANS

Lower Arch.

12:€

16.0

19-3

20-0

(20)

)

¢ 1)

16.0

18:0

21:0

29-3

18-8 | 22-0

20:6 | 24-0

23-6

18-6

17-6

SeSOonmoows

1d 1D CO CO =~ G2 CH

29Hooonnms

DONDABAAGS

il C=

NODOONS

Morr rooe

SODDODSD

ww oeoLOOG

Pal

Ss eae

~ fe)

aie cotcsharer

—_— a Lael — . .

oO oo e)

fs Sl |

fo) a

Te Gg Ty Teg

ay 1S} pe Ors

masSnanags

SNUANANM

pes

“ANNALS

OF THE

SOUTH: AFRICAN MUSEUM

VOLUME XXIV.

PART II, containing :-—

5. Preliminary Report on the Anthropological Researches carried out

by Mr Drury of the South African Museum in South West

Africa. (With Plates XII-X XII and 1 Text-figure.)

| pen Wasting tS

éy = ia ss aw v4

erronae muse CA

ISSUED JULY 1935. PRICE 2s. Gd.

PRINTED FOR THE

TRUSTEES OF THE SOUTH AFRICAN MUSEUM, CAPE TOWN

BY NEILL AND CO., LTD.,

212 CAUSEWAYSIDE, EDINBURGH.

5. Preliminary Report on the Anthropological Researches carried out

by Mr Drury of the South African Museum in South West

Africa.

(With Plates XII-X XII, and 1 Text-fig.)

[In 1919 the South African Museum was enabled to extend to the

region embraced in the mandated territory of South West Africa

the Anthropological Researches on the Bushmen and other native

races which it had been pursuing in past years in other parts of South

Africa. Every facility was extended to the Museum Expedition by

the Administration and the late Sir Howard Gorges.

In 1920-21 the Administration of South West Africa voted a

grant-in-aid to the Museum for the purpose of continuing these

researches, and the trustees desire to take this opportunity of expres-

sing their thanks to the Administration and the then Administrator,

Mr. Guys Hofmeyr. To Dr. Fourie, ex-Medical Officer of Health at

Windhoek, who is a keen student of the Bushmen and other native

races, the Museum is under a debt of gratitude for the interest he

took in the work, and the considerable amount of time and trouble

he spent in helping to organise these expeditions.

The task of obtaining anthropological and ethnological data was

entrusted to Mr. J. Drury, the Museum taxidermist and modeller,

who has had considerable experience of the primitive peoples. In

1919 Mr. Drury visited the Grootfontein district, taking casts, photo-

graphs, and measurements at Grootfontein, Nuragas, and Otjituo.

Mr. Drury then visited Gaub, between Grootfontein and Tsumeb,

for the purpose of obtaining casts of rock gravings.

In 1921 Mr. Drury carried out a further and more extensive ex-

pedition. After casting a large number of the natives of various

tribes employed in the prison at Windhoek, he went out to Sand-

fontein, where Dr. Fourie had reported a tribe of Bushmen still living

in their primitive manner, and apparently but little influenced,

physically or culturally, by Bantu or other races. On this expedition

a number of casts and measurements, a large number of still-life

photographs, and some hundreds of feet of cinematograph film were

taken. All this material is preserved in the South African Museum.

In addition to the material and data obtained by Mr. Drury, other

VOL. XXIV, PART 2. 7

90 Annals of the South African Museum.

members of the Museum staff obtained useful information in the

course of the Zoological and Botanical Survey which was carried

out on behalf of the Administration. In 1921 and 1923 Dr. K. H.

Barnard, Assistant Director, obtained photographs, a large number

of measurements, and many notes on the customs of the various tribes

of Ovambos. In 1925 Mr. R. F. Lawrence, in the course of an expedi-

tion to Zesfontein in the Kaokoveld, visited the rock gravings on the

farm of Mr. C. Schlettwein at Otjitambi, north-west of Outjo.

In the following report I have endeavoured to give, for the most

part in Mr. Drury’s own words, the substance of the very valuable

observations which he made on the habits, customs, occupations, and

physical characteristics of the tribes which he visited.

Mr. A. J. H. Goodwin, Research Assistant in Ethnology at the

University of Cape Town, has read through the original notes and has

helped in various ways in preparing them for publication. I am

indebted to him also for the drawings reproduced in the text-figure.

It should be understood that the photographs, both still-life and

cinematograph, accompanying this paper are only a selection of

those which the Museum possesses. A more comprehensive and

fully illustrated account of the physical characteristics of these

peoples, based on Museum material, especially of certain features

which are outside the scope of the present report, will be published

on a future occasion.

Miss Bleek, who accompanied the 1921 expedition, studied the

language and folk-lore of these tribes, and has published a work on

this aspect of the subject.*—Eprror. |

THe BusHMAN TRIBES AT SANDFONTEIN.

Sandfontein hes due east of Windhoek on the border-line between

South West Africa and Bechuanaland. The railway now runs as far

as Gobabis. (See map: Ann. 8. Afr. Mus., vol. xxix, pl. 4, 1931.)

Two tribes were found living in this locality: the Auen and the

Naron. The latter appeared to be the true local Bushmen, whereas

the former had wandered over from another district some miles away.

Mr. Drury was unable to discover exactly whence they had come.

These tribes were the most interesting with which members of the

Museum staff have come into contact. They appear to be by no

means pure in blood, having apparently a strain of Bantu in them.

* The Naron, a Bushman Tribe of the Central Kalahari. By D. F. Bleek.

Publicat. School Afr., Life and Language, University of Cape Town. 1928.

Preliminary Report on Anthropological Researches in S.W. Africa. 91

Their language and culture, however, appear to have been but little

influenced.

The two tribes were at first sight indistinguishable, and it was

only after they had been studied for some considerable time that

the differences between them were noted; even so, the differences

are difficult to state in concrete terms. The pigmentation of the

skin, the material culture, and the language are similar, but on the

whole the Auen are slightly taller and more powerfully built, and

appear to be more adept in the manufacture of utensils and imple-

ments. In demonstrating the manufacture and use of any article,

the demonstrator was nearly always one of the Auen.

PHYSICAL CHARACTERISTICS.

In colour these people are of a dirty mahogany. Some are darker

than others, and some show a slight yellowish tinge, but not exactly

like that found in the Cape Bushmen. The following table gives an

idea of the colours of different parts of the body according to the

colour-chart of von Luschan:—

Neck. Upper | Lower | Upper | Lower

No. No. Face. Trunk. | Trunk. | Limbs. | Limbs.

| 1 16 25 Bf 25 28 Se

| 2 15 25 i, 28 ie so dle

Auen - 3 14 25 sf 26 a aie 28

| 9 22 25 e820 —2.0) mee 26 29

10 18 24 we 25 24 a 26

12 4 28 33 BE se ay ay

13 2 | 24-26] 26 24 24 25 30

N i Ake) 12 22 5s 25-26 | 28

aron

93 | 91 |21-22/21-22/ 23 | 93 | 25 | 30

Dig) We OOM 3) SOm MSO nl Mead ts

el ee oa ee “i Pi on*ocimos

The hair is a dirty brownish-black. It is usually short and thick,

but in some cases, both among the men and the women, it was worn

in plaits which hung down round the ears and back of neck, resembling

a mop (Plate XIV, fig. 1, and Plate XVIII). The eyes were of a rich

brown colour, with a ring round the iris, varying from dark brown to

pale grey, or sometimes blue.

92 Annals of the South African Museum.

The chart at the end of Mr Drury’s report gives the comparative

measurements of forty of these people taken on the living subject,

from which it appears that the average height of the Naron is

150 cm., while that of the Auen is 158 em.*

DRESS AND ORNAMENTATION (Plates XIV, XVII, XVIII).

Dress.—The costume of both peoples is similar, the material used

being skin brayed to the consistency of chamois leather, the hair

having been removed. The men wear a loin skin, only sufficient to

cover the genitalia, and this is usually the only garment. In some

cases, however, the men weara mantle. This would consist of a large

skin, hair removed and dressed soft, which is slung from the shoulders.

This skin would serve as a mantle by day and a blanket by night.

The boys often made use of this skin as a protection against the

sun’s rays.

The women, like the men, wear a loin skin, though not always, and

an undermantle hung from the shoulders, over which lies a still larger

skin, serving the same uses as the man’s garment. A few of the

women wear a small semi-circular pendant apron, often edged with

beads, and sometimes laced, through cuts in the fabric, with pieces

of leather and string. The women almost invariably carry a powder-

puff (Plate XIV, fig. 2): the outfit consisting of a tortoise-shell, sealed

at the one end, a small puff of soft fur from a hare or other animal,

and a sweet-smelling powder of herbs, brown in colour. The women

powder their faces with the puff.

Ornamentation.—This consisted of strings of ostrich egg-shell beads

for the most part, and in a few instances of European trade beads.

A well-dressed woman would have bangles, and pendants of ostrich

egg-shell beads tied to the hair and suspended to hang between the

eyes. Besides these, thin medallions or short strings of beads are

hung all round the head. Round the neck, arms, and legs would be

a few strings of beads: on the arms and legs would also be bracelets

and armlets cut solid from raw hide and slipped over the member to

dry in place, in some instances indenting the flesh to a considerable

degree. These were usually cut in such a way that two tassels hung

down from the bracelet, one inside the limb and one on the outside.

The older women are not so fond of dressing, their karosses being

in most instances old, having acquired the gloss and smell of a well-

* Cf. Schapera, The Khoisan Peoples of South Africa, London, 1930, chap. iii,

and frontispiece.

+ Cf. Schapera, loc. cit., plates vii and vii.

Preliminary Report on Anthropological Researches in S.W. Africa. 93

worn garment. It is customary for the older women to give al]

their good things to the daughters on their marriage.

The men wear very little in the way of ornamentation, such finery

being confined to the unmarried men, who would wear a few bangles

at most. The older men content themselves with a few leather

bangles similar to those worn by the women. These are usually

worn on the legs, glass beads being preferred for the arms.

Huts (Plate XII).

The Naron huts (the huts of the Auen were not seen) are constructed

of a number of light branches, the thicker ends of which are trimmed

and forced into the earth. When sufficient have been erected, they

are brought over to form an arch, and tied. The hut is formed on a

ground plan which consists of the greater portion of a circle, the one

side being left entirely open on the east as an entrance. The whole

is then covered with grass, little attempt being made at thatching.

The hut is usually 6 or 7 feet across, and some 5 to 6 feet high. In

summer this shelter is comfortable and holds off the west winds,

which are prevalent. In winter, however, which is the rainy season,

they cannot be of much use. The opening faces the east to allow of

the morning sun entering. Such a hut would be inhabited by a

family of three to five, consisting of the parents and two or three

younger children. These tribes have seldom more than two children

to a family. The elder children are separated from their families

until they marry.

It was noted that each hut was sacred to its owners, and that

visitors were always received outside.

ContTENTS OF Hut.

The furnishing is sight. The beds consist of dry grass along the

back of the hut. Besides this, there are a few ostrich egg-shells

for holding drinking water, and a few skins. In warm weather the

fire is made just outside the hut. This fire consists of a few sticks

of burning wood resting on the accumulated ashes of past fires. In

the cold weather, however, which here coincides with the rainy season,

the fire is brought into the hut, where the inhabitants sit and sleep

round its blaze.

Near the entrance to the hut are to be found two or three large

stones used as anvils or for grinding berries, and a few smaller hammer-

stones and rubbing-stones.

94 Annals of the South African Museum.

VILLAGE.

There would appear to be no village form, but the huts are scattered

about in a clearing in the bush, the distance between them varying

from 10 to 20 paces.

One very poor hut, merely a shelter, was occupied by two unmarried

girls under the charge of an old woman. These two girls were

awaiting marriage. Youths of suitable age are similarly segregated

in bachelors’ quarters. They were in a worse condition, having only

a tree with a few bunches of grass stuck in the branches to protect

themselves. Here they stayed until such time as they should marry

and erect their own huts.

UTENSILS.

These are few, being governed by their needs as a hunting people.

Ostrich egg-shells are laid on their side, the blow-hole being made

on the one side towards the end to allow of the egg being placed in

a position of greater stability.

In the hut was usually a pointed stick for digging roots and bulbs,

and very often a wooden pestle and mortar, used for pounding up

the dry Rasyntje * (Grewia cana) berries. Nets for carrying their

belongings, made of string, with a mesh of some 4 inches, were also

present. These would be packed with eggs, bags, etc., when on the

move, and slung on the back, on the digging stick in most instances.

There was often a reed sieve, a mat woven of slender reeds in such a

way as to leave a fine mesh, suitable for sifting termite larvae. Drills

for the manufacture of the ostrich egg-shell beads, and others for

making their soapstone smoking-pipes, and a gemsbok horn for the

making of string were also present.

For the work of braying and dressing skins they had a piece of

iron inserted into a piece of wood, the whole 18 inches long (text-fig. 3),

and pegs for stretching out the drying skin. Wooden spoons, and

wooden scoops for hoeing (text-fig. 1), completed their outfit as a rule.

Ostrich egg-shells were sucked dry through a hole made in the

side towards one end; what was wanted was eaten, and what was

to be kept was spat out into another vessel. Each hut would have

from two to three of these vessels, and in one case as many 4s SIX.

They would constitute the Bushman’s most valued possession, and

would be stood against the edge of the hut, hidden in the grass,

out of harm’s way and cool. To steal water is considered a criminal

offence, and is sometimes punishable by death where water is scarce.

* Afrikaans =a little raisin.

Preliminary Report on Anthropological Researches in S.W. Africa.

‘dus pasr ‘4 fSurpurq qqns “qh ‘uostod “d sdy UOMI-paqaeq ‘"q

. Q 2

(posuepusa ATQY SITS) (osn roy Apraa) Morte uony—'e

. aS (“Ee Ques ‘5 J 3 V ¢

te te ‘ (pesiepsue ATVySTS) (posdoAer) MOG UONW—'F “DTT ‘(aTvos oures 04) (uoat) 1adeaios UTyS UOIRN—'¢

(opissuoye opvos YATA\) (ou04s) Avg s19qus}jo[q WoT JUTENASUT—’ (

IT,

‘(poor Jo) yyeys “s Spoom 10 suo jo 4yeyseaoy “sf

WL (epissuoye opvos YIM) ,,Sureoy,, 10f onbelgG—'T *

“Oly

“OTN

OTT

96 Annals of the South African Museum.

In dry country, water is obtained from spots known to the Bushman,

where he can dig in the sands to find it. Sometimes he is driven to

seek further, and will take a reed, remove the pith, and tie fibres

round the end to act as a filter. This done, he inserts the wad into

the ground at some chosen spot and sucks the end of the reed until

water comes up. This is spat into an ostrich egg-shell. In some

cases, so great is the suction necessary to draw the first water that

blood is drawn from the mouth.

The stick used for digging bulbs is usually from 2 to 3 feet long

and a little over an inch in diameter; one end is sharpened to a

point. The stick is often used for carrying burdens over the shoulder,

for a walking-stick, a weapon, a lever, in fact almost anything, but

primarily it is intended for digging. They use. no pierced stones

for weighting this digging stick, nor do they know the use of these.

The pestle and mortar are of wood, and in most instances have a

block of stone inside, as the wood is not sufficiently hard to grind

the pips of the Rasyntje berry, which are usually ground. The mortar

is a hollowed block of wood some 15 inches high, and 7 inches in

diameter (Plate XII, fig. 2). The pestle is about 2 feet long, of

hard wood, and similar in shape to the European implement. The

person pounding would hold the mortar before him, with his feet,

or letting it stand and steadying it with the one hand. The pestle

would be held in both hands or the free hand.

Scoop or PrimitivE Hoe (Text-fig. 1).

In one of the huts was found a flat piece of wood, very similar in

shape to bone and stone specimens obtained from cave sites, as at

Coldstream (Cape). These latter the late Dr. Péringuey termed

“palettes.” They were used, according to his theory, for preparing

ochre for smearing the body (Ann. 8S. Afr. Mus., vol. viii, p. 160, pl.

xxvi, fig. 196). The interest of Mr. Drury was aroused by the instru-

ment, and he asked for a demonstration of its use. The owner of

the object proceeded to dig with his digging stick, then taking the

wooden plaque in his right hand, in the bend of the finger, the thumb

controlling it from behind, he proceeded to scrape out the earth.

On being asked why this was used, the man replied that the constant

digging of roots and bulbs was wearing to the fingers, and such an

implement was necessary. One is thus brought to the view that the

so-called palette of the Strandloopers is in reality, perhaps, the earliest

form of the hoe.

Preliminary Report on Anthropological Researches in S.W. Africa. 97

Weapons (Text-figs. 4, 5).

The only weapons possessed by these tribes are bows and arrows,

though they occasionally obtain assegais from Bantu tribes by bar-

tering ostrich egg-shell beads. Quite young children have bows and

arrows of proportionate size, and learn to use them by playing games

and shooting birds with them.

The bowmaker (it is not a special trade) had the string ready pre-

pared beside him, and had cut a suitable length of Rasyntje bush

(Grewia cana). He removed the bark of the branch and marked

off the required length of the bow by two rings, cut round the stick,

one towards each end. These ends were not, however, removed at

this stage. The stick was first shaped, the two ends being thinned

down as far as the marked rings. When the taper towards the ends

was complete, the thick portions past the marked rings was removed,

the ends trimmed, and the bow was ready for stringing. In some

instances pieces of gut were slipped over the ends of the bow and

round the middle in a raw state, where they would dry in position

and prevent the bow splitting, and also form a grip. The string

would then be tied to one end, the bow bent by holding it firmly

between the legs and pressing the end on the ground, then the other

end would be attached in such a manner that the string was just taut.

Arrows.—As reeds were scarce there was no chance of seeing these

being made. The arrows actually seen, however, were composite.

_ The foreshaft consisted of a piece of bone, some 23 inches to 3 inches

in length, sharpened to a point at each end. The hinder end fitted

into the reed shaft, the end of which was bound with gut, but not

sufficiently to grip the bone foreshaft. The fore end fitted into a

short section of reed, in which was a small iron point, sometimes

barbed, but more usually merely a spike. In most cases these arrows

were poisoned. When not in use, to protect the user, the bone

foreshaft, together with the head, is removed from the socket in the

reed shaft and reversed into the socket, thus concealing the poisoned

portions (fig. 4). The arrows are quite unfeathered. In the large

collection of arrows in the South African Museum the examples

most nearly approaching this type are from N’Gamiland and the

Grootfontein district. They are of a different type from ‘‘ Bushman”

arrows from Namaqualand. In using the bow and arrow the man

would have no difficulty in shooting an arrow with considerable force

a distance of over 100 feet. These tribes do not use quivers.

98 Annals of the South African Museum.

MANUFACTURE OF BEADS AND OTHER ARTICLES

(Plates XV and XX).

Unique opportunities were obtained of observing the method of

manufacture of ostrich egg-shell beads, rope, string, and soapstone

pipes, and the dressing of skins.

As varying accounts, mostly based on conjecture and not actual

observation, have been given regarding the manufacture of ostrich

egg-shell beads, I give Mr. Drury’s notes here in extenso. It is possible

that the details of manufacture may vary in different tribes and

localities. One of the casts in the Museum shows a woman engaged

in the work, together with the actual objects used by her.

Ostrich Egg-shell Beads.—These beads are made by the women.

An ostrich egg is first broken into large pieces and from these pieces

smaller pieces about half an inch in diameter are bitten off with the

teeth. After accumulating a number of these pieces the woman

proceeds to drill the holes. She takes her drill, which is a stick about

20 inches long, having an iron point at one end, and, placing the chip

of egg-shell on a hard piece of skin, she twirls the drill rapidly between

her palms. When she has drilled a hole nearly through the egg-shell

she reverses it and starts drilling from the other side, until the two

holes meet. She next proceeds to enlarge and smooth the edge of

the hole with the aid of a poimted rimer—a tool somewhat corre-

sponding with our bradawl, and consisting of a small piece of wood

tipped with a piece of pointed iron. The next step consists in roughly

trimming the perforated piece of egg-shell to the shape of a disc.

This is done by means of two suitable stones used as a hammer and

ananvil. The woman holds the piece of shell on the edge of the anvil,

and with the hammer proceeds to chip away the unnecessary corners

until it assumes a roughly circular outline. These roughly trimmed

dises are now threaded on a strong piece of gut, about 12 to 14 inches

in length. The discs are pressed closely together and the gut is

knotted at both ends. Bits of fibre, such as are used for making

string, are then twisted round between the discs, so that the whole

becomes a solid and rigid mass like a stick.

This stick of discs is next placed on a piece of hard skin or wood

and the operator proceeds to grind it down smooth with a rough-

grained stone. This process reduces all the beads to a uniform size

and beautifully rounded.

It has often been assumed that the beads, either singly or strung

together in bunches as above described, were rubbed on the stone.

Preliminary Report on Anthropological Researches in S.W. Africa. 99

These observations show that, so far as these tribes are concerned,

the reverse process is practised.

The final process consists in unstringing and cleaning the finished

beads, and restringing them on long strings until the desired length

has been made. In threading the beads these natives put the bead

over the tip of string, contrary to the method of threading beads on

needles which is in vogue among Europeans. The whole process of

manufacturing these beads is long and tedious. The woman who was

watched broke three out of four discs before completion. She had

taken just over three months (moons) to make a string of beads nearly

12 feet long, and was willing to part with the whole string for two

handfuls of tobacco.

Rope and String (Plate XX, figs. 1-5).—This is manufactured by

the men. In this industry a gemsbok horn (Oryx gazella), fashioned

to a blade along one side at the tip end, and a piece of hard skin

were used. The man has beside him a number of Sansevieria spines

or leaves. He holds the horn in the right hand and beneath the

right foot, the knife edge downwards, resting on the dry skin. He

takes a Sansevieria leaf, passes it beneath the blade, and allows the

foot to weigh upon it. He withdraws the leaf, thereby removing

with the blade the flesh, and leaving the fibre clean. This fibre is

thrown aside and the flesh put over it to keep it moist. A length of

prepared fibre is taken and the one end is held in the left hand, the

body of the fibre being twisted against the right thigh by the right

hand. The next piece is twisted into the last, and itself twisted to

string. This goes on until a sufficient length is obtained. The string

is even and strong, comparing well with the machine-made article.

Soapstone Pipes (Plate XX, fig. 7).—A drill and a rimer are used.

These consist of two pieces of iron, about 2 feet in length and a quarter

of an inch in diameter, flattened at one end to a “diamond point.”’

The soapstone is prepared by grinding it to a cigar shape, on a hard

rock. This is then placed between the maker’s feet as he squats,

and he proceeds to drill from the end. He holds the drill between

the palms of the hands, and rotates it with a rapid twirling motion,

in alternate directions, the hands moving down the stem of the drill;

when they reach the bottom he starts again at the top. When this

hole has reached a sufficient depth, he takes the larger drill in his

right hand, the pipe in his left, and proceeds to enlarge the hole

with a riming motion. He reverts to the smaller drill, and in time,

working from both ends, he bores right through the stone. This

done he proceeds to shape the outside of the bowl by rubbing it on

100 Annals of the South African Museum.

large rough boulders. This pipe takes several days of continuous

work. Many of these Bushmen are now using brass cartridge cases

converted into pipes.

Skin Dressing (Plate XX, fig. 6 and text-fig. 3).—This is done

with rough stones, and a stick with a blade inserted at the one end

(text-fig. 3). The flayed skin is stretched out on the ground and

pegged to dry. The man takes it up when dry, places it on the

ground, and, sitting on it, proceeds to scrape it on the fleshy side

until it is reasonably pliable. The skin is then taken in the hands

and wrung with a twisting motion until soft. Sometimes rough-

faced stones would be used. If the skin is too hard the maker places

it in the shade, covered with moist earth. Once the skin is soft it

is rubbed with grease, left to absorb this, and twisted and wrung

again. It is rubbed with ochre, giving it a reddish appearance, and

the fat is allowed to dry out. It is now as soft as kid. The time

taken to dress a cat skin may be three hours, but one man may take

three days to dress a goat skin—working, walking about, talking,

smoking, sleeping, and returning to the work.

A point of great interest is raised by comparing the iron scraper

used by these people to-day with a stone implement from Pletten-

berg Bay in the South African Museum (text-fig. 2).* The latter

consists of a stick covered with a resinous substance, in which is set

a stone flake of rough chisel shape. The instrument used by the

Bushmen is similar, but longer in the handle (18 inches long), the

thickened end corresponding with the resinous mass on the Pletten-

berg Bay specimen, and having inserted in it a broad, flat, chisel edge

of iron. This suggests a possible use for the Plettenberg Bay -

implement.

SUBSISTENCE.

These tribes are, in common with all the Bushmen, a hunting and

collecting people. Their food at the present day consists largely of

the berries of the Rasyntje (Grewia cana) bush, together with other

vegetable foods, such as the wild cucumber, wild melon, and several

varieties of plant roots.

These Rasyntje berries are small brown berries, a little larger

than a peppercorn. They grow on a bush that is common in these

parts, and reaches a height of some 5 to 6 feet. The taste is not bad,

but they are slightly astringent, and one does not take to them readily.

* Cf. Péringuey, Ann. 8. Afr. Mus., vol. viii, p. 159, pl. xix, fig. 150, 1911; and

Hewitt, Trans. Roy. Soc. 8. Afr., vol. x, p. 49, pl. i, 1921.

Preliminary Report on Anthropological Researches in S.W. Africa. 101

There is more seed than fruit; the Bushmen, however, either eat

stone and all, or dry the berries in the sun and pound them up,

including the stones. A drink is made from this pounded mass.

Even the dogs eat the berries off the trees. The wild cucumber

(Citrullus sp.) is about 4 to 5 inches long, and grows on a creeper which

grows into the trees. The skin is smooth and a pale olive green in

colour, marked with whitish stripes and spots, and ripening to a

brilliant red. The fruits are easily visible hanging from the bushes

on which the creeper has grown. This fruit is only eaten green,

never ripe. Haten raw they are far from palatable; heated on the

coals, however, they are better.

The melon (Acanthosicyos horrida) is about the size of a cricket

ball, and is covered with blunt spines. It is green, ripening to a

lemon yellow. They are even more bitter than the cucumber, and

are not liked by the Bush people.

Among the roots eaten is one large variety with a brown skin

like a potato, white, fleshy, and full of moisture inside. Raw, these

have an earthy taste, but cooked on the open fire they are better,

and with the addition of salt and pepper make a dish fit for any

table. The plant to which this root belongs is insignificant, showing

only two long thin stems growing from the bulb, similar to a creeper,

with a few small leaves.

Snakes, rats, and mice are never used as articles of diet; in fact,

Bush people have a dread of them in these parts. Looking through

the middens before the huts no remains of any such animals or reptiles

were found.

The wooden spoons used for eating mealie-meal porridge are prob-

ably a modern innovation from the surrounding Bantu. In general,

the life led by these folk is lazy, consisting of the search for food,

dancing, sleeping, and a few industries, which are often pure pastimes.

Dancine (Plates XIX and XXI).

There are several dances, all of them somewhat similar. Both

tribes took part in these dances. Dancing is natural to them, and

it hardly needs but a clap of the hands to bring them together and

start them off. The women would often appear to be wrapt up in

little songs and dances of their own, singing softly to themselves as

though in a dream, so vacant is the face. Very young children,

though hardly able to stand, would clap their hands and try to dance,

imitating their elders.*

* Cf. Schapera, loc. cit., chap. viii, and pl. xii.

102 Annals of the South African Museum.

When in the right mood the men and women going for water

would dance. They would line up in single file and dance their

way along, doing a sort of step-dance as they sang and clapped

their hands. Then after each “set” of the dance, they would reform

into two ranks, facing each other, dance a few steps thus, and proceed

again, and so on until reaching the water-hole. The return journey

with the water would be similar. Their voices sounded very sweet

as they came over the hills in the distance, and gradually grew clearer

as they approached.

The night, however, is the real time for dancing. They will then

dance from sunset to sunrise, particularly at the new moon or full

moon. The day before a big dance is spent in collecting food, roots,

bulbs, and such fare. The women and girls paint their faces in red

and black; a few dots here and there and a ring round the eyes

would be sufficient to bring out the beauty of the lady to the very

best advantage. The men wear rattles of the cocoon of a large moth

tied to their legs, and this is all their clothes, except for the loin

cloth. Among the dances noted were the eland dance, the ostrich

dance, the reed dance, and the porcupine dance—the latter being

more of a game than a dance.

Eland Dance (Plate XIX).—Two men were chosen to act as bulls,

and they had horns of wood tied to their heads. The five women

taking part as cows danced round in a figure of eight, divesting them-

selves of their clothing the while, the two eland bulls dancing after

them. While this dance went on a number of men and women looked

on, clapping and singing to the dancers.

Ostrich Dance (Plate X XI, fig. 4).—This is the men’s dance, con-:

sisting of a sort of leap-frog; each of the men in turn stopping and

stooping, the others jumping over him, as they danced in a circle.

Reed Dance (Plate XXI, figs. 6, 7).—Not of much interest; the

men and women all shuffle round together, singing and clapping

and grunting together. This they keep up for a considerable length

of time.

Porcupine Dance (Plate X XI, fig. 5).—Men’s dance. This is danced

in two lines, each facing the other. These lines dance backwards

and forwards to each other all at once, as though each man were

trying to out-dance his partner.

The music accompanying these dances is apt to pali and become

monotonous to the European after a time.

Games.—A few games are played, as, for instance, the “war game”

(Plate X XI, fig. 5). The men and boys would sit on the ground facing

Preliminary Report on Anthropological Researches in S.W. Africa. 103

each other in pairs, each sitting about 5 feet from his opponent. They

would then proceed to mimic throwing dirt at each other, thumping

their chests, and hitting the ground and grunting. This is done to a

tune and time similar to the dances. Each goes on until either he

or his partner is exhausted, the more enduring winning. While the

game goes on they hurl epithets at each other, and emphasise these

epithets of inferiority by their actions. They never strike each other,

but from time to time pretend to spring forward as if to do so.

Stick Game.—The men each had a slender stick of “ Rasyntje”’

bush, about 4 inches long. This they would throw on the ground in

front of them in such a way that it would bounce away from them-

selves; they would follow the stick and throw it again, the game

being to cover the greatest distance in the least number of throws.

This is a very much faster game than it sounds, and after it the players

would perspire very freely.

MusicaL InstRUMENTS (Plate XXI, figs. 1, 2).

Only one instrument was observed, the “Rhumkie,”’ which is a

sort of fiddle. The basis of this instrument is a hollowed-out bow,

which has a wire stretched along its length. In playing it the one

end rests over the right shoulder, while the other end is held with

the left hand, the little finger gripping the wood of the instrument,

and the other fingers controlling the pitch of the note.

The playing bow consists of a twig strung with some hair from a

buck’s tail, which is wetted before use by applying it to the mouth.

This hair the player controls also with the thumb of the right hand

as he plays, keeping the string taut. With this instrument the boys

play a soft and rather eerie tune.

CONCLUSIONS.

In the light of the observations made on these people we may

regard the Kalahari Bushmen and Cape Bushmen as widely diver-

gent, and neither of them appear to be pure Bushmen. The people

we have observed in the Kalahari, or Sandfontein, Gaberones, Kanye,

and Grootfontein, appear to be bastard tribes, composed of a Bushman

basis mixed with Hottentot and Bantu blood.

This fact is also recognised by Miss Bleek, Dornan, and others.

In discussing this question Mr. Drury obtained the opinion of four

old traders whom he met, respectively, at Vryburg, Prieska, Kanye,

104 Annals of the South African Museum.

and Sandfontein, and who had known the Bushmen for some years,

and had trekked through the Kalahari from end to end in the early

days. These traders, quite independently of each other, agreed in

their statements that the Bushman proper was smaller than his

present descendant and a dirty yellow in colour, had no fixed habita-

tion, and was extremely difficult to approach. Their huts were

mere temporary shelters, made of branches in the same way that the

modern Bushmen make a bivouac when on trek.

It may be noted that these descriptions as regards the physical

appearance apply well to the old Bushman, “‘Jantje,” who died in

1924 in Prieska, at the estimated age of one hundred years, and of

whom there is a cast in the South African Museum.

Remnants of these purer people are still to be found scattered in

the Kalahari, and are also difficult to approach; in fact, it seems

impossible to obtain any intercourse with them whatever. Other-

wise they might reveal to us what the pure Bushman really is.

Culturally, the Sandfontein Bushmen are probably as pure as, or

purer than, any of the other tribes. Even they, however, have been

somewhat corrupted by the Bantu and other peoples in the same

districts.

Rock Gravines (Plate XXII).

Rock gravings have been reported from several localities in South

West Africa. The South African Museum has a very fine series,

consisting of original slabs and, where these could not be removed,

casts. Brief notices of these gravings have been given by the late

Dr. Péringuey in the Annual Reports of the South African Museum

(1919; Ann. Rep. for 1918, p.4; 1920; Ann. Rep: tor 19195 pealiGe

1923, Ann. Rep. for 1922, p. 6), together with reproductions of some

of the unusual types (Ann. Rep. for 1919, pls. 1 and ii).

The first lot of specimens was obtained at Gaub in the Grootfontein

district in 1919. The objects represented are the hoof marks or

spoor of various bucks and of zebras, and human footprints. Most

of them are of about natural size. Thus the zebra spoors are about

80-90 mm. across. The buck spoor vary from 40 mm. up to 140 mm.

in length. The human footprints are small, 75-110 mm. in length

(S.A.M., Reg. Nos. 3828, 3829). All these gravings have been pro-

duced by the method of picking (pointillage) with a pointed implement.

Somewhere in the same neighbourhood a graved elephant had been

reported, but this could not be found. Mr. R. W. E. Tucker, who in

1920 also visited the locality, was likewise unsuccessful.

Preliminary Report on Anthropological Researches in S.W. Africa. 105

Further rock carvings were obtained at Klein Wit Vlei in the

Gobabis district (S.A.M., Reg. No. 3830). These consisted of the

usual representations of bucks (eland, gemsbok, wildebeest), and of

a giraffe; but also of a few buck spoor and human footprints, as at

Gaub. The footprints are larger than those at Gaub. The markings

on the giraffe are represented by scattered spots. There were, in

addition, two snakes, occupying a space about 4 feet long (Plate X XII,

fig. 2.)

These carvings have also been produced by the picking method,

but are not of a very high standard of execution. As in other places,

Drury noticed that some of the carvings appeared to be of more

recent date, and to have been executed, not by picking but by line-

drawing, as if some stone or bit of iron had been used to form grooves.

It appears probable that these recent carvings have been done by

herd-boys and others in imitation of the older carvings.

Reproductions of human footprints have been found in other

localities (e.g. Vryburg, Saltpan, Schweizer Reneke), but the buck

spoor appears to be peculiar to South West Africa.

Various explanations have been put forward as to the purpose of

these hoof marks. Péringuey (loc. cit., 1920, p. 16) quotes a state-

ment by Th. Hahn to the effect that the Bushmen had signs to denote

places where they had found water or game, but he is uncertain of

the nature of these signs, whether they were paintings or gravings.

It has also been stated in popular journalistic accounts that these

spoors point in the direction where water is to be found. This latter

explanation is quite fanciful and is positively disproved by Drury’s

observations. At Gaub and Schweizer Reneke these carvings occur

on rocks actually in the (dry) watercourse, and without any attempt

at definite orientation. It seems more probable—as these particular

carvings are only found in, or in the neighbourhood of, watercourses

or vleis, where the animals would leave many spoors—that they have

been copied from natural imprints.

From the promiscuous manner in which the carvings were

scattered about it appears that the artists worked “just as fancy

took them or space permitted,” and it seems certain that they

took their inspirations from the actual animals that they saw in

the neighbourhood.

The representation of snakes at Klein Wit Vlei is interesting from

the rarity of representations of these animals. Only one other example

is known. This was at Britstown, and covered all sides of a large

boulder, being about 30 feet in length.

VOL. XXIV, PART 2. 8

106 Annals of the South African Museum.

The tortoise found at Klemm Wit Vlei is paralleled by another

example from Klipfontein, near Kimberley, and other localities.

At Otjitambi Mr. Lawrence found representations of gemsbok,

eland, elephant, and giraffe (Plate XXII, fig. 1). The markings

on the latter are indicated by intersecting lines (cf. an elephant from

Vryburg, Péringuey, loc. cit., 1909, pl. vin, fig. 2). The same method

has thus been used to represent, in the one case, colour-markings; in

the other, the wrinkled, corrugated, or mud-caked surface of the skin.

The technique of these carvings is similar (7.e. pointillage) to that

of the Gaub and Gobabis examples.

It seems impossible to gauge the antiquity of these carvings.

Péringuey (1909, Trans. 8. Afr. Philos. Soc., xvi, 4, pp. 406 seq.) has

discussed this question. Certain it is that the carvings antedate

the present tribes of Bushmen. Drury emphasises that the local

Bushmen in the Gobabis district know nothing whatever about

these carvings or their origin, expressing blank surprise when cross-

examined on this point. Other observers have testified to the same

ignorance, though herd-boys have frequently been observed to make

crude imitations.

The occurrence of these rock gravings in the river beds in several

localities, as e.g. Gaub and Schweizer Reneke (see also Péringuey,

1906, Trans. 8. Afr. Philos. Soc., xvi, 4, p. 409), raises an interesting

point in connection with the volume of water in these rivers in the

days when the artists were working. Hither the rivers were no

fuller than at the present day or else they must have been inter-

mittent. From the evidences of denudation it is clear that these

Kalahari rivers at one time carried an enormously greater volume of

water than nowadays.

A question of greater import concerns the geographical distribution

of these rock gravings and the race or races who sculptured them.

In North Africa similar sculpturings have been known for some

while. To the references given by Péringuey (loc. cit., 1906) may be

added Frobenius and Obermaier (Hadschra Maktuba: Urzeitliche

Felsbilder Kleinafrikas, 1925; Zelizko, Felsgravierungen der Siidafri-

kanischen Buschmanner, 1925).

Further localities have recently been brought to light by the travels

of Hassanein Bey at Owenat in the South Soudan (Geogr. Journ.,

Ixiv, 5, p. 353, pls., 1924). A noteworthy feature of these carvings

is the absence of any representation of the camel, indicating that

they were executed before the introduction of this animal into the

country.

Preluminary Report on Anthropological Researches in S.W. Africa. 107

Péringuey (loc. cit., 1909) mentions a locality in Nigeria which

forms to some extent a link between the North African and South

African localities, and in 1906 (loc. cit., p. 412) he argued in favour

of unity of race. A detailed comparison between the carvings of

Northern and Southern Africa might lead to important results.

All these carvings have been found in open situations, mostly

with a wide view over the surrounding country. They are carved

on large immovable rocks projecting out of the ground. In the

Transvaal, on the other hand, smaller boulders have been utilised,

some of them quite easily moved. At Klein Wit Vlei, although

there were numbers of small boulders suitable and handy, the artists

preferred large expanses of rock, like tables, almost on the ground

level. One exceptional case was observed where carvings had been

executed on the floor of a rocky overhang or shelter at Gaub.

108 Annals of the South African Museum.

EXPLANATION OF PLATES.

Puate XII.

Figs. 1 and 2. Naron huts at Sandfontein, South West Africa, showing method

of construction, contents (bedding, ostrich egg-shells), and fireplace outside. The

lower figure shows the wooden pestle and mortar.

Puate XIII.

Fig. 1. Group of unmarried Naron men, under their shelter.

. 2. Two Naron girls in their hut, isolated previous to marriage.

PLatTeE XIV.

Fig. 1. Naron woman and child, to show dress, coiffure, and ornaments.

Fig. 2. Naron girl in dancing attitude, photographed from the cast in the South

African Museum. Shows dress, ornaments, powder puff, and *‘vanity bag.”

PuatTe XY.

Figs. 1 and 2. Photographs of two casts of Auen women in the South African

Museum, illustrating method of making ostrich egg-shell beads.

The lower figure (2) shows a woman biting off pieces of egg-shell. The upper

figure shows the process of drilling the beads, and the two stones which are used to

reduce the drilled beads to a roughly circular shape.

Prats XVI.

Fig. 1. Head of Auen woman.

Fig. 2. Head of Naron man.

Puate XVII.

Group of Naron men.

Puate XVIII.

A group of Naron girls and women. ‘The centre figure is a married woman, the

others are unmarried.

PuaTE XIX.

An incident in the eland dance. The two centre figures are men representing

the bulls; the five figures on the left are women representing the cows. On the

right, “orchestra”? and audience.

PuaTE XX.

(Selections from the cinematograph films taken by Mr. Drury in 1921.)

Figs. 1-5. Manufacture of string. In fig. 1 the man is removing the flesh from

a Sanseviera leaf with the aid of a gemsbok horn which has been sharpened to

Preliminary Report on Anthropological Researches in S.W. Africa. 109

form a blade. 2. Preparing the fibre by moistening it in the mouth. 3. Adding

a fresh piece of fibre to a length already twisted. 4. Twisting the fibre by rubbing

it on the right thigh. 5. The finished article.

Fig. 6. Braying a skin.

Fig. 7. Boring a soapstone pipe.

Puate XXI.

(Selections from the cinematograph films taken by Mr. Drury in 1921.)

Figs. 1 and 2. Front and side views of boy playing the ‘‘ Rhumkie,”’ showing

position of hands and fingers.

Fig. 3. The war game.

Fig. 4. The ostrich dance.

Fig. 5. The porcupine dance. Two incidents are shown.

Figs. 6 and 7. The reed dance.

fo)

Puate XXII

Fig. 1. Giraffe engraved on rock at Otjitambi, Outjo district.

Fig. 2. Snakes and buck spoor engraved on rock at Klein Wit Vlei, Gobabis

district. Photographed from the cast in the South African Museum.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XII.

Photos: J. Drury, 1921. Neill & Co., Ltd.

SANDFONTEIN BUSH PEOPLE.

Ann. S. Afr. Mus., Vol. XXIV. Plate XIII.

Photos : J. Drury, 1921.

Neill & Co., Lid.

SANDFONTEIN BUSH PEOPLE.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XIV.

Photos: J. Drury, 1921. Neill & Co., Lid.

SANDFONTEIN BUSH PEOPLE.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XV.

Photos: J. Drury. Neill & Co., Ltd.

SANDFONTEIN BUSH PEOPLE.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XVI.

Photos : J. Drury, 1921. Neill & Co., Ltd.

SANDFONTEIN BUSH PEOPLE.

Ann. S. Afr. Mus., Vol. XXIV. Plate XVII.

Photos : J. Drury, 1921. Neill & Co., Ltd.

SANDFONTEIN .BUSH PEOPLE.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XVIII.

rire et

Photo: J. Drury, 1921. Neill & Co., Ltd.

SANDFONTEIN BUSH PEOPLE.

CTX.

Plate

Ann. 8. Afr. Mus., Vol. XXIV.

Nee

BL LEO OTS.

Co., Lid.

Neill &

Photo: J. Drury, 1921.

BUSH PEOPLE.

SANDFONTEIN

Ann. S. Afr. Mus., Vol. XXIV.

Photos :

J. Drury, 1921.

SANDFONTEIN

BUSH PEOPLE.

Plate XX.

ll & Co., Ltd.

Ann. S. Afr. Mus., Vol. XXIV.

Photos: J. Drury, 1921.

SANDFONTEIN BUSH PEOPLE.

Plate X XI.

Neill & Co., Ltd.

Ann. S. Afr. Mus., Vol. XXIV. Plate X XII.

Photo: R. F. Lawrence, 1925.

Photo: J. Drury. Neill & Co., Ltd.

ROCK CARVINGS, SOUTH WEST AFRICA.

PRINTED FOR ‘THE

: eve ‘NEILL AND £05 JRE

or” THE SOUTH AFRICAN MUSEUM, CAPE TOWN

(111)

6. Archaeology of the Cape St. Blaize Cave and Raised Beach, Mossel

Bay.—By A. J. H. Goopwin, M.A., F.R.S.8.Af., and B. D.

Matan, B.A., University of Cape Town.

(With Plate XXIII and 7 Text-figures.)

Part I.—HIsToRIcau.

Tue Schulpegat, as it was called, at Cape St. Blaize, has had a very

considerable history. It seems at one time to have been a show cave,

and many of the early South African travellers have visited and com-

mented upon it. Barrow * gives an account of the cave up to which

he scrambled from the shore below, and “‘disturbed some thousands

of sea-birds, and found as many thousands of living shellfish scattered

on the surface of a heap of shells that, for aught I know, would have

filled as many thousand waggons. The presence of shells therefore

in my opinion is no argument for the presence of the sea.”

It was this passage which inspired Lichtenstein’s visit a few years

later.t “‘ Ascending a steep and narrow path with the high towering

cliff on one side, and a deep precipice on the other, we arrive on a

sudden at the arched entrance of a cavern, the bottom of which is

entirely overspread with muscle-shells. The breadth of the cave

is about twenty paces, its depth about half as much, and its height

at the centre of the arch may be fifty. The sea is 400 feet below.”

(The actual height is 95 feet.) . . . ““We did not find here the least

traces of any sea-fowl . . . not one of the shells besides appeared

the least fresh; and all were in a state of decay and half buried in

sand and dirt. . . . (Barrow) has, besides, been rather guilty of

exaggeration when he says, the quantity of empty shells in the

muscle-cave is so great that several thousands of waggons would be

necessary to carry them away. Ten or twelve would be quite

sufficient for the purpose.’{ “The common opinion among the

* Travels into the Interior of Africa, London, edition of 1801, vol. ii, p. 67.

+ Travels in Southern Africa, first English edition, London, 1810-12. Van

Riebeeck Society edition, Cape Town, 1928-29, vol. i, p. 219.

{ The translator of Lichtenstein here flatly denies that Barrow ever made such

a statement.

VOL. XXIV, PART 3 8a

112 Annals of the South African Museum.

sensible inhabitants of the neighbourhood, and of the postholder

himself, is, that the cave was formerly a common resort of the

Hottentots, and that they lived very much upon the shellfish, which

are taken here in abundance.”

The history of our knowledge of the Mossel Bay types of implement

has similarly been a long one, and in its beginnings relatively clear.

At the South African Museum there is a fairly large and comprehensive

collection, mainly from the Cape St. Blaize Cave, extending over a

long period of years. Unhappily the material collected has not been

typical, as early enthusiasts were interested chiefly in those conven-

tionalised forms which could be described by such terms as “‘arrow-

heads and flaying knives.”

In about the year 1888, Mr. George Leith, having had his interest

aroused by Dr. D. R. Kannemeyer,* excavated a portion of the cave,

and his account he published ten years later.t He here explains

how the quarrying of the stone immediately previous to the former

date rendered the cave more accessible, so much so that the local

gardeners had sifted and removed much of the cave-earth for manure.

It would seem that Leith had little choice, and examined the refuse

from their sievings before turning his attention to what remained of

the undisturbed parts of the floor-deposit.

“In the undisturbed part of the deposit I found very fine skinning

knives of quartzite, some of them 5 or 6 inches long, and as thin as

a shilling. The number of perfectly formed implements, however,

bore a very small proportion to the number of broken implements

and mismakes, certainly not one to a thousand. The prevailing

type of implement was the knife-like or leaf-shaped flake implement,

with two cutting edges, terminating in a sharp point opposite the

bulb of percussion. These may be classified into arrow-tips, javelin-

points, club-spikes, and half a dozen more varieties, according as they

vary in size and shape, but I regard such classifications as more or less

fanciful and tentative. . . . Early in my cave-hunting experiences

at Mossel Bay, I happened to find an implement of indurated shale,

and a close examination showed its edge so carefully and regularly

notched that I concluded that it was a saw. ... I may mention

another type that I came across that was not figured in the books

* See Kannemeyer, “Stone Implements,” Cape Illustrated Magazine, i, 1890,

p. 120.

+ Journal of the Anthropological Institute, vol. xxxviii [N.S., vol. i], 1898,

pp. 260-262, part of “The Caves, Shellmounds, and Stone Implements of South

Africa,”

Archaeology of Cape St. Blaize Cave and Raised Beach. 113

at my disposal, but which I am told is known as ‘bone-splitters’

(sic). The implement I refer to is shaped like one of the bevel-edged

chisels used by cabinetmakers. The cutting edge is a straight line

at right angles to the direction of the blow which struck the flake from

the core; in other words, the line of concussion.”

It will be noted that Leith, for his time, understood the subject,

and described three types of implement, the point in a great variety

of shapes and sizes, the saw or perhaps an “‘oak-leaf” type, and also

what would appear to be some form of tranchet. This last appears

in some of the collections at my disposal, but would seem to be a

normal form of rejected flake. The local enthusiasts at Mossel Bay

state that Leith recovered a skeleton from this cave, but his docu-

mentary evidence is against this, and the only specimen he mentions

is from another site. So far as I can discover, no skeletal remains

have ever been recovered from here, either from the overlying midden

or from the lower deposits, but in excavation we found a few minute

fragments of human bones in the disturbed layer.

Leith’s implement collection seems to have been divided between

various Museums, the Transvaal Museum at Pretoria receiving the

major part. The South African Museum, the National Museum at

Bloemfontein, and perhaps also the British Museum, have smaller

collections.

The difficulties he met with seem to have precluded any attempt to

note stratification. A number of illustrations submitted by Leith with

his original paper were unhappily never published, though they are re-

ferred to in the text, and were on view at the Anthropological Institute.

During the final year of last century, Drs. Rogers and Schwartz

of the Cape of Good Hope Geological Survey submitted a number of

points from a site “near the Cape St. Blaize lighthouse” (S.A.M., 148),

while later in the same year further material “from the cave below

the lighthouse” was presented by Mr. G. 8. Corstorphine (S.A.M.,

157). These are the spoils from a later search at the cave which is

dealt with in the present paper.

In 1904 Mr. J. Proctor presented the Museum with another series,

this time from a site “situated on the coast between the Gouritz

River and Mossel Bay” (S.A.M., 506). In the following year similar

material was presented by Mr. J. Bain, from various “caves near

the St. Blaize lighthouse.’ It is quite possible that this material

too was a product of Leith’s work, as Bain had been collecting for a

number of years, and in 1905 his entire collection became the property

of the South African Museum.

114 Annals of the South African Museum.

The type therefore had a very considerable history when Dr.

Péringuey wrote his work on the “Stone Ages of South Africa.” *

In this work he describes and illustrates a number of these specimens,

and relegates them to his “Chelleo-Moustierian” period, with the

added rider that “these flakes are doubtless Moustierian in shape.”

While he suggests that the type is “coeval with the boucher,”’ he

admits that, “‘ with the exceptions of Cuts 7 and 12, they were not found

in connection with bouchers, although these paleoliths do occur in

the same district.”

He further describes the typical tool as a “knife or scraper” or as a

“scraper-knife,’ and notes that in some cases “it bears at the base

a dent that looks like a notch.” Reference to the plates indicated

shows that this is intended to be a description of the very common

appearance of a typical step-flake on the mid-line of the outer face,

and not, as might be thought, of a notch on the edge of the flake.

It will be seen that Péringuey realised the importance of his material,

recognised it as a “‘ Moustierian” industry, described it rather loosely,

illustrated it, but did not specifically name it. He would also appear

to have differentiated, though not at all clearly, between this material

and the specimens depicted in his Cuts 106 and 107, which latter

material I have elsewhere described as of the Pietersburg Variation

and Still Bay refuse, respectively. In his subsequent arrangement

of museum cases at this museum the Mossel Bay material was very

sharply differentiated from the other superficially similar forms.

As a direct result of this publication Mr. Hall presented the South

African Museum with a further series (S.A.M., 1755) of points from

the original Cape St. Blaize Cave. A rather different collection of

material from this site is housed at the Albany Museum, Grahams-

town, and presents non-typical Wilton forms, presumably from the

surface.

This was the position of our knowledge when Dr. E. C. N. van

Hoepen of the National Museum, Bloemfontein, published a paper

on the age of South African stone implements,j which included a

redescription of Mossel Bay material in Afrikaans, without illustra-

tion and with no reference to Leith’s publication though it is a

description of that portion of Leith’s material which is housed in the

National Museum. In this article he formally names it the Mossel

Bay Culture. The description seems to be partly taken from

* Annals of the South African Museum, vol. viii, 1912.

+ Ouderdom van die Suid-Afrikaanse Klipwerktuie, S. Afr. Journ. Sci., 1926, vol.

Xxlli, p. 797.

Archaeology of Cape St. Blaize Cave and Raised Beach. 115

Péringuey, and fails to add or mention many of the most important

technical criteria whose value he seems not to have understood. I|

have translated the relevant parts of the paper, and append the

description.

“The National Museum has a number of long thin scrapers from

Mossel Bay, presented by Geo. Leith. They are probably from the

local caves. The scrapers are generally long; they are formed by

the removal of three flakes from a core, after which the scraper is

so struck off that the new cleavage face is parallel with the middle

flake scar of the first three. Generally the midmost of these three

flake scars does not run right to the point. The edges of these imple-

ments show no secondary working at all. There are no bouchers

found with these implements. . . . It would be helpful to name the

culture presented by the long scrapers described above with a South

African name, and I propose to name them the Mossel Bay Culture.

It compares most closely with the Magdalenian Culture of Europe.”

This last allocation he revokes, and substitutes the Aurignacian

Culture, in a later paper.*

Whilst this description of the Mossel Bay “‘scrapers”’ (skrapers)

covers the Mossel Bay points fairly adequately, it is extraordinarily

wide in its interpretation, more especially as it is not accompanied

by any illustration. How inclusive and adaptable the description

is may be seen in the later publication quoted, in which the entire

paper (nominally on the Mossel Bay Culture) is devoted to a descrip-

tion of valuable material in indurated shale from the Modder River,

which should perhaps more properly be included in the Glen Grey

or some similar culture,f and does not agree with any material so

far known from the Mossel Bay district.

In 1929 the present writer included a fuller but crudely illustrated

description of the Mossel Bay material in a paper on the Middle

Stone Age, in the Annals of the South African Museum,{ while a

section on typology was also added to two papers published in the

South African Journal of Science a few years later. The descrip-

tions in these latter papers are certainly the most comprehensive

and detailed so far published, though both the earlier papers are

marred by a failure to acknowledge and refer to the previously

* Argeologiese Navorsing, No. 4, 1932.

t+ See Goodwin, Annals of the South African Museum, vol. xxvii, 1929.

t Loc. cit., pp. 135-140, pl. xxii.

§ “The Chronology of the Mossel Bay Industry,” S. Afr. Journ. Sci., vol. xxvii,

1930, and “The Cape Flats Complex,” S. Afr. Journ. Sci., vol. xxx, 1933.

116 Annals of the South African Museum.

published descriptions of Messrs. Leith, Péringuey, and van Hoepen,

omissions for which I must apologise. The fact that the papers of

all these authors were easily available and widely known would seem

a sufficient guarantee that the writer had no intention of appropriat-

ing the honours which accrue to Leith’s priority of description.

While rightly and courteously drawing attention to the omission

of any reference to his previous paper, Dr. van Hoepen has at the

same time attacked me for having included the Mossel Bay as a

‘Variation,’ within the Middle Stone Age. Reasons for this inclusion,

and for having changed the denomination from a “culture” to a

“variation” have been sufficiently clearly stated in the “Chronology

of the Mossel Bay Industry,” and in “The Cape Flats Complex”

mentioned above, also in my Presidential Address to the South

African Association for the Advancement of Science, Section KH,

1931, while they are further supported in the body of this present

paper.

Part I].—Tae Cave.

Introductory.

The Cape St. Blaize Cave or shelter is a semi-dome-shaped cavern

facing south-east. The cave measures 28 m. (90 feet) across the

extreme lips of the mouth, by 13 m. (40 feet) from front to back,

and the present height of the cave roof above the deposit level is

some 8m. It is situated at a height of 30 m. (95 feet) above sea-

level, at the rocky point of Cape St. Blaize, the top of which is an

extension of the 300-foot peneplain above the town of Mossel Bay.

On a shelf 200 feet above the sea, and directly over the cave, stands

the Mossel Bay hghthouse.

A hundred metres to the north of the cave is a fine exposure of a

raised beach (Plate XXIII). The terrace, starting at present some

45 feet above sea-level, forms a wide common, shelving down to the

present storm beach. Both the cave and the beach are rich pre-

historic fields, and both show large numbers of stone implements of

Mossel Bay type, belonging to the South African Middle Stone Age.

The relation of this beach exposure to the well-known 20-foot beaches

has hitherto been uncertain, but a quarter of a mile away caves cut

by the 20-foot level are observable.* The St. Blaize Cave does not

seem to bear any relationship whatsoever to the beach, nor is it

* Compare Dreyer, “The Stratification of the Superficial Deposits at Mossel

Bay,” Trans. Roy. Soc. 8. Afr., vol. xxii, 1934, pp. 165-169.

Archaeology of Cape

MAIN TRENCH.

Distursep -° -

‘MUDDEN - _ :

St. Blaize Cave and Raised Beach. 117

\y BARE Rock

FLOOR _

—- NO

ase i

i) es:

ans’ exe

eee

9°

CAPE S’ BLAIZE CAVE.

MOSSEL Bay.

[ee eee ee —ee YT,

a= SECTIONS OF MAIN TRENCH =

ON FRONT MID-LINE OF CAPE.

3ST. BLAIZE CAVE

OLD TRENCH.

El, BLACK EARTH.

Fic. 1.—Cape

BROWN EARTH.

St. Blaize Cave.—Plan of cave and sections of main trench,

118 Annals of the South African Museum.

necessarily a sea-cut cave. It would seem to have been primarily

due to a lack of conformity between the horizontally bedded Tertiary

deposits of Enon conglomerate and the steeply dipping Table Moun-

tain sandstone. The Enon has here formed a thick layer of quartzitic

sandstone, which is macroscopically discernible from the Table

Mountain sandstone by its extraordinarily rich botanical fossil

content. This deposit overlies the Table Mountain sandstone,

which here dips at an angle of some 70° and strikes north and south

to form reefs parallel to the coast at this point. The area between

two such reefs forms the natural bathing-pool known as the Poort.

To understand some of the stratigraphical difficulties entailed in

dealing with this cave, Leith’s account of the relevant history of the

cave is of value.*

“When first I visited the cave some years previously (7.e. previous

to 1888) the floor was level. The dark soil of sand and wood ashes

was littered with the shells of large edible molluscs, fragments of

stone from the roof, flakes of quartzite, broken bones, and sea-worn

pebbles. It could only be reached by scrambling up a steep talus

of rubbish at the risk of slipping downwards into deep water or from

the top by a still more dangerous route. . . . The cave, by the way,

is described by the travellers, Lichtenstein and Sir John Barrow (both

cited above). . . . On my second visit to the Lighthouse cave I

found that a great change had taken place. So much of the rock

had been quarried for building purposes that access to the cave

had become easy—so easy, in fact, that farmers and townspeople

had sifted out a great deal of the cave-earth and carried it away

for manure to their gardens. Heaps of the siftings, consisting chiefly

of quartzite chips, shells, bones, and fragments from the roof, were

lying in the cave, while cartloads had been thrown over the edge

into the sea. Here and there, however, especially near the outer

edge of the semicircle, considerable masses of deposit were still

undisturbed.”

This haphazard digging for guano and garden fertilizer must thus

have been the cause of the extremely disturbed state in which we

found the greater part of the deposit, and accounts also for the number

of excellent implements still to be retrieved from the overlying dis-

turbed and crushed midden deposit. If any further excavation is

to be attempted in this cave it would seem that the best method to

apply would be to sieve the entire deposit once more, removing all

material as excavated, and by this means reaching the undisturbed

* Loc. cit., p. 260.

Archaeology of Cape St. Blaize Cave and Raised Beach. 119

layers below, and eventually lowering the floor-level of the cave to

bedrock. A number of permissions would have to be obtained and

the work undertaken by an experienced excavator with extreme care.

Course of the Excavation.

Permission having been obtained from the Minister of the Interior,

the Minister of Railways, and the Mossel Bay Town Council, work

was started. on Wednesday, 6th July 1932, by A. J. H. Goodwin

and B. D. Malan.

The screens used consisted of a rectangular sieve of six holes to

the inch, and two riddles of three holes and two holes to the inch

respectively. The sieve, after careful testing, was found to be un-

necessarily fine for the deposit and the material collected, and was

eventually discarded. All material found was passed through the

two riddles.

Excavation was commenced midway across the mouth of the cave

(fig. 1) and a trench dug towards the back of the cave from a slight

step formed by the swing of the drip-line from the cave lip, some

5 feet from the present talus edge. As it was conditional that we

return all earth excavated to the trench after we had finished, the

debris was sorted and stacked on this shelf, and eventually returned.

At this point the top one-foot layer consisted of a heavily compressed

mixture of shell, nacre, black charcoal, and white ash, with a con-

siderable admixture of dark soil and implements of Mossel Bay type,

while also scattered here and there were found the epiphyses of

fragmentary human bones. It was noticed also that many of the

shells in the deposit were perpendicular, and not horizontal, as they

would normally have been if deposited over a long period of time.

These facts were discernible throughout the whole of this layer, and

we were led to expect that the entire cave had been excavated

previously, and the material from the rear of the cave thrown forward

to cover the front portion of the deposit. A careful examination

of the rear of the cave confirmed this suspicion, as the wall showed

disintegration, oxidation, and an almost complete absence of lichen

to a height of some 5 feet, and it was deduced that the original midden

floor (before this previous excavation) had dipped fairly steeply

from the back of the cave towards the front, the subsequent levelling

being the result of excavating the rear of the cave and throwing

the debris forward and over the edge. It was afterwards realised

that this debris consisted of a mixture of various layers; a midden

VOL. XXIV, PART 3. 8a*

120 Annals of the South African Museum.

layer some 3 feet in thickness, and the dark earthy layer underlying

the midden and containing Mossel Bay types of implement. It is

therefore impossible to reconstruct the true depths from the original

surface of the implements found, and the levels given are relative to

the disturbed present-day surface of the midden as we encountered it.

The removal of the material for garden manure, and the subsequent

excavation of part of the remainder by Mr. George Leith, is a sufficient

explanation of the condition of the deposit.

Considerable numbers of implements of Mossel Bay type and also

implements of midden type were found in this debris which we call

layer A. No microlithic implements were found.

Below the level of this disturbed midden (1 foot below present cave

surface) was encountered a small lenticular layer (layer B) some 3 feet

in diameter and 8 inches thick at the centre, which contrasted strongly

with the overlying midden (see section, fig. 1). Here the shells

were horizontal and in most cases whole and in good condition.

The interjacent material consisted of yellow sand and fish-bones in

a good state of preservation and apparently containing a large per-

centage of their original fats. No carbon or white ash was discernible,

and it became apparent that we had here met with a small section

of the original midden deposit entirely undisturbed. Whereas the

overlying debris was crushed and heavily compressed, and contained

shell finely broken as though thrown out from the more heavily

populated part of the cave, the undisturbed midden layer B was very

loose and contained only whole shells. The finding of a small frag-

ment of a modern glass tumbler in the disturbed midden immediately

overlying this confirmed our earlier conclusion that this surface

had been exposed within the European period. Whereas layer A

had contained material from all depths of the deposit, layer B was

completely sterile of all implements of any conventional type though

specimens of the usual chipped pebbles of water-worn origin, typical

of coastal middens, were found. Neither microlithic nor Mossel

Bay types of implement were found in the undisturbed midden,

layer B.

Immediately beneath this layer was encountered a heavy black,

sandy soil, composed entirely of sand and vegetable matter. The

shell ceased abruptly and completely, and no single fragment of

shell, trace of lime, nor of fish-bones (which might have suggested

that the cave-dwellers of this period were fish-eaters), were found

below this level. There is no suggestion that the inhabitants ate,

caught, or collected fish until the period of the heavy midden deposit

Archaeology of Cape St. Blaize Cave and Raised Beach. 121

(layer B) is reached.* The only animal remains encountered

in layer C were an unidentifiable buck, and a tooth identified

by Dr. S. H. Haughton as of Phacochoerus dreyeri (Dreyer and

Lyle). The undisturbed black deposit yielded a very large number

of flakes and implements in Table Mountain sandstone of Mossel

Bay type.

Throughout the remaining depth of the cave, work was hampered

by the embedded fallings from the cave roof, weighing up to 200 lbs.

As a result the implements found were in many cases broken and the

stratification within this level so disturbed that any attempts to

differentiate between minor variations in strata were here impossible.

The implements were, however, easily recognisable as of Mossel Bay

form from both the whole and the broken implements found. The

relative positions of the blocks of fallen stone prove that at this point

roof scalings had been falling throughout the whole period of habita-

tion. Implements were found beneath boulders, between them

andabovethem. Ifit were possible to regard the rate of accumulation

of these fallings as constant, then the black sandy soil would represent

a far longer period of occupation than the thicker overlying midden.

The apparent discrepancy between the actual volume of material

in each layer is easily accounted for by the tremendous shell content

of the midden deposit.

Towards the bottom of the layer the colour of the deposit lightened

to a nut-brown fairly suddenly, but with no marked division. It is

within the limits of possibility that the darker colour of the upper and

later period was due to the presence of a greater amount of vegetation

at the cave mouth, and that the nut-brown portion was the product

of a relatively drier period. The analysis of the chemicals composing

the layer does not give evidence in any way conclusive on this point

(see analysis of specimens | and 2).

An attempt was first made to differentiate between the black

and the brown layers, but the fallen debris had so disturbed

the material contained that no differentiation was here possible.

A comparison between implement types showed no _ distinct

change.

Below this level a large expanse of rock was found forming a hori-

zontal surface. This would not seem to have been the original cave

floor, but no edge to the block was reached during excavation and no

attempt could therefore be made to discover whether the cave had

been inhabited previous to this huge fall, for such it would appear

* Compare Dreyer, op. cit.

VOL. XXIV, PART 3. 9

122 Annals of the South African Museum.

to have been. The rock is of Enon lying horizontally, and is appar-

ently an immense solid falling from the roof. Geological indications

would suggest that the original floor was necessarily composed of the

steeply dipping Table Mountain sandstone which here underlies the

Knon.

At one point a heavily cemented concretion was found between

layer C and this rock surface. It had originated from C, but had been

strongly concreted by a natural cement. Embedded firmly in this

deposit were a number of implements identical with those from the

bottom of layer C. The concretion, as exposed, was found to be

circular, partly extending under the untouched deposit, and was

subsequently found to be a cone of cemented material effected by a

drip from a point of rock above.

The work in this trench was continued for a distance of 6 feet

towards the back of the cave, the layers remaining the same and

horizontal throughout save for the small lenticular fragment of un-

disturbed midden (layer B) which disappeared.

At this point a sudden change was noted. The top layer (A) dipped

vertically, and we realised that we had come to the transverse trench

dug by the early guano diggers. The infilling was loose, containing

holes and air-pockets, and the midden had throughout the same dark

colour with vertical shell-fragments, as previously observed in layer A.

This infilling was watched as it was encountered, and while no attempt

was made to excavate and sieve the material numbers of excellent

specimens were found. The dark colour of the earth still adhering

to some of these proved that the implements had come originally

from a layer such as the dark earthy layer (C) rather than from the

light sandy undisturbed midden (layer B), and had been incorporated

in their present position by mechanical infilling. At the bottom of

this old trench, proving that the infilling was of recent date, was found

an easily recognisable fragment of tanned leather bearing signs of

machine-stitching. The trench does not appear to have been dug

to bedrock, and the underlying remnants of C may still prove worth

digging.

Before proceeding further with this excavation, a trial trench was

dug from the rear of the cave, with the intention of joining up with

this previous excavation and thereby obtaining a complete section

of the cave from front to back (see Plan, fig. 1).

The rear wall of the cave comes forward as the spring of a half-

dome resting upon a shelf of rock which starts a few feet above the

present floor-level, but well below the original floor. This horizontal

Archaeology of Cape St. Blaize Cave and Raised Beach. 123

rock steps sharply down to disappear beneath the deposit. The step

is formed of layers containing a rich content of fossil wood and wood

impressions. This is obviously a sedimentary deposit—the base of

the Enon probably—and it has formed a number of pockets consisting

of wood-impressions filled with loose yellow sand. It has also given

off pockets of crystals, some yellow, some sulphur coloured, apparently

of gypsum.

In commencing the work from the rear of the cave the surface of

this shelf was searched and a slight accumulation of undisturbed

deposit was removed. A number of specimens of Mossel Bay type,

stained yellow, were found, and among these a single specimen in a

light green surface quartzite, bearing signs of the Howieson’s Poort

technique, was noted. The relationship of this specimen with the

others on this shelf is unknown, but the shelf would appear to have

been some 2 or 3 feet below the original surface of the cave deposit

before it was first rifled. This trench was abandoned after a day’s

excavation, as it was found that the material throughout this portion

of the cave had already been dug by previous workers, a burnt match

being found at the depth of 2 feet.

A return was now made to our original trench in the centre of the

cave-mouth, and work was continued towards the large rectangular

blocks of stone, fallen from the ceiling, which lie on the cave floor.

The rock-bottom below the trench had so far remained level. On

digging farther it was found that the rock dipped slightly from the

edge of the concretion, towards the east, to form a slight hollow. The

basin thus formed had filled with a clay-like sandy, damp deposit,

free of all implements. Comparison with modern rock pools in the

cave shows that the precipitate at the bottom of these is similar to

that found in this sandy-clay layer, which thus represents an old pool,

completely sterile. The nature of the lower portion of layer C is

here better revealed, the concretion is obviously part of a cone

cemented by drip action, and C at this point immediately overlies

the sandy clay.

Up to this point the fallen rocks from the roof, throughout layer C,

had rendered impossible any attempt to differentiate between minor

variations in culture in this layer. Now, for an area covering a

square yard and rich in implements, the deposit was found to be

free of rock-falls, and variation could be noted. It was observed

that the implements resolved themselves into two layers of maximum

density, separated by layers of relative scarcity. Layer C was now

divided into four.

124 Annals of the South African Museum.

C 1, (top section) not very rich.

C 2, extremely rich.

C 3, relatively poor.

C 4, relatively rich, but small.

This last division coincides with the change from black soil to brown

at the base of layer C.

In the course of excavating layer C, a number of flakes and worked

implements of surface quartzite and quartz crystals were found, all

showing workmanship resembling that of the implements from

Howieson’s Poort rock-shelter.* An attempt was now made to

localise this type more exactly within layer C. They were found to

belong to layer C 2; that is, they form an intrusion after the main

period of the Mossel Bay habitation of this cave, and before the end.

They do not form a separate sub-layer within C 2, but appear to be

thinly present throughout. Later, with such flakes, was found a

single lance-head of Howieson’s Poort type, im situ within this layer.

It was an inch and a half in length, worked over the whole of one face,

to a slight keel, the cleavage face being left unworked. It resembles

exactly the Howieson’s Poort material (see fig. 5, 4).

This bears out the suggestion implied by the finding of a specimen

of the same type on the shelf at the rear of the cave, and shows that

the Howieson’s Poort Variation at this site cut into the Mossel Bay

Variation. This fact becomes more important when we realise that

evidence at Peer’s cave (Skildegat) (Cape Peninsula) proves that this

same Howieson’s Poort Variation cuts into the Still Bay layers at

that site. We may presume, therefore, that the Mossel Bay Variation

at the type site, and the Still Bay Culture at the Skildegat site, were

at least partly contemporaneous. This presumption would appear

further to support the classification made previously; f in which

the Mossel Bay Industry is regarded as a Variation of the Still Bay

Culture, and not as constituting a separate and distinct culture.

Work in the cave ceased on Wednesday, 20th July 1932.

Soil Analysis.

The following is the analysis of the organic matter in three samples

of soil from layer C. They are: (1) Specimen of black soil, C1. (2)

Specimen of brown layer, C 3. (3) From clayey soil underlying C 4.

* See Stapleton and Hewitt, S. Afr. Journ. Sci., 1927, vol. xxiv, pp. 547-587;

and 1928, vol. xxv, pp. 399-409. Also Goodwin, Ann. 8. Afr. Mus., vol. xxvii,

1929, pp. 130-134.

+ Goodwin, ‘“‘“The Middle Stone Age,” Ann. 8. Afr. Mus., vol. xxvii, p. 139.

Archaeology of Cape St. Blaize Cave and Raised Beach. 125

Fic. 2.—a, flake core; 6, point; c, Levallois flake; d, burin. Implements

from layer C 1.

126 Annals of the South African Museum.

Soils from Cape St. Blaize Cave.

Il, mes, 33.

Per cent. Per cent. Per cent.

Moisture . 2-04 3:12 5-09

Loss on ignition 2-65 2-44 2°38

Organic N, 0-072 0-063 0-047

a Je Oe". 3-4 5-45 5-4.

Wes, EGO) 0-077 0-089 0-098

CO, ‘ 0-15 0-24 0-24

Aue CRO ] 3°85 7-84 6:56

W.S. Cl, as NaCl 0-44 0-65 0-95

12-61 19-83 20-718

Acid-insoluble silicates, etc. 87-39 80-17 79-282

The matter of probable organic origin in specimens 2 and 3

constitutes some 20 per cent. of the whole, while in specimen | it is

represented by only 12 per cent. It may eventually be possible to

relate this organic change with climatic factors, or with the drop

in sea-level or with both.

Analysis and Description of Artefacts.

The following is an analysis of the various types of implements

found in the different layers of the major layer C, grouped by sub-

Cie | CO | GS | CL Nes) Bane

Total of specimens . | 50 75 115 5O 287 88

Flake cores. ; of € 3 2 17 »

Points . : s 26 46 48 31 149 47

Worked points : 1 1 3 0 of 4

Blades . ; é 13 12 22 10 42 a

Burins . Wyo 3 ) 2 -3

Disc scrapers il 0 1 0 0 0

Round scrapers 0) ) 2 0 2 0

Disc cores 4 2 5 1 8 0

Disc flakes 1 1 10 4 14 0

Oak leaves : 0 0 3 0 10 3

Discoidal artefacts 2 4 12 1 23 12

Tranchets : i) 4 2 i) if 0

Redirecting flakes . 0) 0) 0 1 12 4

Archaeology of Cape St. Blaize Cave and Raised Beach. 127

Fic. 3.—1, Levallois core; 2, lance head of Howieson’s Poort type;

3, chalcedony scraper; 4, burin. All from layer C2.

layers and compared with the Slang River material. This shows

the amount of homogeneity to be expected within a culture, and also

reveals certain interesting discrepancies of a minor order.

The Slang River material is taken from a previous paper,* but

* Goodwin, S. Afr. Journ. Sci., vol. xxvii, 1930, op. cit., p. 572.

128 Annals of the South African Museum.

we have omitted the twelve nondescript flakes collected there, as

we have similarly omitted a large number of nondescript flakes from

this present site.

Descriptions —In many cases no adequate description of the imple-

ment types is here given, and readers are referred to the general

descriptions given under the same terms in any good book on

prehistory.

Material.—The material used in the Mossel Bay Industry is a

fine-grained Table Mountain quartzitic sandstone, obtained locally,

but in some few instances a surface quartzite, a quartz, or quartz-

crystalis used. The Howieson’s Poort types are generally of the sur-

face quartzite, though one or two are of each of the other materials.

It is not possible to differentiate completely between the debris

rejected from these two industries.

Flake Cores.—The specimens found were generally completely used.

The striking platform was in most instances prepared in such a manner

that the flake produced would have a curved butt, and that the flake

scars would converge slightly to produce a pointed flake (fig. 2).

Flakes.—These are of two types, long flakes, with slightly converg-

ent fluting, struck from flake cores; and circular flakes, with radially

converging flake scars, struck from a Levallois or disc core.

Points.—These are pointed flakes, without secondary retouch and

with no further preparation after their removal from the flake core.

They show the cleavage scars of two or sometimes three slightly

convergent longitudinal trimming flakes. The third and central scar

is short and sometimes stepped, with the result that the bounding

ridge between the two outer and major flake scars extends down to.

the point forming a shght keel. The butt is generally, though not

always, faceted, showing either a simple curve or a “Cupid’s bow”

curve (figs. 2, b, 4, 1, 2, and fig. 5, 6).

Worked Points.—The same as points, but with secondary trimming

along both edges from the under face, which leaves resolved flake

scars on the edge of the outer face.

Oak Leaves.—Points which have been trimmed by the removal of

deep secondary flakes at more or less regular intervals along the edges,

in such a manner as to leave a series of denticles between the scars

(figs; 2):

Blades.—These are really the same as the points save that the

central scar runs the whole length of the flake, dividing and separating

the two lateral flake scars. It is possible that both the point and the

blade are rejects and that they merely represent normal spalls pro-

Archaeology of Cape St. Blaize Cave and Raised Beach. 129

duced by the technique employed. This does not invalidate their

inclusion and description, as normal products of a technique are basic

to its study.

Burins.—These are generally prepared on a flake by the removal of

a sliver from the forward end, down one edge, in such a manner that a

transverse chisel edge is left at the intersection (figs. 2, d, and 3, 4).

Redirecting Flakes and Refacing Flakes.—The striking platform of

the flake core is not always faceted, and in some cases a triangular

sectioned flake is found which seems to have been the product of a

lateral blow struck in such a manner as to remove the old striking

platform.

Side Scrapers.—Exactly similar to the Huropean form, especially

as seen at La Quina and Le Moustier (figs. 4, 5, and 5, r).

Discoidal Artefacts —Levallois or dise cores (unstruck), pebbles

écaillées, “‘pseudo-coups-de-poing,” etc. These have been grouped

together, as it is often impossible to differentiate between them with

certainty. Certain artefacts are immediately placeable, but many

more might fall into either two or even three groups. They are all

generally lenticular in shape, having been formed by the removal

of a number of radially converging flakes struck from about an

equator.

Disc Scrapers.—Two specimens were found of small discs with a

worked edge about the perimeter.

Round Scrapers.—Probably belong to the side-scraper class. The

work is on a flake, and roughly resembles the circular scraper of the

Smithfield.

Levallois or Disc Cores (struck).—These are very similar to the

Somme specimens, but are smaller, the flake removed measuring

perhaps a little more than an inch in diameter (2-5 em.). The strik-

ing platform is sometimes plain (figs. 3, 1, and 5, 5).

Levallois Flakes (struck from dise cores).—None of these show any

sign of secondary retouch. The butt (fig. 5, 6) is generally faceted,

but sometimes plain (fig. 2, c), agreeing with the striking platform on

the nucleus.

Tranchet.—This has only been separated as a type, as early writers

refer to it as a “‘bone-splitter.”’ It consists of a wide, short, parallel-

flake, with no signs of secondary working or usage.

Howieson’s Poort Types.—These have been sufficiently well described

by Stapleton and Hewitt (see footnote, page 124), and no description

is necessary here. But compare figs. 3, 2, 5, 3 and 4, with their

illustrations.

130 Annals of the South African Museum.

During the course of the excavation it had been thought that

certain changes could be discerned between layers C4 and Cl.

Fic. 4.—1 and 2, points; 3, worked blade; 4, worked chert fragment;

5, side scraper; 6, point. 1 to 4 from layer C 3, 5 and 6 from layer C 4.

Layer C 4 seemed to show a number of long flakes of fine workman-

ship, and a proportionately large number of wide, flat flakes, lanceo-

Archaeology of Cape St. Blaize Cave and Raised Beach. 131

late in form, all made from a fine-grained Table Mountain sandstone.

C 1 appeared to show smaller implements and a less fastidious choice

of material. With the intention of further analysing and defining

this development careful measurements were taken of all points (the

total numbers of cases are given above), with the result that no real

change could be postulated with certainty.

In spite of the fact that we believe the majority of the points and

the parallel flakes or blades to have been in the nature of rejects,

yet they are normal products of the technique employed and hence

are more likely to give us some idea of any change in skill or desir-

able forms than a study of the few available “finished” (7.e. re-

touched) implements.

Length

of Cl C2 C 3. C 4 C (gen.).

points

Max. . | 10-0 cm.| 14:0 em.| 15:1 cm.| 11-0 cm.} 10-0 cm.

Aver. . Bos) 5. 6:49 ,, Po 5 8-38 ,, io)

Min. . BO 54 BIA0) i ASO! es AO PAID) ahs

Widths of

points.

Mae: 5-5 em. 5-4 em. 5-3 ecm. 4-4 em. 6-5 cm.

Aver. . 30) 5 By gs 3°46 ,, Soil 2:54 ,,

Min. ; eB). ap esa ies: ilo | Re Ope 28

It will be noted that there is a change from 6-5 cm. in the average

length in layer C 1 to 8-38 cm. in layer C 4, but no great weight can

be attached to this change, as the flakes in layer C 1 were often dug

out of the concretion, and possibly concealed small material may have

escaped our notice. It is therefore necessary to withhold judgment

until further analysable material is at our disposal.

Part III.—TuHe Ratsep Beacu.

Mossel Bay war memorial stands on a slope which shelves up from

the sea, where it 1s edged by Table Mountain sandstone reefs, to a

height of some 45 feet, forming a wide sloping common. Behind

the memorial a road has been cut into this shelf at its highest point,

and reveals a continuous section of an old raised beach for some

thirty metres, standing about 6 feet above road-level at its highest

point. Above this cutting the land surface rises sharply, and is

132 Annals of the South African Museum.

y >A

BEAZLAY

NAL:

Zo) Lig)

SLL

Fic. 5.—1, Side scraper; 2, oak leaf in quartz; 3, concave scraper (Howieson’s

Poort): 4, lance-head (Howieson’s Poort); 5, Levallois core; 6, ‘‘Cupid’s

bow” type of butt to show effect on convergence of fluting. All from

layer C, general.

Archaeology of Cape St. Blaize Cave and Raised Beach. 133

formed by the steeply sloping talus from the 300-foot peneplain

behind Mossel Bay. The peneplain is composed of Tertiary deposits

of Enon conglomerate which have here formed a quartzitic sandstone,

horizontally bedded and lying unconformably on the steeply dipping

Table Mountain sandstone.

Dr. A. V. Krige refers to this raised beach, but does not seem to

have studied this cutting, which may not have been made at the time

of his visit.* He suggests that “the most recent event in the coast-

line history is that of shght emergence in two distinct stages. We

recognise a major emergence, from nil to 80 feet, followed by a minor

emergence between the extreme limits of 15 and 25 feet. This has

brought the strandlines of the former emergence to altitudes of from

20 to 100 feet, averaging from 50 to 60 feet.”

Discussing the evidence for the major emergence he quotes the raised

beach at Cape St. Blaize as an example: “ Problematical evidence

is afforded by the caves at Cape St. Blaize. The larger cave below

the lighthouse is 90 feet above sea-level and the smaller ones are

about 110 feet. These caves may be of subaerial origin, as they occur

in the base of the Enon sandstones, where these overlie the T.M.S.

unconformably. To the north of Cape St. Blaize a little terrace rises

to only 45 feet at the foot of the higher escarpment.”

Returning to his discussion of the terrace he later states that it

has been cited as evidence of the major emergence. “This is not

necessarily so, on account of the probability that it may have been

built up to this level by rain-wash from the cliffs. Rock floor is

seen round the edge to a height of only 15 feet. It is possible that

the terrace was cut in part during the higher stand, and subsequently

modified during the lower stand” (p. 56). The cutting made for the

road, apparently subsequent to Dr. Krige’s visit, precludes the

hypothesis that the shelf has been built up by rain-wash from the

steeply dipping talus slope above. The beach deposit is obviously

undisturbed and shows a normal history up to the time when the

surface was covered to about a foot in depth by scree. Farther to

the north-west the scree entirely replaces the beach at this height.

It stands to reason that the section exposed must either coincide

at its base with the base of the beach or stand above that level. If

the surface of the beach is 45 feet high, and the exposure cuts down

to a depth of 8 feet, then the base of the cutting must stand at a

maximum height of 37 feet above sea-level.

* A. V. Krige, “Examination of the Tertiary and Quaternary Changes of Sea-

level in South Africa,” Annals of the University of Stellenbosch, v, A.1.

134 Annals of the South African Museum.

Most helpful is the statement of the Clerk of Works to the Mossel

Bay Municipality, that the builder of the war memorial was still

finding beach material at a depth of 16 feet below surface (5 m.).

The memorial stands about 10 m. away (farther down from the slope

of the terrace) from the foot of the exposure, and is probably below

the storm portion of the old beach. We can thus safely subtract

this 16 feet from the 37 feet of the road-level, making 21 feet the

maximum height of the base of the old beach above present sea-

level. This coincides more nearly with the 20-foot beach described

by Rogers at Little Brak,* which certainly belongs to the minor

emergence, and with which Tomlin } has associated an extinct type,

Calliostoma mosselense Tomlin.

The beach exposed by the road cutting at Cape St. Blaize is easily

recognisable, and is composed of large spherical boulders perhaps up

to 18 inches in diameter, originating from the talus but since com-

pletely rolled by sea-action. These are packed with smaller pebbles,

while here and there are small patches of sand, and scattered through-

out are a considerable number of recognisable shells. No species

not usual to this part of the coast were noted. The broken-down

fragments of other shells are plainly visible, and many have gone to

form a lime which has been carried to the surface of the beach by

water action to form a distinct surface, which dips towards the sea

at an angle here of about 45°. The lime formation suggests that the

beach surface was exposed for some time before the encroachment of

the talus rubble, while the steepness of the angle would further suggest

that we are dealing with the highest or storm portion of the beach.

Below this visible surface the lime has loosely cemented the beach

material, rendering the entire deposit a dirty white. The cemented

beach is stratified more or less horizontally, but follows the contours

of the large boulders and the old land-surface.

Overlying the lime crust the soil changes very considerably, and

there is a marked difference discernible between the original beach

material and the overlying rubble. The soil in the latter consists

entirely of reddy brown sandy material a foot in depth (obviously

the product of a dry climate) containing humus and angular frag-

ments of rock from the Enon. Here and there, as a result of erosion

and subsequent infilling, the talus material has cut into the lime

surface of the beach to a depth of 2 or 3 feet, but at these points the

disturbance is obvious and is always easily definable.

* Rogers, Geological Survey of the Cape of Good Hope (Tenth Report).

+ Trans. Geol. Soc. 8. Afr., vol. xxviii, 1925.

‘popfor-uou “yIsodep Yoveq posted OAOGR WOIF 9109 STOT[VAST Yonaysun “q+ po[or ‘qisodop Yovoq posted WO, 0100 SIO[[VAVT YonsysuN “Y—'Y “OTT

136 Annals of the South African Museum.

We are thus presented with two fortuitously associated layers;

the beach deposit, exposed at this section, and varying in height

above road-level from a few centimetres to a little over two metres;

overlying this is a foot of hill talus, necessarily later as a deposit,

and containing material of all ages. The importance of this section

is considerably increased by the fact that it occurs in a locality

extremely rich in caves, and therefore in stone implements, many of

which have at all times been included in the talus material. The

beach has obviously been fed, at least partially, from this rubble,

while the lime crust limits the beach admirably both in time and

space. It stands to reason, therefore, that any rolled implements

from the talus slope included beneath the lime and in the beach

must necessarily have been manufactured before the recession of the

sea (or the land uplift), and also before the formation of the lime

crust.

The finding of implements attributable to man in beach deposits

of a similar type, at a level of 20 feet sea-level at Little Brak

River, has been reported by Rogers.* Péringuey, quoting Rogers’s

report, adds that the implement found in the quaternary lme-

stone of the raised beach at Little Brak “is a fragment of a

knife-scraper flake, in shape and composition corresponding with

examples found on the surface in the Mossel Bay district, especially

at Cape St. Blaize.”+ The implement thus belongs to the Mossel

Bay type of stone implement. Goodwin reported the finding of

similar types immediately overlying this same beach, but not from

within the beach material itself.t

Little human material has so far been found in the St. Blaize

beach (37-foot exposure), and during the present visit a careful search

was made for datable material, which might be related to a definite

prehistoric period and used as a datum line in South African chron-

ology. Asa result of this search a number of humanly made flakes

were found, and one apparent cowp-de-poing, all considerably rolled.

The flakes are generally formless, and might be attributed either

to any Earlier or Middle Stone Age culture. The butts are crudely

faceted, and in form they resemble the Nooitgedacht material de-

scribed by Péringuey (op. cit., pl. xin, 105), which has since proved

to be a collection of spalls from the Stellenbosch Industry. None

show, on account of their rolled state, any further definable form.

* Op. cit., p. 293.

+ Péringuey, Ann. S. Afr. Mus., vol. viii, p. 80.

t “Chronology of the Mossel Bay Industry,” S. Afr. Journ. Sci., vol. xxvii, 1930.

Archaeology of Cape St. Blaize Cave and Raised Beach. 137

The coup-de-poing would seem to be a more easily identifiable

fossil, and we identify it completely with the pseudo-coup-de-poings,

perhaps unstruck disc cores, described in association with the Mossel

Bay Industry both here and at Slang River.* Several specimens of

this type from the talus overlying the raised beach were also found

associated with Mossel Bay type at the site at present under discussion.

This implement was found cemented into the beach deposit and

in a position which it could not have reached as a result of a recent

fall from the talus (see Plate XXIII). It was heavily rolled, was

more than half enclosed in beach deposit when found, lay directly

below unbroken lime surface, and was completely surrounded by

shells, pebbles, sand, and similar authentic and undisturbed beach

material. It is therefore to be regarded as an integral part of the

raised beach. In order to ascertain whether it would be possible

for talus material to be incorporated in the beach deposit from above,

a careful search was made for non-rolled implements from the per-

pendicular beach section. Some few were found, but all lay in

positions easily reached in their fall from above, and were all caught

in the surface, but neither buried nor cemented in. Rolled artefacts

were in all cases attached, and partially buried into the lime.

We may, therefore, take it as proved that the Mossel Bay Industry

is, at least in part, contemporaneous with or older than the raised

beach exposed at Mossel Bay. It had previously been proved that

this industry is partially contemporaneous with or older than the

20-foot raised beach at Little Brak, on the evidence of Rogers.

It remains, therefore, to suggest the relationship of the St. Blaize

beach with that at Little Brak. Rogers states that the beach at

Little Brak lies some 15 feet above the alluvial plain and about 20

feet above sea-level. Krige (quoted above) states that the terrace

at St. Blaize rises to 45 feet above sea-level. The cut in the beach

exposing the section totals 8 feet, and does not reach the bottom

of the beach deposit. We were voluntarily informed that the builder

who contracted for the war memorial dug to a depth of 16 feet, and

was still finding sea-shells, ime and water-worn pebbles. We may

take it, therefore, that the beach deposit is 16 feet thick from the

present road-surface, and that we are dealing with a beach the base

of which is not more than 21 feet above sea-level. This brings the

St. Blaize beach within the scope of the minor emergence described

by Krige, and not, as he suspected, with the major emergence.

The Mossel Bay Industry is therefore partially contemporaneous

* Goodwin, S. Afr. Journ. Sci., vol. xxvii, op. cit.

VOLA XV PART oe 10

138 Annals of the South African Museum.

with, or previous to, the minor elevation marked round the southern

South African coast by the “twenty-foot”’ beaches.

Further research in raised beaches, together with our previous

knowledge, reveals other features of interest, which all point to the

validity of the association. At Shelly Beach, Port Elizabeth, a shelf

covered by old dunes and standing to a height of up to 50 feet has

been left by the Minor Emergence. Beneath the hardened sand-

dunes which form the surface are to be found a number of imple-

ments of Mossel Bay type. They are often slightly waterworn, but

no further proof of their associability with the Minor Emergence is

forthcoming. The section of 20-foot raised beach exposed by the

road cutting behind the boat-building yard at Humewood, Port

Elizabeth, is completely sterile of implements.

On a recent search in a railway cutting, a hundred yards east of

the Bitou railway bridge, near Fairey Knowe, George district, a

fragment of raised beach was found, and incorporated in it were a

few water-worn implements of Mossel Bay type. The deposit is

too much disturbed to be conclusive in itself. The height above sea-

level is a little over 20 feet.

Goodwin * has already pointed out that Middle Stone Age sites

near the coast are all situated above the raised beach level, and it

should be further noted that they are generally situated ammedzately

above that level, but never below. The recent contour survey of

the Cape Peninsula (Trigonometr. Survey, Cape Peninsula, Sheet 2,

1934) has shown that the Noordhoek lagoon or vlei, which would

seem to be a remnant of the raised beach, is situated just below the

25-foot contour. This becomes important when we realise that on

the seaward shore of the vlei and at the 20-foot level is an extensive

Still Bay site, which les beneath midden material containing Wilton

implements.+

CONCLUSIONS.

From the Cave.

1. That the Mossel Bay Industry precedes the midden cultures of

the Southern Littoral, and is certainly not a midden culture itself.

2. That the Howieson’s Poort Variation cuts into the Mossel Bay at

this site, much asit cuts into the Still Bay Culture at the Skildegat site.

* “Chronology of the Mossel Bay Industry,” 8. Afr. Journ. Sci., vol. xxvii,

1930, p. 570.

+ See Haughton, 8. H., The Geology of Capetown and adjoining Country.

Explanation of Sheet 247 (Cape Town). Geological Survey. Pretoria, 1933._

159

Archaeology of Cape St. Blaize Cave and Raised Beach.

Fic. 7.—a, worked point; 6, Levallois flake; c, small disc core. From

bove the raised beach, non-rolled

140 Annals of the South African Museum.

3. That we may therefore presume that the Mossel Bay Industry

is at least partly contemporaneous with the Still Bay Culture, of

which it is a Variation, due to a choice of materials used.

From the Raised Beach.

4. That the Mossel Bay Variation is partly contemporaneous with

the Minor Emergence of Krige.

5. That the only extinct marine fauna so far associable with the

raised beach (at Little Brak) is Calliostoma mosselense Tomlin. The

type is described from that source.

From the Cave and previously known Sources.

6. That the land fauna associable with the Mossel Bay Variation

is in part extinct, and included Hquus capensis and Phacocoerus

dreyeri.

ACKNOWLEDGMENTS.

Our thanks are due to the South African Railways and to the

Municipality of Mossel Bay for permission to excavate, also to the

University of Cape Town for financial help.

Ann. S. J Plate X XIII.

sruck Levallois core in situ (cf. text-fig. 7, a).

Neill & Co., Ltd.

Ann. 8. Afr, Mus., Vol. XXIV. Plate XXIII.

Exposure of raised beach deposit at Cape St. Blaize, showing position of cave. The area outlined in black is enlarged in the left-hand figure, and shows an unstruck Levallois core in situ (ef. text-fig. 7, a).

Neill & Co., Ltd,

( 141 )

7. Some Native Snuff-Boxes in the South African Museum.

By Miss M. Suaw, B.A., Assistant in Ethnology.

(Plates XXIV-XXXI, and 2 Text-figures.)

SNUFF-TAKING is a widely-spread practice among the Bantu tribes

of South Africa, both men and women alike being “passionately

fond” of it (9, 10, 18, 15, p. 237; 22, p. 683; 27, p. 179; and 28).

Tobacco and snuff were among the most important objects of barter

and gift carried by the early travellers. The Herero are a notable

exception, and they learned the use of tobacco from the Nama

Hottentots, who, like the Bushmen, do not grind it to make snuff (6).

Tobacco is believed to have been introduced by the Portuguese

in the sixteenth century (26); and it is possible that snuff-taking was

introduced at the same time, though the fact that powdered dagga

leaves are frequently used instead of tobacco suggests that the practice

might have been known beforehand. Even more than smoking it is

the accompaniment of nearly all social intercourse, but whereas

smoking always remains a more or less individual affair, snuff-taking,

at least among the south-eastern tribes, has reached such a height of

formality that it has become a social usage in itself (25, p. 223).

There is little variation throughout the tribes in the method of

preparing snuff. Tobacco is used more frequently than dagga and

is grown about the huts (8, pl. vi,; 10, 15, p. 237; 28, and 27, p. 180),

excepting by the Ovambo who import theirs from the north (28).

The dried leaves are ground to a powder between two stones, usually

with the addition of the white ashes of burnt twigs, or dried aloe

leaves or some other sharp ingredient, to increase the strength and

quantity of the snuff (12, 18). It is kept in bulk in large calabashes

which are hung up in the huts. The Suto tribes of the Northern

Transvaal damp the snuff after the first grinding and allow it to dry

in blocks which can be used for trade. When wanted for use a portion

is broken off and ground again, and the powder is put into a snuff-

box (9).

As to the manner of taking snuff, although it seems to have

attracted the attention of travellers in most parts of the country,

142 Annals of the South African Museum.

the only adequate description available is for the South-Hastern

Bantu. The Zulu, both men and women, all carry a snufi-box

(25, p. 8), and those who can afford it, a small snufi-spoon carved out

of ivory. They take snuff squatting on the ground; in a circle if

there are several of them. Having first cleansed the nostrils with an

iron spatula, which serves as a handkerchief, they pour a certain

amount of snuff into the left hand and inhale it, either directly from

the palm of the hand, or by taking a large pinch with finger and

thumb, or by scooping it up with a spoon. After a short pause the

thumbs are drawn down the sides of the nose from the eyes, as if

to make a channel for the tears which soon gather and overflow.

This is the height of enjoyment and apparently the crux of the per-

formance, which is terminated by a loud sneeze (7, 18, 25, pp. 8-9).

According to Delegorgue, if any of the company was so hardened that

the snuff did not make him sneeze, he tickled his nose with a few

blades of straw so as not to be outdone by the others (7). Shooter

states that “they seldom meet without indulging in a concert of

snuffing’’ (25, p. 223), and it would be regarded as very rude, if not

as a definite insult, to interrupt them (7,18). If there is an important

person among them, or one richer than the others, he is expected to

supply snuff for the party, but he must be asked for it once or twice

and show reluctance to comply. If he seemed over-eager, or offered

it without being asked, he might be suspected of evil motives; because

in order to take snuff a man must necessarily lay down his arms and

place himself in a defenceless position, and it is reported that the

Zulu despots frequently used this means of treachery to get rid of

people whom they found undesirable. It is considered polite to

pour a quantity of snuff into the left hand and pass it round without

looking to see how much the others take, or better still to hand over

the box, and let the others help themselves (25, p. 223).

The Xosa and other tribes of the Eastern Province and the Basuto-

land Border take snuff in much the same way as the Zulu, and make

as great a business of it. All work is set aside if there is snuff to be

taken. According to Fleming both sexes take snuff, but Kropf,

writing much later, says that it is only the men. They shout with

joy when the tears flow and go on till they are quite stupefied. When

they sneeze they say “May the Chiefs bless me” (17). Most of them

use bone, ivory, wooden, or less frequently iron spoons, with which

first to clean the nostrils, then to carry the snuff to them, and finally,

to scrape off and return to the box any particles of snuff remaining

round the edges. One nostril is held with the finger while the other

Some Native Snuff-Boxes in the South African Museum. 143

is provoked. Fleming mentions a small brush that is used to give

the nose a final tidying (10 compare 7, p. 355). He also describes an

iron point with which to stir up the snuff when it has become caked

in the box, but the pointed end of the snuff-spoon or the finger-nail

might be used for the same purpose (11, 18, 21). Those who can

afford neither spoon nor box rub the snuff on to a very hairy piece

of skin, which is carried rolled up and applied directly to the nose (12).

The Basuto of Basutoland use spoons when they can, or even the

ordinary iron sweat-scraper (8, 28). The Transvaal Suto inhale

directly from the palm of the hand, or if they do not possess snuff-

boxes, from a piece of skin, in the same way as mentioned above.

They too take great enjoyment in snuffing-parties and a good flow

of tears. Grutzner gives an account of the part played by snuff in

the betrothal ceremony of the Northern Suto. When the boy’s

father has obtained the girl’s father’s consent to the match, and has

returned to tell the boy, a messenger is sent back to the girl’s kraal

to ask for snuff. The old women grind some and put it into a gourd

snuff-box which is sent two or three days later, with a member of the

gitl’s kraal, to the boy’s. The boy’s family and relatives are then

called together to take the bridal snuff. The box is opened by the

husband (or future husband if she is as yet only betrothed) of the

boy’s eldest sister. He opens it with his awl, shakes some into his

left hand, and passes the box on, to go the round of the whole family.

Next day the empty box is sent with the first instalment of cattle to

the girl’s kraal, and is given to the girl, who covers it with beads and

wears it round her neck as a sign of betrothal. She calls it her child,

Noana, and removes it only after she has borne her first child, when

she makes a necklace of the beads and hangs them round the child’s

neck (9, 14).

The Bechuana do not appear to use spoons, but “pour a large

quantity of snuff into the palm of the hand and draw the whole of

it up the nostrils at once’ (4). In Burchell’s time the southern

Bechuana tribes obtained their tobacco by barter from their northern

neighbours. The leaves had been prepared by steeping them in

hot water, which made them much milder, so that the colonial brand

was preferred.

The Matabele are inveterate snuff-takers (8), but with less formality.

Unlike the South-Eastern Bantu, among whom snufi-taking makes

any other occupation impossible, one writer reports of them that the

women were seated “weaving mats, snuffing, and chatting” (27,

p- 214). Apparently they have no prejudice against offering snuff,

144 Annals of the South African Museum.

as a young man may try to win the love of a girl by giving her a pinch

of snuff in which a special medicine has been mixed (27, p. 278).

Among the Bathonga of Portuguese Hast Africa if two men meet

they will always offer each other a pinch of snuff and thereby

establish a friendly atmosphere before starting a discussion. In the

same way snuff is taken at peace conferences as a sign of goodwill.

In the final part of the marriage ceremony of one of the clans the

bride kneels to offer the bridegroom snuff, but when he has approached

and taken a little between his fingers, she throws the rest in his eyes

and runs away, pursued by his friends, who must catch her before

she reaches a certain point, or else further presents must be given

to persuade her finally to return to the village (16).

Another example of snuff being used ceremonially is reported by

Bent, of the Makalanga in the south-west of Southern Rhodesia.

He says that whenever they enter a new territory “they kneel before

a tree and burn snuff, saying as they do so, ‘Muali, we have brought

knives, give us meat.’ Then they do the same at another tree, asking

the same for their children” (2, p. 271).

The snuffi-boxes are carried either attached to a necklet, tucked

in the girdle or attached to it by a strap, or ina small bag. The Zulu,

Matebele, the Natal tribes now in the Eastern Cape Province (Fingo,

Hlubi, etc.), and some of the Mashona tribes, use a small piece of

reed, decorated and stopped at each end, which they carry in a hole

made in the lobe of the ear. The spoons are frequently attached to

the boxes, or else stuck in the hair, or carried somewhere else on the

person, as an ornament (1, 2, 8, p. 20; 10,15, pp. 237-238; 18, 20,

22, 25, p. 8; 27, p-179). |

ZULU.

The earliest types of snuff-box were the reed worn in the ear, of

which the Museum has no specimen, and a small gourd (Lagenaria)

(15, p. 237).

The simplest is a dark brown bottle gourd, unpolished, with a round

hole at the top, fitted with a round of gourd shell (S.A.M., 287 (*)).

Height, 4? inches (upper bowl, 1% inch; lower bowl, 2% inches).

Diameter of upper bowl, 1? inch.

Diameter of lower bowl, 2? inches.

Diameter of neck, 14 inch.

Diameter of mouth, 2 inch.

This type would be carried in a small bag attached to the waistband.

Plate XXVI, fig. 2, shows an example of the same type (S.A.M.,

Some Native Snuff-Boxes in the South African Museum. 145

2870 (°)), though smaller and lighter in colour, with a leather stopper

attached to a thong round the neck, by which the box could be

suspended.

Height, 4 inches (upper bowl, 14 inch; lower bowl, 24 inches).

Diameter of upper bowl, 12 inch.

Diameter of lower bowl, 24 inches.

Diameter of neck, | inch.

Diameter of mouth, 2 inch.

Another very popular type has hitherto been described as a small

round gourd (28, p. 8), but appears rather to be the shell of a tree

fruit, probably a Strychnos. It has a round hole at the top for the

mouth, and next to it two smaller holes through which a loop of

twisted grass 1s passed and knotted at each end. It can be attached

to the waistband or necklet.

Plate XXVI, fig. 5, shows an example (8.A.M., 2870 (*)) of this

type suspended from a necklet of small blue and white beads, with a

centre-piece and ends of three large red tubular beads. The fasten-

ing is at the back.

Diameter of shell, 24 inches.

Diameter of mouth, 3 inch.

Length of loop, 3? inches.

Another example of the same type (S.A.M., 2870 (?)) has a short

thong loop, instead of a grass one, to attach it to a necklet which

fastens in front. The bowl is decorated with scratched figures illus-

trated in the panel, Plate X XVI, fig. 6.

Chiefs and rich men who could afford a lot of snuff used larger

gourds of the same type (Lagenaria).

An outstanding example is illustrated on Plate XXIV (S.A.M.,

3322). Fig. 1 is an outline of the gourd itself, fig. 2 the design on

the upper bowl, and fig. 3 the design on the lower bowl. The gourd

is said to have belonged to Dingaan and to have been decorated by

a Zulu who had visited Cape Town, to give him an idea of European

civilization. It is ight brown and highly polished, and the engraving

appears to have been done with a sharp point before the shell was dry.

The amount of detail shown is quite remarkable. On the lower

bowl there are public buildings and houses showing the stairs inside

with people climbing up them; particular attention has been paid

to the structure of the roofs. There is a tent wagon and an open one,

both outspanned, though the latter has an ox beside it; a carriage

and pair with driver and groom; two trees and a solitary rider; a

barrel of wine; and two processions, one of cavalry brandishing

146 Annals of the South African Museum.

swords with the scabbards at their sides, and the other of infantry,

with fixed bayonets and led by a band! On the upper bowl three

men are out shooting with their dog, and a large flock of birds above

them; in front of each is what might be a dead bird. There is a

grandfather clock standing in a group of unrecognisable furniture;

a man seated on a chair, writing at a desk; and finally what appears

to be a view of the town, showing roofs only, perhaps as seen by the

man through the window.

Height 5? inches (upper bowl, 24 inches; lower bowl, 34 inches).

Diameter of upper bowl, 1% inch.

Diameter of lower bowl, 34 inches (approx.).

Diameter of neck, 12 inch.

Plate XXV, fig. 2, shows another example (S.A.M., 3330). It is

the same light reddy brown with a fairly polished surface, decorated

with rows of triangles, first traced with a sharp point, and then

branded.

Height, 8 inches (upper bowl, 44 inches; lower bowl, 34 inches).

Diameter of upper bowl, 1? inch.

Diameter of lower bowl, 23 inches.

Diameter of neck, 1§ inch.

Diameter of mouth, $ inch.

After their introduction European beads became the most popular

form of decoration for snuff-boxes.

Plate XXVI, fig. 1, shows a plain bottle gourd (S.A.M., 2870 (°)),

covered with a diamond network of beads (about } inch in diameter),

strung on sinew. The top half of the upper bowl is navy blue and

white, and the remainder navy blue, white, and white-eyed red.

Half-way down the lower bowl the network gives way to a long fringe

of shghtly larger beads, mainly sky blue, and ending in white-eyed red,

navy, white, large tubular black, and white. The stopper is missing,

but should be a wad of rag, probably with the upper half beaded.

Height, 4 inches (upper bowl, 2$ inches; lower bowl, 14 inch).

Diameter of upper bowl, 14 inch (approx.).

Diameter of lower bowl, 3 inches (approx.).

Diameter of neck, 1} inch (approx.).

Diameter of mouth, 3 inch.

Length of fringe, 1} inch.

This type, besides serving the twofold purpose of snuff-box and

ornament, rattles splendidly with the least movement.

An entirely different type, possibly a development of the reed

snufi-box, is illustrated on Plate XXV, fig. 3 (S.A.M., 2820 (7)). It

Some Native Snuff-Boxes in the South African Museum. 147

is a cylinder of wood covered with closely set beads strung on sinew.

The actual stringing is diamond fashion, but the beads are set so

closely as to appear in rows. The colours are arranged in spiral

stripes of two beads each—white ($ inch), alternating alternately

with white-eyed red (} inch) and black (slightly smaller). The odd

stripe is china blue (1, inch). The edging at each end is two rows of

apple green and one row of white (,}, inch). The ends are stopped

with rounds of gourd shell, one fixed permanently with two short

bead straps, china blue edged with black (,!, inch), and the other to

open, and attached by one similar strap. Two conical brass buttons

at each end attach the box to a round necklet of cloth covered with

small white beads (4, inch) strung on fibre—a long string wound

round.

Length, 5? inches.

Diameter, 1? inch.

Length of necklet, 28 inches.

Diameter of necklet, $ inch.

Plate X XV, fig. 1, shows an example of a large round gourd used

for storing snuff (S.A.M., 2871). It is light coloured and covered

entirely with a diamond network of beads (about } inch); the points

being two white beads, and each side two china blue beads. There

are two horizontal stripes, one near the mouth and one round the

bowl, made by filling in the sides of the diamonds alternately with

white-eyed red, and black. At the bottom a tassel of four strands

comes from a circle of white-eyed red and black. Each strand is

composed of one white, three dark saxe glass, one pink, three dark

saxe glass, one white (;, inch), one large tubular black, and one white.

The stopper is a wad of rag, the top of which is covered with five

alternate rows of large white and china blue. It is attached to the

bowl by a short string of dark saxe glass, white, black, and pink.

Height, 6 inches.

Greatest diameter, 6 inches.

Diameter of mouth, ? inch.

Length of stopper, 12 inch.

TRANSKEI.

The snuff-boxes of the Transkei differ very little from those used

by the Zulu, except that the reed worn in the ear is only found among

the Fingos and a few others who have migrated from Natal in fairly

recent times (1). The most common form is again the gourd, plain,

decorated by branding, or covered with beads.

148 Annals of the South African Museum.

Xosa.—Plate XXVII, fig. 8, shows a light coloured bottle gourd

(S.A.M., 3328) decorated by branding, the designs having first been

scratched with a sharp point. There is a plain band of branding

round the mouth, and round the lower bowl the figures of a man and

woman in European clothes, discussing a dead fowl; a man and

woman in European clothes walking arm in arm; a man on all fours

being attacked by a dog; a large bird and various incomprehensible

designs. This type would be carried in a bag.

Height, 44 inches (upper bowl, 12 inch; lower bowl, 2% inches).

Diameter of upper bowl, 14 inch.

Diameter of lower bowl, 34 inch.

Diameter of neck, | inch.

Diameter of mouth, } inch.

A less usual and more original type is illustrated on Plate XX VII,

by figs. 1, 2, 3, and 4. These are made when an animal, usually an

ox, has just been killed and skinned. The blood and remaining

tissues that are scraped off the inner side of the skin when cleaning

it are collected and mixed with ochre or powdered clay, and if

necessary a little more blood is added to form a stiff paste. A clay

model of the desired shape is made first and baked in the sun. The

paste is spread over this, and when nearly dry a rough nap is picked

up with a sharp point. When quite dry a round hole is cut for the

opening, and the clay model is picked out. A stopper is fitted, either

of wood or a round of gourd shell. These boxes are carried in a bag,

or hung from the waistband by a short thong (11, 12).

Plate XX VII, fig. 1, shows an example shaped like a bottle-necked

gourd (S.A.M., 3319 (1)). It is a very light neutral shade, and the

nap runs in regular concentric rings, except round the neck and at

the bottom, where it is left plain. It has a round wooden stopper,

the top of which is shaped like an hour-glass.

Height, 34 inches (upper bowl, 1 inch; lower bowl, 2 inches;

stopper, $ inch).

Diameter of upper bowl, 1} inch.

Diameter of lower bowl, 2? inches.

Diameter of neck, ? inch.

Diameter of stopper, +4 inch.

Total length of stopper, 14 inch.

This specimen does not appear to have been used, as the nap is

very sharp and fresh, whereas on two other examples (S.A.M., 3319

(?) and (°)) it is worn almost smooth. No. 3319 (?) has a stopper of

the same material, with the top half covered with white beads.

Some Native Snuff-Boxes in the South African Museum. 149

Plate XXVII, fig. 2 (S.A.M., 3319 (*)), is shaped like a pig. The

opening is in the snout, and is stopped with a round of gourd shell.

The box is reddy brown in colour and appears to have been rubbed

with red ochre. The tail and ears have been left smooth. The nap

is worn rather smooth.

Length, tail to snout, 4 inches.

Greatest height, from ears, 24 inches.

Greatest width, 1? inch.

Plate X XVII, fig. 3 (S.A.M., 58 (1)), represents an ox. Itis pinkish

in colour, and the tail, feet and belly, horns, eyes, ears, and nose

have been left plain. The opening is between the horns.

Length, from tail to horn tips, 6 inches.

Greatest height, at horns, 44 inches.

Least height, at tail, 2 inches.

Greatest width, 24 inches.

Plate XXVII, fig. 4 (S.A.M., 58 (?)), also represents an ox, with

the horns curved down and a hump behind the head. The nap is

in very regular rows, and the colour brownish red. The opening is

below the tail, which with feet and belly, ears, horns, and nose

is left plain.

Length, 44 inches.

Height at head, 27 inches.

Height at tail, 1? inch.

Greatest width, 2 inches.

The Museum has another example of this type (No. 1339), made

to represent a sheep.

Another very common form of snufi-box is made of the pointed

end of a goat- or ox-horn (18). The solid point is usually carved,

and the hole at the base is filled in with a thin disc of wood cut to

fit exactly, and nailed in.

Tembu.—Two Tembu snuff-boxes of this type are illustrated on

Plate XXVII, figs. 5 and 6, and fig. 7.

Figs. 5 and 6 show two views of a black and white horn (S.A.M..,

988 (1)); the bowl of which has been pared down to make the base a

definite and very symmetrical oval. The light-coloured wooden base

is nailed in with iron tacks. The top is carved to represent a buck’s

head, with short ears, long straight horns and a conventionalised

shield-like face, in which the hole for the aperture is cut.

Length along upper curve, 8 inches.

Base, 24 inches by 12 inch.

Oval aperture, 2 inch by } inch.

150 Annals of the South African Museum.

Fig. 7 shows a dark brown horn (S8.A.M., 988 (?)), thinned at one end

to a straight nozzle, divided from the bowl by a flanging ridge. The

wooden base is oval and nailed in with brass tacks. It appears

to have been covered with red ochre. A crack near the base has

been stuffed with a rag. The stopper is a short joint piece of reed.

A leather strap is knotted round the base of the nozzle.

Length along upper curve, with stopper, 6? inches.

Length of nozzle, 12 inch.

Diameter of nozzle, =, inch.

Base, 1% inch by 12 inch.

Length of strap (double), 6 inches.

Width of strap, } inch.

Tambuki.—Plate XXVIII, figs. 3 and 6, shows two beaded gourds

used by the Tambuki of the Cala District, Eastern Province. Both

are light-coloured bottle gourds (S.A.M., 4981 (+) and (?)).

Fig. 3 has only the neck and two-thirds of the lower bowl covered

with a network of small white beads (;4; inch), with a single stripe of

black beads round the middle, and alternate triangles of black and

white round the neck. At the bottom there is a band of four rows

of china blue beads, and below that a short fringe, every two strands

of which are joined in alternate pairs top and bottom, at the top by a

white and two black beads, and at the bottom by a small diamond

of two black and two pink beads. The strands themselves are divided

in the centre by a black bead, the top half being dark jade, and the

lower half china blue, except for a short patch where white-eyed

red take the place of the green. The stopper is a wad of pale blue

cotton material, covered at the top with rows of pale blue and pink

beads. A long bead handle of china blue and white is attached to

the top of the beadwork at each side.

Fig. 6 is entirely covered with beads (1, inch), with the exception

of a small ring round the mouth, and an irregular seven-sided patch

at the bottom. The beads are threaded on black cotton on the upper

bowl, and twisted sinew on the lower. The colours are in vertical

stripes of white, black, jade, white, china blue, black, and white.

The beads are threaded on a long string wound round and round the

gourd and supported at every change of colour in each row by a

vertical strand of wood fibre which is twisted round the thread.

In between these seven vertical columns the strands are free, and in

some places hang fairly loosely. The stopper is a wad of dark blue

cotton, with some small white beads sewn irregularly on top. It is

attached by a short string of alternate black and white beads. There

Some Native Snuff-Boxes in the South African Museum. 151

is a loop handle attached to the top of the beadwork, a single strand

of black, white, red, and yellow beads.

Fig. 3. Fig. 6.

Total height : . 4 inches. 4% inches.

Height of lower bowl . . 2 inches. 23 inches.

Diameter of lower bowl . 3 inches. 34 inches.

Height of upper bowl . . 1 inch. 12 inch.

Diameter of upper bowl . id inch. 12 inch.

Stopper . inch. 4 inch.

Depth of net- Worle : . 12 inch.

Depth of fringe. 2 inch.

Fingo.—A typical product of the i der’s store is the Fingo snuff-

box illustrated on Plate XXVIII, fig. 2 (S.A.M., 969). An empty

vaseline tin holds the snuff. Two holes are Boted in the side, about

1} inch apart, through which there is a short loop of twisted fibre.

The tin is attached to a flat pendant of beadwork by loops of china

blue, black, and white beads, which pass under the fibre loop. The

ground of the pendant is white (4 inch and ;}, inch), with an edging

of yellow (4 inch) and stripes of two rows of coloured beads; from the

bottom, black and red, green and black, blue and black, red and black,

yellow (single), black and pink, red and black, green and black, blue

and black, black and red. At the top there are two groups of six

strands each of white beads, one topped with yellow and black and

the other with blue and black, gathered together and attached each

group to a brass button, by means of which the pendant is buttoned

on to a necklet.

Diameter of tin, 1% inch.

Height of tin, % inch.

Total length of pendant, 6 inches (slightly shortened in illustration).

Width of pendant, 1% inch.

Length of loops, at top, 14 inch, at bottom, | inch.

BasuToLAND BoRDER.

A variety of snufi-boxes comes from the Colonial side of the south

and south-eastern borders of Basutoland.

Plate XXVIII, fig. 5, shows an ordinary bottle gourd (S.A.M., 4932),

covered with a network of pale blue and white beads (4 inch), with a

single row of white round the neck, and a spiral stripe of dark blue,

black, and white-eyed red round the lower bowl. On the upper

bowl the beadwork is a regular diamond pattern, with two white

beads for the points, and a single blue between. On the lower bowl

152 Annals of the South African Museum.

there is a similar arrangement of the white beads, but the blue ones

are alternately in ones and twos, and only in occasional patches on

the bottom half is there an appearance of regularity. The top of

the rag stopper is covered with pale blue and white beads, and there

is a short handle of blue beads interspersed with red, white, and black.

Height, 5 inches (upper bowl, 1$ inch; lower bowl, 3 inches;

stopper $ inch).

Diameter of upper bowl, 2 inches.

Diameter of lower bowl, 33 inches.

Plate X XVI, fig. 3, shows a specimen (S.A.M., 4956) very similar

to the Zulu type on Plate X XV, fig. 3, but considerably smaller. It

is a hollow piece of bamboo covered with rows of beads (;, inch)

threaded on a string of twisted simew and wound round. The colours

are in five stripes, black and red alternate, separated by a single

row of white, and there is an edging of pink and blue at each end.

The box is attached to a short double-bead chain, of white-eyed

red at the bottom, and black at the top, divided at each change of

colour and in the middle of the black by a short flat pattern of blue

and white.

Length of box, 2 inches.

Diameter of box, ¢ inch.

Length of chain, 64 inches.

Plate XXVI, fig. 4 (S.A.M., 4937), is a brass cartridge case covered

with a diamond network of beads (,!; inch), white for the points, with

the sides filled in at the top and bottom with green, and in the centre

half with black and half with pink. It is edged top and bottom with

white-eyed red. The rag stopper is made to look like a head, with

black beads for the hair edged by a row of black and white alternate,

and a frill round the neck of pink, black, green, white, and yellow.

The snuff-box is attached to a single-bead chain of green and pink

not long enough to go round the neck.

Height, 22 inches.

Length of stopper visible, # inch.

Diameter of cartridge case, 2 inch.

Finally, there are horn snuff-boxes similar to those previously

mentioned.

Plate XXVIII, fig. 1, shows a greyish black horn (8.A.M., 4933)

with natural spiral rings. It is round in section and at the bottom

has a convex-faced wooden plug, thick enough to remain in without

nailing. The nozzle is eight-sided and separated by a step from

the bowl. Near the mouth it has three narrow bands of roughly

Some Nate Snuff-Boxes in the South African Museum. 153

plaited brass wire. At the base of the nozzle two thongs are knotted

round the horn, twisted, and knotted together.

Total length over outer curve, 8 inches.

Basal diameter, 1,'5 inch.

Length of nozzle, 24 inches.

Diameter at mouth, ;°; inch.

Diameter at base, | inch.

Plate XXVIII, fig. 4 (S.A.M., 4958), resembles the type found at

its best in Basutoland. It has a short, black, fairly polished bowl,

oval in section, with a wooden base nailed in with three brass tacks,

and a round hole with raised lip near the top in front. It is sur-

mounted by a very conventionalised buck’s head in which the two

horns curve inwards and join in one piece.

Height, 3? inches (bowl, 23 inches; head, 13 inch).

Basal diameters, 13 inch by 14 inch.

Width across horns, 14 inch.

Basuto.

The Basuto use plain or beaded gourds similar to those of the

South-Eastern Bantu, but their most distinctive snufi-boxes were

carved out of horn, an art that is now lost. The horns are black and

highly polished and filled in at the bottom with flat discs of wood as

has been previously described.

The Museum has three specimens, which are illustrated on Plate

XXIX.

Figs. 1 and 2 are two views of No. 448, which represents a man.

His head and neck are the full size of the horn at that point. His

head is turned obliquely and he has a beard. The short nozzle is

in the top of the head and has a raised ring round it. The rest of

his body is entirely disproportionate and is carved in relief on the

back of the horn. Each leg coincides with a natural patch of white.

The wooden base is not nailed.

Height, 5 inches.

Length along outer curve, 5? inches.

Basal diameter, 14} inch.

Diameter of nozzle, >, inch.

Figs. 5 and 6 are two views of a similar snufi-box (S.A.M., 561).

It represents a woman with a child on her back. The skin cloak which

supports the child is shown in the front but not at the back, where

it would naturally cover the child. The woman wears a round girdle

and the old-fashioned skin skirt, long at the back and short at the

VO lie XOX PART oe 1

154 Annals of the South African Museum.

front. Up to that point, though not proportionate, the body is

carved according to the size of the horn, but it is terminated by two

spindly legs, right-angled to show the thighs, and standing out in

relief in the front of the horn. The toes are turned in. The hole is

in the top of the head and has a raised lip. The wooden base is not

nailed in.

Length over outer curve, 6+ inches.

Height, 5} inches.

Basal diameters, 2 inches by 1? inch.

Diameter of mouth, 3 inch.

Figs. 3 and 4 illustrate No. 449, of which only the top part is carved,

and that in the shape of an antelope’s head and neck, with graceful

horns. The hole is in the front near the top and has a raised rim.

The horn is shaped to a sharp edge down the centre of the back and

the round wooden base is nailed in with two brass tacks.

Height, 62 inches.

Length over outer curve, 84 inches.

Diameters of base, 1% inch.

Diameters of nozzle, ;% inch by $ inch.

MASHONALAND.

The snuff-boxes from Southern Rhodesia show much less European

influence than those of the South-Eastern Bantu. Again the gourd

is one of the most common types, but it is less frequently decorated

with beads.

Plate XXX, fig. 8 (S.A.M., 1806), is a plain gourd (Cucumis or.

Citrullus) snuff-box, from the Manyika tribe on the south-eastern

border. It is reddish brown and highly polished, but entirely without

decoration. Its stopper is a small wooden plug, hour-glass shape.

Inside are a quantity of melon seeds, which serve perhaps to sift the

snuff and certainly make a very good rattle.

Height, 1% inch.

Greatest diameter, 34 inches.

Diameter of stopper, 3 inch.

Height of stopper, 3 inch. .

Another type very characteristic of Mashonaland is illustrated

on Plate XXX, figs. 4 and 5 (S.A.M., 2774 and 1807) (2). Both come

from the Manyika tribe and are hollowed out of very hard dark brown

wood, with the design cut in ridges, but whereas in fig. 4 both the

ridges and the ear handle are quadrilateral in section, in fig. 5 they are

triangular. The inside hollow does not appear to be very large.

Some Native Snuff-Boxes in the South African Museum. 155

The twined brass wire round the nozzle of each appears to be European,

but that round the base of fig. 4 is thicker and probably of native

make; it is covered with a black gummy substance.

On fig. 4 the pattern is the same back and front, and the flat back

of the handle has three rows of short parallel ridges. The loop through

the handle is a two-ply twist of black rag.

Height, 9 inches.

Greatest width, 1 inch.

Greatest thickness, 12 inch.

Length of nozzle, 14 inch.

Diameter of nozzle at base, 2 inch; at top, § inch.

Diameter of hole, 5% inch.

Kar projects, $ inch.

On fig. 5 the whole of the back (on each side of the handle) is carved

in longitudinal ridges, and the front is divided by two longitudinal

ridges into two panels of diagonal ridges, both running in the same

direction, but changing it at the bottom. At the base there is a

deeply cut ring with a ridged cone protruding slightly.

Height, 5? inches.

Greatest diameter, 14 inch (the box is round in section).

Length of nozzle, 1,>, inch.

Diameter at base, 2 inch; at top, # inch; of hole, >; inch.

Ear projects, $ inch.

A similar but less finely finished example is shown on Plate XXX,

fig. 3 (S.A.M., 3335). It is hollowed out of hard black wood, round

in section, and carved in longitudinal ridges. The nozzle is plain

and the mouth stopped with a mushroom-shaped light wooden stopper.

A single piece of leather is inserted in a hole in the side and then split

in two, almost as far as the hole, and knotted to form a loop. A small

piece of tin has been hammered in near the base below the loop.

Height, with stopper, 3} inches.

Diameters of base, 3 inch by $ inch.

Greatest diameter of bowl, 14 inch.

Length of nozzle, 2 inch.

Diameter at base, 2 inch.

Diameter of mouth, ;‘; inch.

Diameter of stopper, ;%; inch; height above mouth, 3 inch.

Length of loop, 3$ inches.

The specimens shown on Plate XXX, figs. 1, 2, and 7 (S.A.M., 3318),

were used by the Makalanga of the Victoria District, Southern

Rhodesia. They are made of moth cocoons (Lasiocampidae, probably

156 Annals of the South African Museum.

Gonometra), with the top cut off and fitted with a small stopper of

light wood. The cocoon is attached to a strap of leather or twisted

fibre string, which passes once and is knotted under the ridge made

by the junction of the cocoon and the twig from which it was

broken. Some still have the remains of the twig. The boxes could

be attached to a necklet or worn in the hair.

Fig. 2 is a single cocoon with a leather strap decorated with a large

white-spotted sky-blue bead and divided to form a loop by which

the box might be suspended. The stopper is hour-glass shape.

Total length, 22 inches.

Greatest width, | inch.

Diameter of stopper, upper circle, ;4; inch; lower circle, $ inch.

Height of stopper, $ inch.

Length of strap, 44 inches.

In fig. 1 two cocoons of similar proportions and with conical stoppers

are fastened together on a short fibre string.

In fig. 7 the ends of a short strip of leather, about 24 inches by

jj; inch, are split in two and a cocoon attached to each of the four

corners.

Plate XXX, fig. 6, shows a small leather snuff-box (S.A.M., 1791).

It is a round of leather worked into a conical shape while wet. The

greatest diameter when open is 2$ inches; when closed the length

is 24 inches. The edge is decorated with a double row of beads, in

bands of pink and black alternately (three times). About ? inch

from the top is a leather thong, sewn to the main portion in two

places and threaded through three metal beads, brass, iron, and

copper, which can be slipped up or down to close or open the box.

The rest of the thong forms a loop and is decorated with several

brass beads. Its length when doubled is about 54 imches.. When

closed the edge is inclined to frill.

OVAMBO.

The Museum has three specimens of Ovambo snuff-boxes all quite

modern and unused. They are carved out of light-coloured wood,

probably of a palm, and decorated with incised and branded diagonal

grooves.

Plate XXXI, fig. 6 (S.A.M., 4769), is shaped like a stemmed cup.

A leather strap passes vertically through a projection at the top, and

is held by the insertion of a transverse piece of leather. At the other

end it is divided in two, one end being attached to a round stopper of

gourd shell, and the other to a wooden spatula.

Some Native Snuff-Boxes in the South African Museum. 157

Height of box, 3? inches.

Diameter at greatest width, 1% inch.

Diameter of base, 1 inch.

Diameter of neck, 1,3, inch.

Diameter of mouth, ? inch.

a b

Fig. 1.—Ovambo Snuff-boxes.

Total length of spatula, 4,4, inches.

Length of bowl, 1 inch.

Greatest width, 2 inch.

S.A.M., 5435 (text-fig. 1, a) is a cylinder of wood, hollowed out

and open at one end, which is fitted with a round stopper of gourd

shell. The stopper is attached to a leather strap which is passed

through the wall of the box near the top and knotted. At its base

a short piece is divided from the main strap to attach an iron spatula

which appears to have been made out of a European spirally grooved

file.

158 Annals of the South African Museum.

Height of box, 3 inches.

Diameters, 1 inch by 1,4; inch.

Length of spatula, 2? inches.

The third example (S.A.M., 5436) (text-fig. 1, b) is similar to the

second, but wider, and having a different design in decoration. It

has no spoon or spatula.

Height of box, 3 inches.

Diameters, 1,°,; inch by 12 inch.

The two latter examples are very smoothly finished and look as

if they have been sandpapered; the former shows the marks of a knife.

BECHUANA.

The Bechuana use mostly goat- or buck-horn or bone for their

snuft-boxes (24).

Plate XXXI, figs. 1, 2, and 3, shows three snuff-boxes made of

highly polished horn. They all have the bottom filled in with a

round of wood, which is nailed in with small wooden pegs. The

nozzles are inserted separately and cut to fit the hole in the top of

the horn. The stoppers, which are missing, would probably be wads

of rag.

Fig. 1 (S.A.M., 3344 (7)) is a black horn striped with grey. It has

an oval base and a round ivory nozzle decorated with three vertical

rows of black dots, and with no apparent means of securing it except

exactness of fit. The mouth has a raised lip.

Height of bowl, 2% inches (approx.).

Basal diameters, 2 inches by 1? inch.

Length of nozzle, ¢ inch.

Diameter of nozzle, 2 inch.

Fig. 2 (S.A.M., 3344 (3)) is a light-coloured almost transparent horn

with a round base filled in with a convex piece of wood nailed with

three black wooden pegs. The ivory nozzle is not quite circular

and has three vertical rows of black dots. The mouth has a raised

rim.

Height of bowl, 27% inches.

Basal diameter, 14 inch.

Length of nozzle, inch.

Diameter at base, 1} inch by $ inch; at mouth, $ inch by 3 inch.

Fig. 3 (S.A.M., 3344 (?)) is a round white horn with a scallop decora-

tion in black dots. The nozzle is made of black horn and like the

wooden base is held in place with four small black wooden pegs.

The mouth has a raised lip.

Some Native Snuff-Boxes in the South African Museum. 159

Height of bowl, 2,°, inches.

Diameters of oval base, 142 inch by 14% inch.

Length of nozzle, 14 inch.

Diameter at base, inch; at mouth, ;°, inch.

Plate XX XI, fig. 4, shows a snuff-box (S.A.M., 1686) made of a piece

of bone, decorated with two bands of treble chevron pattern, first

engraved with a sharp point and then filled in with some black

substance. There is a plain band round the bottom, with the initials

in black, M.K.B. The base is triangular and filled in with wood

held by three tiny wooden pegs. The mouth is oval and has a wide

flat rim. The stopper is a piece of hide with a short double leather

tag.

Height, 43 inches.

Length of each side of base, 13 inch.

Top, § inch by 2 inch.

Mouth, 3° inch by 3 inch.

Length of tag, % inch.

Plate XX XI, fig. 5, shows an unidentified specimen (S.A.M., 2876)

which is probably of Bechuana make. It is an unpolished light-

coloured horn, oval in section and slightly twisted. The stopper is

a wad of leather with a long double leather tag. The bowl is quite

plain except at the bottom, where it has been pared away to leave

a slightly raised ring round each of the eleven tiny wooden pegs

which hold the wooden base in place. The outstanding feature is

the base itself (fig. 5,a@), which has been branded in a conventional

pattern after it was fixed in place.

Height, 32 inches.

Basal diameters, 1% inch by 1,% inch.

Diameter of mouth, 5); inch.

Length of tag (doubled), 2,7, inches.

The Museum has few examples of snuffi-spoons. Of those shown

in text-fig. 2, a and b come from East Pondoland, ¢ from the Basuto-

land Border, and d, e, and f from Basutoland. The Zulu spoons are

the same type. Speaking of them Holden says: “In making the

comb part great patience is requisite, as to shave the solid piece of

ivory to the proper size and then saw the long teeth with a rough

piece of iron is no small task” (15, p. 254).

a is made of bone. The cross-hatching on the back of the stem

has been first incised and then filled in with dark red ochre. The

front is quite plain. In cross-section the stem is a segment of a

circle and the prongs rectangular. The bowl projects shghtly forward.

160 Annals of the South African Museum.

Bowl, 1 inch by 2 inch.

Stem, 1} inch by } inch.

Fork, length, 3% inches, width at top, 2 inch, each prong 4 inch.

Thickness, } inch.

bis made of bone. The round dots in front are first incised and

Sop © ® © &® @ wom.

a b c d e fig

Fic. 2.—Snuff-spoons.

then filled in with dark red ochre, as is the cross-hatching on the

back, which stretches from the base of the bow] to the third last dot.

In section the prong is the segment of a circle. The back of the bowl

is in a line with the prong, curving up slightly at the tip.

Bowl, 1 inch by +4 inch.

Prong, 5 inches by } inch (at greatest width) by 4 inch (at

thickest).

Length of dotting, 1; inch.

ce is made of bone and rather more elaborately carved, but

Some Native Snuff-Boxes in the South African Museum. 161

otherwise undecorated. It keeps the same horizontal plane throughout.

The prongs are round in section.

Bowl, 7 inch by 3°5 inch.

Stem, +3 inch by 75 inch (across the wings) by 3°s inch thick.

Fork, 33 inches by 35 inch (at greatest width), each prong, } inch.

d is an ordinary iron sweat-scraper, sometimes used by the

Basuto to take the place of a snuff-spoon. The iron is bent. over

and welded down at the top of the stem to make a hole for a suspen-

sion cord. The stem is almost square in section except where it is

flattened and cross-hatched. The blade is thin, with a midrib on the

upper face only and slightly askew. It curves up at the bottom

about ? inch above the general plane.

Stem, 34 inches by 3% inch.

Band of cross-hatching, #? inch by 3'2 inch.

Length of blade, 23 inches.

Width of blade across wings, 13's inch.

Width of blade at centre, 7's inch.

Width of blade at bottom, 7° inch.

Total length, 52? inches.

e is made of ivory. The front of the prong is plain and flat,

and the underside curved and decorated with triangles of cross-

hatching. The bowl curves very slightly forward. On the back of

the prong, near the top, there is a short triangular projection.

Bowl, 14 inch by 3 inch.

Prong, 4% inches by 4+ inch (at widest) by $ inch (thick).

Length of cross-hatching, 2% inches.

Length of triangular projection, ? inch: height, 35 inch.

f is made of bone. In section it is the segment of a circle and

it is curved up at each end, narrowing gradually from the top, and

hollowed for rather more than half its length to form a bowl. The

front is plain, the back has a length of cross-hatching exactly similar

to that on No. 5.

Total length, 63 inches.

Length of bowl, 32 inches.

Greatest width of bowl, 7%5 inch.

Width of stem, 3° inch.

YOM. P whys

ie.)

Annals of the South African Museum.

REFERENCES.

ALEXANDER.—Narrative of a Voyage of Observation, vol. ii, p. 99. 1837.

. Bent.—Ruined Cities of Mashonaland, pp. 64, 271, 313. 1892.

Brown, On the S. African Frontier, p. 197. 1899.

BuRcHELL.—Travels in 8. Africa, vol. ii, p. 321. 1822-24.

Casatis.—The Basuto, pp. 140-41. 1861.

CHaPpMAN.—Travels in the Interior of S. Africa, vol. i, p. 335. 1868.

. DELEGORGUE.—Voyage dans l’Afrique Australe, vol. i, pp. 354, 355. 1847.

. DuGGAN-CRONIN and LESTRADE.—Bantu Tribes of S. Africa, vol. ii, sec. iii,

Introduction, p. 20, and pl. vi. 1929.

. EnDEMANN.—Zeitschrift fiir Ethnologie, vol. vi, p. 23. 1874.

. Fireminc.—Kafiraria, p. 112. 1853.

. Fremine.—Southern Africa, pp. 204-205. 1856.

2. FrrrscH.—Die Eingeborene Siid-Afrikas, p. 68. 1872.

Or HB Co

for)

bw be bw b bo

[o a)

. GARDINER.—Journey to the Zoolu Country, pp. 106 and 107. 1836.

. GRUTZNER.—Zeitschrift fiir Ethnologie, vol. ix, p. 8]. 1877.

. Ho~tpEeN.—Past and Future of the Kaffir Races, pp. 237-238 and 254. 1866.

. JuNop.—Life of a S. African Tribe, vol. i, p. 115 and vol. ii, p. 1038. 1912.

. Kipp.—The Essential Kafir, p. 56. 1904.

. Kropr.—Das Volk der Xosa Kaffern, p. 103. 1889.

. LAuFER, HamBiy, and Linton.—‘Tobacco and its Use in Africa,” Field

Museum of Natural History, Chicago, Anthropology Leaflet 29. 1930.

. Moopre.—Ten Years in South Africa, p. 260. 1835.

. NavuHavus.—Zeitschrift ftir Ethnologie, vol. xiii, p. 344. 1881.

. Passarce.—Die Okawangozumpfland. Zeitschrift fiir Ethnologie, vol. xxxvii,

pp. 668 and 683. 1905.

. ScoRuMPF.—Die Basuto, wie sie sind, p. 45. 1861.

. ScHuLTze.—Aus Namaland und Kalahari, p. 626. 1907.

. SHooTER.—The Kafirs of Natal, pp. 8-9, 223. 1857.

. THEAu.—Yellow and Dark-skinned Races, p. 226. 1910.

. THomas.—Eleven Years in 8. Africa, pp. 179-180, 214, 278. 1872.

. TéngES.—Ovamboland, p. 85. 1911.

Ann. S. Afr. Mus., Vol. XXIV. Plate XXIV.

Shaw, del,

Neill & Co., Lid.

ZULU SNUFF-BOX.

eh i

a Ae f ey

| ae esd

Plate X XV.

Ann. 8. Afr. Mus., Vol. XXIV.

OO aa

—wier-h © LJ

Sore

Neill & Co., Ltd.

Shaw, del.

SNUFF-BOXES

ZULU

Ann. 8S. Afr. Mus., Vol. XXIV. Plate XXVI.

Shaw, del. Neill d: Co., Lid.

ZULU AND BASUTOLAND BORDER SNUFF-BOXES.

Ann. S. Afr. Mus., Vol. XXIV. Plate XX VII.

Shaw, del. Neill & Co., Lid.

XOSA AND TEMBU SNUFF-BOXES.

Ann. 8S. Afr. Mus., Vol. X XIV. Plate XXVIII.

S ee

eek

stein)

5 AAO

Shaw, del. Neill d Co., Lia.

TAMBUKI, FINGO, AND BASUTOLAND BORDER SNUFF-BOXES.

Ann. 8. Afr. Mus., Vol. XXTV. Plate X XIX.

Shaw, del. Neill & Co., Ltd.

BASUTO SNUFF-BOXES.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XXX.

Shaw, del. Netll & Co., Ltd.

MASHONA SNUFF-BOXES.

Plate XX XI.

Ann. S. Afr. Mus., Vol. XXIV.

Neill & Co., Ltd.

Shaw, del.

NUFF-BOX

BECHUANA AND OVAMBO §

( 163 )

8. Vosburg: Its Petroglyphs.—By A. J. H. Goopwin, M.A.,

F.R.S.8.Afr., University of Cape Town.

(With Plates XX XII-LX and 13 Text-figures.)

THE SITE.

VosBurG is a small village with a population of less than 400

inhabitants. It was founded by the Vos family, and stands 23°

East of Greenwich, by 30° 40’ South. The village lies just within

the Victoria West district of the Cape Province, where it meets the

Britstown, Carnarvon, and Prieska districts. It is served by road

from Britstown 50 miles to the east, from Prieska 60 miles north,

and from Victoria West and Pampoenpoort 60 miles south. It is

accessible from this latter station and from Britstown by a road-

motor run in conjunction with the railways.

The whole area shows rock-engravings, many of which have never

been visited by scientists, but we shall here confine ourselves to those

on the farm Keurfontein, lying 5 miles east of Vosburg on the Brits-

town road.

CLIMATE.

Except in extremely dry seasons, Keurfontein can hold consider-

able stocks of sheep. The country as a whole is arid, and lies within

the Karoo, a region of summer rain but subjected to long periods of

intense drought. The annual rainfall is between 5 and 10 inches,

and this is confined almost entirely to the summer months when

evaporation is at its maximum. When rain does fall it is short and

heavy, but supplies reservoirs which, with use and the enormous

evaporation, will generally last the year. Should these fail, bore-holes

normally supply sufficient water to carry the stock over until the

following year. In extreme cases the stock is railed 200 or 300

miles to a region of winter rains at the Cape, an action which more

or less reproduces the normal game-migrations of pre-Huropean

times.

No such certain sources of water were available to prehistoric

man, and both he and the game he hunted must necessarily have

depended upon the precarious springs of the district. If these

VOL. XXIV, PART 4. 12

164 Annals of the South African Museum.

failed as the dry season came, so game and man must have moved

to more amenable fields for their sustenance. The Orange River at

Prieska is only 60 miles away.

In the Karoo to-day, sheep, hemmed in by herdboys and fences,

die in their tens of thousands. It is necessary to stress this point as

it is one seldom considered in attempting to draw conclusions on

prehistoric climate and animal population in areas whose condition

is known to-day.

As to other evidences, the suggestion of a greater water-supply

during prehistoric times is present. Here and there in this part of

the Karoo are to be seen the remains of ancient lakes, originally

dammed by dolerite dykes which have since broken through. Simi-

larly it can be shown that the very evident drying up of springs

to-day is in part due to a lowering of the water-table by means of

excessive dependence upon bore-holes. Most of this water is brought

to the surface and lost.

The scenery of Keurfontein is typical of the Karoo—startlingly

treeless. The whole area consists of a wide plain from which rise

small dolerite kops, ridges, and plateaux. The alluvium of which

the plain consists is desertic and typical of the region. It is composed

almost entirely of red-brown earth and sand, formed from the

weathering of broken dolerite blocks, which slowly wash down to

find the valley level, leaving the hills relatively bare of earth and

exposing the shiny black desertic blocks which are a monotonous

feature of the country-side.

It is on the exposed surfaces of these blocks of dolerite that the

engravings and other petroglyphs occur, in the open, and never

covered by any rock-shelter.

Evaporation.—It is the constant evaporation of the water which

falls during the hot summer that is the key to the area. When rain

falls, that water which does not immediately flow off in wide shallow

erosion gullies soaks into the ground. In soaking downwards it

dissolves various soluble salts, mainly of calcium, and keeps these

in solution. As time passes the hot sun and dry wind demand

moisture, and this water is sucked up once again bearing its burden

of soluble salts. Towards the surface of the ground these are pre-

cipitated, the crystals forming as the volume of water becomes too

small to carry them.

The local flora is of a peculiar type; it must be able to exist for

long periods without rain, and it must be capable of obtaining water

and nourishment from a soil filled with various salts, and rendered

Vosburg : Its Petroglyphs. 165

“prak.” There is thus little grass, and indeed little apparent

nourishment for game, though sheep seem to have adapted them-

selves. What is possible for stock is possible for game, and in the

Kalahari, farther north, considerable herds of game are supported

by an even more arid climate for longer periods.

Evaporation and Patina.—Of great importance to us is the effect

of this high evaporation, engendered by sun and wind, upon the

surfaces of the dolerite boulders with which we are concerned. The

dolerite is an igneous rock, formed into dykes, sills, and pipes, as

intruded by volcanic forces. Denudation has left the rock bare,

while forces such as lightning, insolation, and original contraction and

stresses of the cooling lava, have broken up the dolerite into misshapen

masses. Where these are in situ they are piled up in ruiniform

groups, but more generally they lie scattered down the talus slopes

of the hill-sides, or over the plain, forming a firm rubble of rocks up

to 6 feet in diameter. Typical ruiniform groups are the Painted

Shelter and High Rock, but these are unusual.

The forces which have broken up these rocks still continue to act,

and I have been able to note three new fractures of large rocks during

the four years covering my visits. The newly broken surface is

olive green, and shows a recognisable freshness and sparkle, which

is dulled after rainfall.

The normal surface patination or oxidation of the dolerite is of

two types, the one affecting rock above ground, the other (which

does not here interest us) seems to continue below ground in the

presence of moist soil. This latter type forms a matt, powdery

crust of ochreous red, easily washed away by running water to join

the alluvium of the plain. No such surfaces have been used for

making petroglyphs. This type of patination may also affect rela-

tively deep rock-pittings, which hold water and thus provide the

continued moisture necessary.

The patination of rocks above ground is largely a product of

oxidation due to evaporation. As in the case with the soil, water

falling on the exposed portion of the rock spreads over most of the

surface of the boulder. It penetrates to a slight depth, generally

a little over a millimetre, dissolves out soluble iron, manganese, and

other salts, then evaporates, bringing the salts to the rock surface

and depositing them there. Here they oxidise to form a brown-

black matt patina, to which the wind eventually gives a fine glaze.*

* A similar state of affairs is described from Australia (see Basedow, ‘‘Abo-

riginal Rock-carvings of Great Antiquity in Australia,” Journ. Roy. Anthr. Inst.,

vol. xliv, 1914, p. 199).

166 Annals of the South African Museum.

Once soluble salts have been removed from immediately beneath

the surface, the way is left open for further inroads of the action;

thus patina may in rare cases affect the rock to a centimetre in depth,

though the average thickness of the affected rock is generally about

5mm. This is not all true patination, and in a patinated rock three

general layers are to be observed. There is an outer skin or film of

true patina, almost black in colour, forming a partly protective

“paint.” The thickness is fairly constant and is maintained by the

action described above, while any growth is counteracted by wind

polishing. Beneath this and of varying (2 mm. to 7 mm.) thickness

is a light brown layer from which soluble salts have been leached;

while beneath this is the unaffected rock.

Factors IN PATINATION.

Various writers, notably Basedow, and, in South Africa, Péringuey

(S. Afr. Phil. Soc., 1906, p. 401, and 1909, p. 401), have attributed

extreme antiquity to petroglyphs which have been heavily patinated.

Since their time various attempts have been made to fix the relative

order in which groups of petroglyphs were made, by means of a study

of apparent patination.

It might appear obvious that the deeper the patina the older the

petroglyph, and that is the presumption underlying this method.

In estimating the age of petroglyphs from the relative patina observ-

able, I would suggest that a large variety of important and complex

factors come into play which render such a presumption less safe,

unless they are taken into constant consideration, or a sufficient

number of cases are observed to cancel out these variative factors.

1. Climatic conditions of alternating wet and very dry, through

succeeding years, with extreme evaporation.

2. Distribution of water over exposed rock-surfaces.

3. Direction of wind and sun.

4. Presence or absence of previous patination and pitting.

5. Action of birds.

All these factors affect the distribution, the thickness, and the

intensity of patination, and all act unevenly through time with no

definite relationship to the period over which they act. It is not time

which produces patina, but the various factors mentioned acting

alone or together, and more or less intensely over varying periods.

The general process of patination as a result of evaporation has

been described above, and we need not recapitulate here, but the

various other factors need amplification.

Vosburg : Its Petroglyphs. 167

The distribution of water over the rock-surface is governed by the

direction and intensity of a rain-bearing wind, the flow of water over

the rock, the drip-line and so on. The more intense distribution

would necessarily occur on the exposed upper face of the rock, the

least intense beneath the rock and within the drip-line. The direction

of wind and sun will also have their effects. The side protected from

both these evaporating agents will remain moist for longer, and

infiltration of moisture would thus be allowed to go on for longer

periods. The abrasive action of the wind in controlling and checking

the growth of more than a skin of complete patination must also be

taken into account.

The presence or absence of previous patination is important. A

new surface will dull rapidly, but an old patinated surface reacts very

slowly to further inroads. If the surface patina has been scratched

to leave the under layer exposed, the water will have greater pene-

trating powers, and repatination will be quick. Where the scratch

is sufficiently deep to hold water, patination will quickly heal the

wounded rock-surface, as the amount of immediately available water

is greater. It seems obvious that designs facing upwards on the rock

will hold water, and patination will be fast. Where the petroglyph

is Shallow, the speed will be affected by the exposure of the unprotected

second layer.

The action of birds does not affect all rocks, nor the whole surface

of any one boulder. As there are no trees, birds are given to settling

on certain rocks, generally those higher than their surroundings.

Not all such boulders are chosen, and the birds show marked pre-

ferences. Those perches they choose are affected by their droppings,

which form a cap and seem to have two effects. They coat the rock,

keeping it damp and cool and impeding evaporation. More important

is the chemical action of the various acids, etc. in the droppings,

which effectively remove the black outer skin of patina, exposing

the second layer. Those rocks upon which birds have perched are

thus chemically bleached to a light brown and bear no polish, while

any shallow petroglyphs on rocks so coated are destroyed. This

action is plainly visible on Big Rock, where it would appear to have

been made use of by man as a medium of artistic expression.

It is worth noting that while the intensity of patination has gener-

ally been used to prove the relative age of petroglyphs, it has been

just as readily admitted that the patination on certain rock-engravings

equals that on the exposed rock-surfaces about it. This admission of

such a maximum of patination (admitting as it does that an engraving

168 Annals of the South African Museum.

can “‘catch up” with the patination of the original rock-surface) is

important, and to some extent logically defeats the end in view. If

all engravings lay at the same angle and had consistently suffered

from the same atmospheric actions, and if they were all of the same

depth, we might presume to give a table showing the relative age of

the various artistic periods, phases, or schools represented. Too many

factors enter into the question of patination to do this with entire

safety, but despite these a considerable amount can be done.

I shall give a suggestion of order in this paper, but it is tentative,

and cannot, except where superposition is shown, be considered as

absolute. But it must be noted that relative order, as observed from

patination, agrees with the relative order obtained from palimpsests.

Each period (except the first) is based upon a large number of examples,

and the direction of the petroglyphs in relation to the prevalent rains,

winds, and the sun is therefore cancelled as a factor. Other important

factors still remain. The most important is the depth of the original

incisions in the rock.

DISTRIBUTION.

When first I visited the farm Keurfontein in 1926, with the late

Mr. F. J. Jansen and Mr. C. H. T. D. Heese, it was thought that the

engravings were confined to a single hill, and in 1928 when I visited

the site once again, with Mr. and Mrs. Burkitt, only the series on this

hill were shown to him. His conclusions have long been published

(Burkitt, South Africa’s Past in Stone and Paint, pp. 145-149,

Cambridge, 1928). Since that time I have been enabled to pay three

visits to the site, ranging from three weeks to four days, and on each

visit numbers of new specimens have been found, many some distance

from Burkitt’s Hill. These I have divided into four regional groups.

The various points referred to are shown on the accompanying plans.

The rock known as High Rock, which I have taken as a fixed point,

is a prominent ruiniform arrangement of stones on the hill described

by Burkitt (“Burkitt’s Hill’’).

Group I.—About half a mile east of High Rock is a long dyke

running directly towards the farmhouse; on this ridge are a number

of petroglyphs, very scattered and seldom of very great interest.

Many of these have been photographed, but not all have been traced.

Certain earlier styles are missing.

Group II1.—High Rock stands on a small kopje (Burkitt’s Hill),

perhaps 50 feet high, which covers an area measuring some 200 by

300 yards. All over the upper part of this kopje, and in the talus

Vosburg : Its Petroglyphs. 169

of fallen boulders strewn about, petroglyphs are to be found. They

are not markedly grouped, and seem to occur only on suitable rock-

ROAD To DISTANT

Farm House + 2 Rock

Big Rockk=s At

ape/ ~ =

~sMlGH *

AP Ra neo

DETAIL ON

BURKITT'S HILL.

3O METRES

Fie. 2.—Detail on Burkitt’s Hill.

faces, but they form a rough circle about High Rock, of which Burkitt

says (op. cit., p. 146):

170 Annals of the South African Museum.

“On the summit of the decorated kopje near Vosburg, a large

boulder, much undercut, had doubtless been used as a shelter,

and there we found a Smithfield industry. Although the con-

temporaneity of this with the engravings cannot be proved, it

would seem probable that at least some of the series which we

are going to describe are the work of Smithfield man.”

A continuation of Burkitt’s Hill forms a small kop, which I have

called Distant Rock, 150 yards north of High Rock. A few badly

made petroglyphs are found here, all in the late scratched style and

of little interest.

Group III.—West of High Rock a road skirts the kopje, leading

from the farmhouse to a windmill. The valley along which it runs

is about 100 yards wide and is bounded on the west by a low

plateau (Big Rock Plateau) which stands a few feet higher than

Burkitt’s Hill. On this plateau are a number of petroglyphs; gener-

ally they are found in groups of four or five, where useful rock-surfaces

occur together. The most important group is found immediately about

Big Rock, which is 220 yards west of the eastern lip of the plateau.

In the valley dividing Big Rock Plateau from Burkitt’s Hill, and

scattered about among the engravings on the plateau itself, are a

number of “homes,” or sites inhabited by prehistoric man. The

High Rock Home (Burkitt’s undercut boulder) now yields nothing,

but seven other homes have given us an abundance of implements,

pottery, bored-stone fragments, ostrich eggshell and so on. They are

generally in earth-filled hollows, between the boulders and not

sheltered to any extent, which seem to have been used for the erection

of rough dwellings, though no hut foundations of any sort are

discernible.

Group IV.—Separated from these other groups by a distance of

about a mile is a long, fairly high dyke. It is capped by ruiniform

dolerite, which holds back a plateau very like the Big Rock Plateau.

This ridge overlooks the Vosburg-Britstown road and is perhaps half

a mile west of the farmhouse. In the main dolerite “ruin” is a small

shelter composed of a number of blocks, on the largest of which are

paintings. This I have called the Painted Shelter. Both on the

plateau behind the shelter and on the small hill which it overlooks

are several petroglyphs.

It is almost certain that search would reveal petroglyphs on all the

hills on this farm, and indeed throughout the district, but search is

a slow process, and it was thought that the four main groups described

above would be sufficiently representative.

Vosburg : Its Petroglyphs. 171

TECHNIQUES EMPLOYED.

The basic method of making petroglyphs at all sites in South

Africa was always the same. It consisted of the removal of the outer

black glaze on the rock to reveal the second dull light-brown layer,

or possibly in some few cases the removal of both layers to reveal the

olive-green rock beneath. The rock was always heavily patinated

before the petroglyphs were made, and a deeply cut design would

originally have shown olive-green below the black surface, while the

more usual shallow designs would be brown.

The methods by which the patina was removed varied with different

artists and with different periods. Some artists incised or engraved

sharp lines, some pecked deeply into the rock, some pecked just

deeply enough to remove the dark outer skin. Others have rubbed

off the skin mechanically, while there is evidence that during at least

one period the artists at Vosburg have removed the outermost film

chemically. This was effected by painting designs on the patinated

rock with bird-dung, and allowing the action of the naturally con-

tained chemicals to take its course.

Burkitt (op. cit., pp. 146 et seq.), in discussing the techniques which

he described after visiting the High Rock group, says:

“Four distinct techniques almost certainly of very different ages

can be determined as follows :—

“Engravings of the earliest series show a fine incised outline, and

the bodies of the animals are filled in with fine lines more or less

parallel to this outline (fig. xxv, 1). In some cases the surface of the

engravings has almost the appearance of having been rubbed over

later, but this may be partly due to subsequent weathering. The

important fact to notice is that the patina of the engraved lines is

exactly as dark as that of the rock on which the figure is engraved.

This indeed often makes it very difficult to see the figures, and it is

not to be wondered at that examples of the incised style are not seen

in museum collections. At Vosburg we noted elephant and eland

figured in this series.

“Another series comprised figures of animals made by a pocking

technique without any definite outline; sometimes there is merely

an outline formed by a more or less wide band of coarse pockings

(fig. xxv, 2); sometimes the whole of the body of the animal, some-

times the head and neck only, are covered with pock marks. This

pocking technique clearly shows that a punch was used, and in view

of the hard nature of the dolerite one would imagine that punches

172 Annals of the South African Museum.

must have been made by shaping some still harder pebble to produce

a sort of cigar-shaped point. But no such tools, or indeed any tool

suitable for making these engravings, had yet been found. The

patina of this second series is not quite so deep as it is in the case

of Series I; on this ground alone we came to the conclusion that the

pocking technique here was invariably rather more modern than the

incised technique of Series I. A variation of Series II can be recog-

nised where the pock marks, instead of being little more or less

circular hollows, consist of dashes. It is quite easy to see how this

variation probably arose. If the punching was effected by means

of direct blows at right angles to the rock, the normal kind of pocking

would result. If, however, the punch were held obliquely to the

rock-surface there would result a long dash in place of the circular

pock mark. Generally, the body of the animal is covered all over

with such dashes, but less densely than in the case of the normal

pocking. The patina is similar. In Series II we noted figures of

eland, springbok and rhinoceros.

“Series III is the most important and the most frequently seen of

all, and, judging from the fact that the patina is sometimes fairly

deep and sometimes very slight, it must have lasted through a long

period of time. Especially when only slightly patinated, the drawings

are lighter in colour than the patinated surface of the rock itself,

and are therefore very visible. Engravings of this series show a clear

outline, the body of the animal being, as it were, deeply rubbed all

over. That this ‘rubbing’ was done with a fine punch making a

minute pocking all over the surface seems to me to be proved by the

fact that definite pock marks can occasionally be made out. Buck,

ostrich, and eland are figured, and also some metal spears which

occur in superposition over the body of an elephant belonging to

Series I. This fact is important from two points of view. Firstly,

because here stratigraphy confirms the deductions obtained from a

study of the preservation, 7.e. degree of patination; such a confirma-

tion is necessary, for of course weathering is always a capricious

factor, although the engravings being all in the open and close to-

gether, and the rocks being all of the same material, it generally forms

a fairly safe guide. Secondly, because the occurrence of what must

have been meant for metal spears indicates that a part, at least, of this

third series only dates back as far as the introduction of metal into

South Africa, which was due, in the first instance, to the Hottentots.

“Lastly, at Vosburg we noted a series showing no patina and made

in various ways, generally with a metal knife; this was clearly the

Vosburg : Its Petroglyphs. 173

work of modern herdsmen, doubtless in imitation of the older

engravings around. The bodies of the animals are sometimes filled

in with cross-hatching, an uncommon feature in the earlier series.”’

Since Burkitt’s visit I have been able to return to the sites at

various times and to make a considerable survey of the field. The

evidence collected points to the petroglyphs being the products of

a long series of prehistoric artists. The data here obtained did not

contradict Burkitt’s findings, but amplified them very considerably.

With the discovery of considerable numbers of other petroglyphs

generalisations were possible which he had not the data to make.

The material will best be understood if the general findings and

sequences are first stated and the proofs given afterwards. These

proofs are of two types; the absolute proof of direct superposition;

and the somewhat less satisfactory proof of comparative patination.

While I regard patination as inconclusive proof, it must be remembered

that the various drawings composing the palimpsests show just

those differences in patination which we are employing in the second

type of evidence. This very considerably strengthens our case in

making use of this latter method.

From the findings I have divided the various types into eleven

styles. Some of these depend upon patination alone, some on both

sources of evidence. These styles I have numbered in Arabic figures,

1-11. Some can be redivided into substyles, which may represent

either the return of the old artists a few years later, or imitations in

the same technique by later artists.

These styles may be grouped into general phases thus:

Early, deeply engraved. Styles 1 and 2.

Heavily pecked. Styles 3, 3a, 33, 3c.

Chemically made. Styles 4, 4a.

Middle engraved. Style 5.

Lightly pecked. Styles 6, 6a.

Mechanically rubbed. Styles 7, 8, 10.

Late engraved. Styles 9, 11.

At this site Styles 9, 10, and 11 may all be relegated to the age

of metals.

SUPERPOSITIONS OR PALIMPSESTS.

Big Rock.

This rock stands out on the centre of its plateau, somewhat reminis-

cent of a Central American monument, and is as completely covered

with designs, though these are not so deeply cut, nor are they all

174 Annals of the South African Museum.

related to one another. The rock forms a rough sphere of about

4 feet in diameter and is banded about by an intricate complex of

lines. Closer examination reveals six differing periods of artistic

execution * (Plates XX XII-XX XVII and LVIII, B).

On Big Rock the earliest artistic efforts show an early scratched

type (Style 2). A well-drawn eland is typical of this old style. The

incisions are completely patinated and are at one point covered by

the second style. This deeply engraved and heavily patinated style

belongs to Burkitt’s Group I.

Second is the peculiar, and apparently local, type of work which

first suggested rubbing to me, but which I have since found to show

little sign of mechanical action. The workmanship does, however,

conform absolutely to the action of bird-dung on the rock, and it

seems almost certain that the patinated rock was painted with dung

(probably to give an immediate white design), and as this washed off,

the pattern remained. Quite possibly only frequent repainting over

the same lines would have so marked an effect (Style 4).

Third is a group of elongated and bizarre human figures, rough

animals of an equally quaint type, ladders and grills, targets or con-

centric circles. This I have called Style 44. In technique it is identical

with Style 4, but is later, slightly less patinated, and here and there

can be seen to overlie Style 4. Where they do not overlap it is

difficult to say which is which, and in fact at many points of over-

lap the superposition is not decipherable. The second series does,

however, show more recognisable animal and human forms than the

earlier.

Fourth (by patination only, for there is no palimpsest present) is

a spotty or dashed technique, showing a delightful study of an eland

head, with a giraffe head imposed upon it (Style 6). The patina is

very much lighter than that visible on the two previous styles, and

there can be little question of the order here.

Fifth, if it can be regarded as a style separate from the fourth,

is an imitation in very deteriorated mode. A scattered patch or two

of badly executed forms is all that represents this. One patch shows

a shape resembling a shark’s egg, another a possible buck. The

patina is slightly lighter than that of the two animal heads which

are on the same panel, and has thus been called Style 6a.

* In the following descriptions Arabic figures refer to the style in the general

series outlined above and developed below; verbal numbers to the order in the

palimpsest which is being described; while references to Burkitt’s series will be

specific.

Vosburg: Its Petroglyphs. 175

The sixth and final style is lightly and faintly scratched. It

overlies the third style in places, and is represented here only by a

single shape very much resembling the light wooden structure of a

domed Hottentot hut. It is obviously relatively recent. One or

two similar hut-like specimens are to be found elsewhere, for instance,

an exactly similar representation 250 yards south-east of Rock 30

(Burkitt’s Hill) on a dolerite rock in the valley. There is also a badly

drawn zebra on Burkitt’s Hill of the same age (Style 10).

Paintings Hill Rock.

On what was originally a single rock, about 100 yards west

of the Painted Shelter, and on the crest of the hill looking towards

Vosburg, is a large representation of a buffalo (Plate XX XVIII).

Since the engraving was made, this rock has split at least twice. The

first split removed the tail, the second cut into the belly-line and hind

legs of the buffalo. The animal is deeply engraved with incised lines,

obviously belonging to Burkitt’s Period I (Style 2).

Superimposed directly upon this, and seemingly confined by these

breaks in the rock, is an eland, lightly incised, and not much patinated

(Style 5). It stretches from the neck of the buffalo to the belly, the

tail extending farther. Overlying the tail is a crude ostrich in a late

scratched style (Style 9). Overlying the hind legs of the eland is

another ostrich in a rather similar late scratched technique. It is

impossible to say what the relationship between the two ostriches

may have been, but from the relative positions of the legs, the stylised

tails, and the general similarities of form and pose, they are either by

the same artist, or the simpler specimen was copied from the other.

The patina of both agrees, and superposition shows them to be later

than the eland, and, therefore, than the buffalo. We may accept

them as belonging to one style.

Rock 6.

Rock 6 is the only palimpsest on Burkitt’s Hill (Plate XX XIX).

It bears an elephant in Style 5, rubbed and scratched. About it and

related to it, though of a much later date, are a few badly drawn

human figures. They are throwing huge lanceolate spears, the heads

being as large as the throwers. The spearheads are certainly of

metal and overlie the elephant. On neighbouring rocks are to be

seen similar individual spears, some barbed specimens proving that

the originals were of metal (Style 10).

176 Annals of the South African Museum.

Rock AK.

Rock AK on Big Rock Hill has a rhinoceros and a zebra deeply

pecked upon it side by side (Plate XL, B). Overlying these are two

elands, scratched all over, the scratches following the natural direction

of the hairs. The superposition here shows Style 3a with Style 5

imposed upon it.

Apart from these, there are a few palimpsests of no importance.

They show either completely modern drawings over earlier styles,

or are not sufficiently definite for us to draw any conclusions—a few

scratched lines over an older style, and so on.

We may summarise the data obtained from these definite sources

by comparing the actual palimpsests on Big Rock, Rock AK, Rock 6,

and the rock on Paintings Hill:

2 4,44 6, 6A 10 | Big Rock.

3A 5 Rock AK.

2 5 9 Paintings Hill.

5 10 | Rock 6.

THE SERIES OF STYLES.

Turning from these sequences to those suggested by the relative

state of. patination on the petroglyphs, a considerably augmented

series may be made. Where the evidence of sequence has been proved

absolutely by the superpositions described above, these are referred

to. In other cases references are given to examples of each style.

These do not constitute a complete list of the petroglyphs studied.

Here the order relies upon the long and tedious task of relating each

patina with every other.

1. Very deeply patinated cross-hatched designs (Plate XLI). A

curious design, cut on Rock AL, is the only specimen of the style

identified with any certainty. The design is deeply cut, and so heavily

patinated that it is not visible as a whole from any angle, but parts

stand out on the curved rock with differing cross-lights. The camera

actually reveals more than is visible to the eye. No attempt to trace

this specimen succeeded, but the photographs taken in a suitable

light show the figure well. The object, be it idea or bird, repre-

sented is too stylised or diagrammatic to be recognisable. It is

finely cross-hatched, and appears to end with a conventionalised

swallow-tail.

2. Animal forms in deeply scratched lines following the natural

Vosburg : Its Petroglyphs. 177

direction of the animal’s hair. The lines were cut with a fairly coarse

tool which gives a gouged effect. This is Burkitt’s first period

(Burkitt, op. cit., plate xxv, 1) and is the oldest style in the palimp-

sests on Big Rock and Paintings Hill. Rock 27 also shows a fine

eland (Burkitt’s specimen) and Rock 28 bears a zebra.

3. Deeply pecked outlines, carefully executed, with some slight

attempts at body marking. This is Burkitt’s second period (Burkitt,

op. cit., plate xxv, 2 and 3) and he has taken his illustrations from

Rock 30, immediately below High Rock (Plate XLII). The one shows

a buck with a ring on its neck and stylistically rounded feet, the other

is an animal with defined and peculiar markings. Rock AL and

several other rocks show ostriches belonging to this style; Rock AO

(Plate XLIII, A) has a pair of ostriches, one in outline, one filled in.

Rock AN has an unfinished buck. All of these agree in patina and in

style.

34. This may be a development from Style 3. It is similar in

technique but the figure is filled in with attempts at body markings.

This represents locally the best known of our South African artistic

periods. Almost all the more striking specimens of rock-peckings

in both the Transvaal Museum and the South African Museum con-

form in type and patina (where this has not been oiled by authority)

with this style.

It is sometimes difficult to draw a clear line between this style

and its neighbours, 3 and 3B. Many specimens waver, and may

perhaps belong to transitional styles. At Vosburg 3a is almost

confined to Big Rock Hill, which explains why Burkitt missed it. It

generally depicts the quagga. The earlier quagga and rhinoceros of

the palimpsest on Rock AK belong to this group. The Styles 3 and 3a

are perhaps linked by the two ostriches on AO, one of which conforms

to each of these styles (Plate XLIII, A). Rock AF bears two deeply

pecked quagga, only the shoulders being striped. These are of interest,

as in both instances and in several other peckings of the same style

on nearby rocks the animal has been naturalistically depicted then

spoilt. The additions consist of added trunks, tails, extra legs,

enlarged bellies and so on. What is most important is that these

additions are in the same laborious technique, and belong to the

same period and patina as the well-drawn quagga which have been

marred. Such studied vandalism, apparently by the artist himself,

is difficult to explain (Plates XLIIT, B, and XLIV).

Rock 29 has an ostrich, and several rocks in the neighbourhood of

Big Rock have quagga depicted in this style (Plate XLVI).

178 Annals of the South African Museum.

3B. Outlined figures, pecked all over, the pecking being “dashed”

to follow the lines of the hair. These figures are generally slightly

conventionalised, but good. Rock 3 has an eland of this type, Rock 10

a buck with a few dashes on the body. Rock 14 has a bearded eland

(Plate XLVIT), while the outline of a second has been started and left.

On Rock 31 is a buck, sparsely “dashed” on the body, and there is a

later design on the same rock. Style 3B is Burkitt’s variation of his

second style. Probably a quagga bearing light graffiti on the body

belongs to this style (Plate XLV).

3c. Now seems to come a deterioration of the deep pecking tech-

nique. The form of the animal is generally bad, so bad that even

the zoological family is not recognisable; some show animals which

might be buck, lions, or baboons. What is lost in form is adequately

compensated for in a fine vivacity and movement, which gives a

crude, but pleasingly humorous effect. The figures are generally

confined to “dancing buck,” what appear to be horns being shown.

Rock 33 has a “lon” pecked deeply all over in this style, Rock 35

has a pair of opposed dancing or fighting buck, and Rocks AQ and AN

have bucks in similar attitudes (text-figs. 3 and 4). It is possible

that this “dancing buck” style precedes 3B.

4 and 4a. These “chemical” styles are almost entirely confined

to Big Rock. Here two periods are visible, the older being the better

executed and slightly more patinated than the later, which it under-

lies. There is little difference either in style or time between these

two phases, and we may introduce them together.

Judging from a careful examination of the rock-surface, it is

extremely probable that the technique consisted of painting the black

rock with white bird-droppings; perhaps the painting was repeated,

but the white paint was left in place sufficiently long to attack the

rock chemically. At the top of the rock the lines of the encircling

“snakes”’ of both these periods merge completely with the patch left

by the natural action of the birds which have perched here, and cannot

be differentiated from this. There is no sign of mechanical rubbing

on the lines of the drawings (Plates XX XIII, XXXIV, XXXYV).

4. To the earlier period belong ridiculous and whimsical animal

figures, drawn at all angles. Two ostriches with necks intertwined,

a pair of quagga, the front legs of the one forming the back line of

the other, a weird animal with beaver-like tail, a baby buck with

umbilical cord, and a number of lines encircling the rock, can all

be relegated to this style (Plate XXXII).

44. To the later period of the two belong very similar snakes,

Fic. 3.—Human figure in Style 4a.

Fic. 4.—Figures in Style 3c.

VOL. XXIV, PART 4.

“Dancing Buck”’ in Style 3c.

“Dancing Buck” type.

180 Annals of the South African Museum.

better drawn, and crossing and recrossing the older specimens. With

these are some weird human figures, one showing steatopygy.

Possibly the devil (a distinctly “medieval” representation with horns,

a tail, and an erect penis) is of this period too, but as he stands on a

different face of the rock his patina is somewhat darker (Plate

XXXVII). It is worth noting that at Vosburg all the grids and

concentric circles belong to this period. These forms are very

common in the Kimberley district, and many have been depicted

by Miss Wilman.* In that region it is certainly a pecking technique,

and shows no sign whatsoever of having been done with the bird-

droppings. One or two concentric circles at Vosburg have been

similarly pecked, but these also conform to Style 4a in patina.

ROE

Fic. 5.—Maroon Paintings. Painted Shelter.

To one of these two periods can be assigned the dancing or squat-

ting human figure on Rock AD (text-fig. 3) and a figure between

Rocks 2 and 3.

5. Deeply scratched design, recalling Style 2 in all but patina

and the animals depicted. Style 5 shows a fairly old patina, and may

be more closely related to Style 2 than I have suggested. The

palimpsest on Paintings Hill shows a very clear gap between the

buffalo of Style 2 and the eland of this style. Similarly the super-

position on Rock AK proves this style to be later than Style 3a.

It can therefore only precede Style 4, but the patina on this latter

type is darker and presumably older than that on Style 5.

Rock 6 is the best example, and here again we have a super-

position, this time the metal spears referred to above (Plate XX XIX).

An elephant 30 yards below the Painted Shelter is of this period

also.

6. A good revival in pecking follows either Phase 44 or Phase 5.

It is typified by light dashes rather than pecking, recalling though

hardly resembling Style 3B. Here again the dashes follow the

* M. Wilman, Rock Engravings of Griqualand West, Cambridge, 1934.

Vosburg: Its Petroglyphs. 181

lines of the hair on the animal, but the strongly pecked outline is

missing, and the patination is very much lighter and of no very

great age.

In this style there is a peculiar tendency, not markedly seen in

other South African prehistoric art, but present in the Huropean

Aurignacian, to concentrate upon the head of the animal. This is

carefully and feelingly delineated, while the body is very much less

carefully executed, or left out entirely. This is the most artistic,

though not the most impressive period at Vosburg.

This style is spacially the most nearly related to the paintings in

the Painted Shelter, and there are two specimens within a few yards

of this site. One of these is a buck 10 yards from the shelter, the

other a buck’s forequarters 3 yards from the shelter. Comparison

with the elephant 30 yards below proves that this style is consider-

ably later than Style 5.

On Big Rock can be seen two “portrait heads,” one of a buck, the

other of a giraffe (Plate XXXVI). This is the only attempt to draw

two animals abreast, a distinct artistic achievement.

6a. Possibly the deterioration of Style 6 accounts for the weird,

loosely pecked figures, sometimes badly outlined with light scratches,

which occur here and there. The style has no interest whatsoever.

One specimen occurs immediately behind the two heads in Style 6 on

Big Rock, and the horns of the giraffe seem to have been elongated

in this deteriorated style. A few other figures elsewhere belong to

this type. Among the Eastern Outlyers is a very fine rhinoceros

(Plates L—LI), which, though spoilt by later scrawlings, shows

that even the artists of this period were capable of a certain

degree of attainment. Rock 28 shows a sparsely pecked animal,

Rock 3 a similar buck, Rock 2 a rhinoceros and a rhinoceros head.

There is also an equid in the same technique.

7. A rubbed series with fairly good animal forms (Burkitt’s third

style). The figures are often outlined, and the incised outlines show

a lighter technique than the rubbing. This can only be accounted

for by the fact that the rubbing has not removed the whole outer

skin of the rock, and hence was never greatly contrasted with the

background. The incised lines, however, have cut through this

outer skin, and show up more markedly.

Rock 10 bears a badly outlined wildebeest, Rock 11 has a group

of ostriches, zebra, and buck. The grouping here is very similar to

that on Rock 9, and the two certainly belong to a similar style.

Rock 15 shows a large, pregnant eland, and both 23 and 24 show

182 Annals of the South African Museum.

elands with bunches of incised hairs at the end of the tail. Rock 22

has a strongly outlined buck (text-fig. 6), 23 a lightly rubbed buck

with no outline, and 26 has a similarly rubbed ostrich. On the

Big Rock Hill, AM has a crude picture of a lion biting off a man’s

head (text-fig. 7), and AQ has a zebra, the mane and stripes of which

have been incised with engraved lines (text-fig. 8). Near Big Rock

itself is a porcupine in much the same style, but the patina has

the appearance of being very much older. Possibly this is due to the

Fic. 6.—Buck in Style 7, on Rock 22.

figure being drawn on a horizontal flat surface which has undergone

the maximum of weathering.

8. Very lightly rubbed outlines. Rock 32 shows a series of pools

much as they would appear in the rainy season in a shallow ero-

sion gulley. Such a gulley is easily visible from this spot, perhaps

200 yards east. Near this, also on Burkitt’s Hill, is a crudely

executed drawing of a human couple copulating. An ostrich on

Rock 34 may belong to this style. A buck, depicted with a queer

form over its head, possibly representing water as seen in perspective

from the hill above, is the only animal which can certainly be relegated

to this style.

9. Lightly scratched and scrabbled figures. The elephant and the

rhinoceros are generally depicted, but a fine eland, some 4 feet long,

is to be seen among the Eastern Outlyers (Plate LII). A buffalo

being followed by a man, in the same region, may belong to this

Fic. 8.—Quagga rubbed, with incised markings and outline. Rock AQ. Probably Style 7.

184 Annals of the South African Museum.

style, or possibly to Style 7 (Plate LIII). The three human figures

below this buffalo belong to the same period (Style 10) as those added

to the elephant on Rock 6.

On Rock 9 is a group of zebra, buck, and ostriches. These are

grouped very much in the same way as those on Rock 11 (see above,

Style 7), but are badly done and weakly sketched. On Rocks 16

and 17 are lightly rubbed elephants, with small trunks and wrongly

attached tusks, suggesting sketches from memory. Near Distant

Rock are several similar elephants, composing almost the only

drawings present on this kopje.

10. We now reach the metal age with certainty. The lanceolate

spears which are drawn, some with recurved barbs on the spearhead

itself, can only have been made of metal. Rock 6 is the best example,

as the spears overlie the figure in Style 5. The human figures often

added to older engravings (Plate LIII) are all of this period. The

Hottentot form of hut shown on Big Rock and elsewhere can be

relegated to this period.

11. Here we first meet European influence. Outlined ostriches

deeply scratched with a metal knife: men on horseback (text-figs. 9

and 10); some with guns; heavily skirted women with Boer sun-

bonnets, all belong to this period (Burkitt’s final phase). These are

followed by contemporary love tokens, initials, and other trivialities.

Interesting in this group is the presence of Nine Men’s Morris boards,

formed of three concentric squares, the midpoint of the sides of each

being joined. This game is not African, and is not known to-day

in the district.

THE PAINTED SHELTER.

The paintings still remain to be described. Although the local

people speak of several sites at which paintings occur, only one could

be located; this was the Painted Shelter. The position of this in

relationship to the engravings has been described above, under

Group IV.

At this point the crest of the hill, overlooking the Keurfontein-

Vosburg Road (Panorama, Plate LX), is composed of ruiniform

dolerite boulders, evidently the edge of a tilted volcanic sheet

(Plate LIV). Among these clustered rocks is a small shelter, 4 feet

wide at the mouth and less than 8 feet deep, and hardly covered by

a roof. Paintings only occur on the three exposed faces of the

largest dolerite block (Plate LV). They are all in one colour, a

-Fic. 9.—Ostriches, and a man on horseback, Style 11.

Fic. 10.—Zebra in Style 11.

186 Annals of the South African Museum.

maroon or plum, and all belong to the same period, no superposi-

tions being present. They have suffered considerably from exposure

to the actions of water and bird-droppings.

There are no human or animal figures depicted; or, if there are,

they are so conventionalised that recognition is impossible. The

paintings have all been drawn directly upon the rock with the tips

of the fingers dipped in paint (Plate LVI). Two types of sign

appear. The one consists of two concentric semicircles, mouth

downward, with perpendiculars occupying the remainder of the space

within the smaller semicircle. The second generally shows a circle,

an inch in internal diameter, with a perpendicular line rising at right

angles to the upper horizontal tangent. The line may bisect the

circle and continue, or it may begin at the uppermost point of the

circumference (text-fig. 5).

It is possible, even probable, that these designs are phallic, the

first representing the female, the second the male symbols. There

is another possible explanation, that the first represent tectiform

pictures, suggesting huts of bent sticks, and the second an extreme

conventionalisation of the human form. A variant of this second

sign does in one instance certainly represent the human figure.

There is no method in grouping, save that the eastern face of the

rock is mainly of the first (female or hut) type, and the northern face

shows the second (male or human) type more generally.

It is important to note that these signs are very similar indeed to

signs described by Dart (R. A. Dart, “ Rock-engravings in Southern

Africa and Some Clues to their Significance and Age,” 8. Afr. Journ.

Sci., vol. xxviii, 1931, pp. 475-476, illust.) from a site on the Chifubwa

stream near Solwezi. These he relates with later Stone Age imple-

ments. It is interesting, too, that the signs at that site are both

pecked and painted.

Three yards north of the shelter is a relatively recent buck in

Style 6; only the forequarters are properly depicted. A few yards

farther is another, the whole animal being picked out in a similar

style; while a little farther away, 10 yards north of the shelter, is a

third. The two last are slightly more heavily pecked than the fore-

quarters of the first, and may form a separate and slightly earlier

form of Style 6.

Thirty yards below the shelter is an elephant in a deeply scratched

style (Style 5) which is earlier than the buck in the neighbourhood

of the shelter.

Considering how nearly the Style 6 approaches the shelter, and how

Vosburg: Its Petroglyphs. 187

little sign there is of the shelter having been regularly inhabited, we

are justified in associating these paintings with Style 6, and both with

the implements found in the shelter.

The paintings would seem to link up with various other sites in

the Cape Province, and Miss Bleek has suggested several affinities.

A site in the Griquatown Hills, Griquatown West; cave in Klein

Aasvogel Kop, O.F.8.; Rooi Kuil, O.F.S.; Lichtenstein, Orange

River, O.F.S. All these sites show similar colouring and technique,

though at most of them the signs depicted are different. They all

appear to be late.

IMPLEMENTS AND Home SITES.

In open sites such as these it is impossible to associate implements

and petroglyphs with any hope of accuracy. Vosburg has no true

deposit-bearing shelters; the only one, the Painted Shelter, contains

an inch or so of sand on a rock bottom, and a dozen or more imple-

ments were retrieved from here.

Scattered about among the rocks surrounding the petroglyphs

are an abundance of artefacts, and considerable collections of these

were made. The evidence afforded by their presence does not

relate them happily with the engravings. As a long sequence of

petroglyphs has been shown, and as a long sequence of artefacts has

been found, we must presume that the implements and _ rock-

engravings were parallel to some extent. But which group of

implements represents the makers of which style of engraving is a

problem which will eventually be solved only by elimination and

the study of distributions of implements and styles at a vast number

of sites. For this purpose those sites with the fewest styles will

prove more valuable than the fuller series.

The only other proof which may be available is to discover the

actual worn tools used in the making of petroglyphs and the relating

of these with their cultural associations. At Vosburg, in spite of

constant search over a period of years, I was unable to associate

any tool with its type of engraving with any degree of certainty. For

these reasons I have only found it possible to suggest the parallels

existing between the petroglyph series and the artefact series.

Farther than this I do not think we can at present go.

At Vosburg the problem is made more difficult by the absence of

any real relation between the “homes” of the prehistoric folk and

the petroglyphs. Here and there on the dolerite hills, and on the

alluvial soil between them, are small flat patches of earth, about

188 Annals of the South African Museum.

30 feet in diameter (Plate LVII, A). These are sedimentary floors

left dammed behind natural rock barriers in the process of denuda-

tion. They are generally surrounded by rocks, the tops of the

boulders forming the natural dam, which yield a slight shelter from

the wind. The prehistoric inhabitants evidently found these pan-

like formations best suited to camping, as their implements are

abundant here. These I have called Homes, as they suggest a

sojourn, showing pottery and here and there a few microscopic rem-

nants of decayed bone, in addition to stone implements (text-fig. 1).

Disregarding the find of a single implement of Fauresmith type,

the implements range from Middle Stone Age to Later Stone Age

material, and show four industries. The earliest we may refer to

the heavily patinated and weathered flakes from Alexandersfontein

Pan, Kimberley. (See Goodwin and Lowe, Ann. 8. Afr. Mus.,

vol. xxvii, p. 140.) The second does not seem to be related to any

described material, and it is perforce called the “ Vosburg Industry.”

The third is fairly typical Smithfield B material, with associable

pottery. The fourth seems to le between Smithfield B and C. It

has a retouch very similar to Mr. C. H. Heese’s Britstown material,

and resembles material found by me on Moonlight Kop, Victoria

West. (See Goodwin and Lowe, op. cit., p. 247.) The advanced

forms found by Mr. Heese are missing from both this and the Moon-

light Kop sites.

EASTERN Home.

There is a small pan-like depression, of the type described above,

near an outlined horse on the Eastern Outlyers Hill. This home is

fairly well sheltered and consists of the usual earthy platform,

nestling between the two ridges cresting the long hill where the

engravings are found. It commands a magnificent view of a vast

plain to the eastward. The home is not rich, and the general assem-

blage is not of much value. The implements consist of half a dozen

heavily patinated flakes in lydianite, assignable to the Alexanders-

fontein Variation of the Middle Stone Age. With these are eight

Smithfield flakes and tools, dulled by time but not patinated. They

belong to Phase B of this industry and are also lydianite.

Alexandersfontein Type.

A crude point with faceted butt. The implement ends in a “nose”

and is coarsely trimmed.

A long triangular sectioned point, similar to a willow-leaf type,

Vosburg : Its Petroglyphs. 189

with crude secondary trimming. This specimen shows usage at the

blunted point. The butt is not visible.

A crude tool, flaked over both faces, probably a fragment of a

disc, with rough edge-trimming. No butt visible.

A flake with faceted butt.

Smithfield Type.

One disc, may belong to an older period.

Portion of a circular scraper.

One heavy end-scraper.

One straight-edged scraper.

Four unworked flakes.

Home Site I.

Situated on the alluvial soil at the foot of Burkitt’s Hill and north-

west of High Rock, this home is of greater interest. It yielded a

fragment of a bored stone which must originally have been ovate,

measuring about 13 cm. by 11 cm. It is bored through the longer

axis. The material seems to be a relatively soft dolerite. With this

were fragments of pottery of a sort typical of several of these homes.

It measures a centimetre in thickness, the outer surface is brown,

while the bulk of the clay has been burned black. The inner face

shows certain quartz inclusions, appearing as fine and not very

numerous grains of white. These quartz inclusions are generally

finer than those visible in the coastal pottery, though one or two

quartz fragments as large as }-inch in diameter are also visible. More

important, perhaps, are the impressions of grass visible on the inner

face of the sherds, and also within the walls of the pot. These are

probably a result of the mode of burning. The outer face bears a

regular pattern, indented or impressed before the baking of the pot.

The tool used seems to have consisted of a stylus or type-stick, bearing

a row of at least three raised squares, the type-face measuring 9 mm.

by 14 mm. This, when impressed on the pot, left three or more

indentations, each measuring about 2mm.by 14mm. This unit was

repeated in parallel to give much the effect of impressed basketry.*

One small fragment shows a slight variation in the tool used, while

another shows that the design was not imprinted over the whole

pot. A few fragments of wind-worn ostrich shell were found,

* Compare Hewitt, Trans. Roy. Soc. S. Afr., vol. xix, 1931, plate xiii.

190 Annals of the South African Museum.

together with splinters of animal bone altogether too meagre to be

of any value.

About forty stone implements were collected here. They are

generally dulled and greyed, but otherwise not greatly patinated.

End-scrapers, cores, a single notched scraper, and two side-scrapers

suggest the Smithfield industry. For the main part the assemblage

suggests phase B, but one or two specimens may denote a somewhat

earlier style than the others. One specimen shows the coarse work-

manship ascribed by Dr. van Hoepen to his Konig culture (E. C. N.

van Hoepen, “Die Konigse Kultur,’’ Argeologiese Navorsing, vol. 1,

Bloemfontein, 1928). Apart from. these uncertain elements, the

assemblage includes the usual small cores and fabricators of the

Smithfield B. With these appears a single burin of the Vosburg

industry. It is patinated a leathery brown, and has an uneven butt

suggesting a prepared striking platform, and has been re-sharpened

twice. The extreme point shows two transverse flake-scars a milli-

metre square, either the product of trimming or usage.

Vosburg Type.

One burin, 4:8 cm. by 1-9 cm. by -6 cm.

Smithfield Type.

Fifteen end-scrapers ranging in size from 3 cm. by 1-2 cm. by -6 cm.

to 8-5 cm. by 3-7 cm. by ‘5 cm.

Fragment of a bored stone.

A disc, used for hammering.

Two flake cores.

A redirecting flake.

A “tea cosy,” probably half a broken disc.

A circular scraper.

A triangular scraper, similar in type and technique to the usual

square and circular forms.

A burin.

Several potsherds and flakes.

Home Site II.

Possibly this is an extension of Home I. It lies 60 yards farther

west and on the talus of Big Rock Hill, and it is cut off from Home I

by a small erosion gulley and a modern farm road.

A few heavily patinated implements of Alexandersfontein type

Vosburg: Its Petroglyphs. 191

occur here, coarsely trimmed and generally formless (text-fig. 11, 5-8).

A few flakes, less markedly oxidised and weathered, and smaller,

belong to the Vosburg type. No tools are recognisable.

With these older implements appear two groups of Smithfield

forms. The first of these may be at once compared with the Smith-

field B material at Home I, but the presence of retrimmed flakes, a

small jasper thumbnail scraper, and a quantity of unpatinated small

tools suggest that we have here something later, perhaps even Smith-

field C. The dulled Smithfield B material suggests that this site was

at that period an extension of Home I. It shows the same typical

assemblage of end-scrapers and cores, and has in addition two

possible burins. The unpatinated material recalls Heese’s material,

mentioned above, but contains no tanged forms nor burins.

Alexandersfontein Type.

A crudely worked point, leaf-shaped, with a heavy ridge, 6-7 cm.

by 2-9 cm. by 1-1 cm.

A small core of Moustierian type.

A few flakes.

Vosburg Type.

A flake. The remainder of the implements of this type have been

retrimmed by Smithfield man.

Early Smithfield B Type.

A heavy scraper.

A burin made on a heavy flake.

A side-scraper.

An end-scraper.

A core scraper or small fabricator.

An older Vosburg flake retouched as a rod scraper.

Two patinated implements of Vosburg type retrimmed, one as a

burin, one with under-trimming, probably usage.

Late Smithfield B Type.

Three large cores.

A fabricator.

A double end-scraper.

192 Annals of the South African Museum.

Three small circular scrapers of thumbnail form.

Two square end-scrapers.

Six graver-like tools.

A pebble “ écaillée.”

Two side-scrapers.

A rod scraper.

Some thirty flakes.

Smithfield B to C.

Three thumbnail scrapers.

Three cores.

An end-scraper.

Two graver-like tools.

A few flakes.

Home Site III.

This is on a considerably larger floor than the others. It lies some

100 yards north-east of Big Rock, with a good view to the north, and

is marked by a large single cube of rock at its centre. No engravings

are to be found on this rock, nor within 100 yards of the site,

though the surface of the dolerite block would appear to have been

suitable.

A single flake from here may be relegated to the Alexandersfontein

Industry. The remainder of the material, some seventy specimens,

reflects again the state of affairs visible at Home II. The dulled

Smithfield B material is here, and also the clean, unpatinated arte-

facts of the later phase. The tendency, already observable at Home

II, for these later flakes to be generally smaller and to show long, flat,

narrow forms, is confirmed here. Thumbnail scrapers and small end-

scrapers of jasper and similar material further bear out the identity

of these two assemblages of implements. While no true burin forms

are represented, a few artefacts suggest usage as engraving tools. A

single specimen suggests the presence of Wilton elements (text-fig. 12).

Alexandersfontein Type.

One small scraper, roughly circular.

Early Smithfield B Type.

A flake, later retrimmed as an end-scraper.

Vosburg : Its Petroglyphs. 193

Fie. 11.—Pottery and implements from Homes I and II.

1. Two views of indented pottery. Home I; 2-4. End-scrapers. Home I;

5-8. Flakes of Alexandersfontein Variation. Home II; 9-10. Smithfield rod-

scrapers. Home II; 11-14. Smithfield end-scrapers. Home II; 15. Two views

of a Smithfield core. Home II.

194 Annals of the South African Museum.

Late Smithfield B, and Smithfield B to C Types.

Three end-scrapers.

An end-scraper, showing crude working on the under-face. It seems

to have been a chisel, resembling European Neolithic chisel-forms

(8 cm. by 2:2 cm. by 1-6 cm.).

Two discs.

A core.

Two side-scrapers.

A heavy circular scraper.

A redirecting flake.

A small battered jasper fragment.

Thumbnail scraper of agate.

A straight end-scraper of agate.

A crescent-like bead borer of lydianite.

Over a hundred flakes.

Wilton Type.

One small backed point or crescent of silcrete.

Home Site IV.

This is another pan-like floor, about 100 yards south of Home III

and 60 yards east of Big Rock. Again there are no petroglyphs

in the immediate vicinity.

One specimen belongs to the Alexandersfontein Variation. It is

steeply worked all round with a crude heavy technique. With this

are some thirty flakes and tools of the Smithfield B type. They

are related to the material described above.

Alexandersfontein Type.

One backed side-scraper.

Smithfield B Type.

Two discs.

Three side-scrapers.

Two cores.

Two heavy end-scrapers.

A small side-scraper with under-working or usage.

Six small end-scrapers.

A score of flakes.

Vosburg : Its Petroglyphs. 195

Home Sire V.

This is in the usual hollow between rocks and affords a good view

to the north. It is 80 yards north-west of Big Rock.

About a dozen artefacts from here may be regarded as belonging

to the Alexandersfontein Variation. They show the usual crude

workmanship, and in this larger assemblage they show more markedly

their origin in a developed Moustierian. Small flakes struck from

disc-shaped nuclei and a few of the cores of this sort are represented.

One of the flakes has been retrimmed by Smithfield man as an end-

scraper. Some forty specimens represent the Vosburg Variation.

They are again patinated to the leathery brown visible at other sites

and are unweathered. Faceted butts occur, but are not general, and

the flaking shows a thorough knowledge of the “Neanthropic” or

Capsian parallel flaking technique. The flakes are generally small,

and some suggest use as gravers, though no true burins have been

observed from this site.

Two or three specimens show similarities with the earlier Smithfield

A-B material found at Home I. Here again this is followed by the

normal Smithfield B material, represented by about sixty specimens,

and divisible into three patination groups. It corresponds exactly

with the comparable material at Homes I, I, III, and IV. Once again

thick pottery is associated. It is baked black, and bears indented

designs and quartz inclusions and again shows the presence of grass.

The type of design is identical with that from Home I, but a different

stamp seems to have been employed. Here the “basketry”’ rows of

imprint have been arranged to meet in a chevron.

The later Smithfield material which I have called B-C is again

represented, but at this site the specimens collected have additional

associable material. Three fragments of a polished (probably natural)

pebble which has been used as an anvil (text-fig. 12, 7), an ostrich

egg-shell bead and several shell fragments, two stone bead-borers,

two fragments of a bone point, and some pottery, all appear to

belong to this later group. The pottery which I associate with

this phase is thin (6 mm.) and has no signs of design on the

outer face. It shows no grass inclusions, is reddish in colour, and

is better preserved than the older pottery. It is important to

note that the association of these last objects rests upon the state

of preservation and the distribution of the forms between various

Homes.

VOL. XXIV, PART 4. 14

Fic. 12.—Pottery and implements from Homes III, V, VI, and the Painted Shelter.

1-3. Flakes. Home III; 4. End-scraper. Home III; 5-6. Small scrapers of

jasper. Home III; 7. Polished or water-worn stone. Home V; 8-11. Bone

arrow-point, ostrich egg-shell bead, and two bead-borers. Home V; 12. Indented

pottery. Home V; 13. End-scraper. Home V; 14. Flake of Alexandersfontein

type. Home VI; 15. Burin, Vosburg type. Home VI; 16. Flake used at point,

Smithfield type. Home VI; 17. Used Smithfield flake. High Rock; 18. Used

water-worn fragment. Painted Shelter; 19. Indented pottery. Painted Shelter;

20. Jasper thumbnail scraper. Painted Shelter; 21-22. Lydianite end-scrapers.

Painted Shelter.

Vosburg: Its Petroglyphs. 197

Alexandersfontein Type.

Two Moustierian end-scrapers.

One Moustierian core.

Ten flakes.

Vosburg Type.

Two flakes of Alexandersfontein type retrimmed.

Six cores, two elongated in form.

One polyhedral baton.

One flake showing a burin blow struck from the butt end to include

part of the cleavage face.

Two circular scrapers.

A redirecting flake.

Two flakes worked on the cleavage face.

An end-scraper.

Thirty-six flakes.

Early Smithfield B.

Three end-scrapers, one large.

Two fabricators.

A point or chisel.

A rod scraper.

A redirecting flake.

Two side-scrapers.

Three core scrapers.

Half a dozen flakes.

A dozen potsherds.

Late Smithfield B.

An early Smithfield B flake retouched as an end-scraper.

One small and one large flake, with under-trimming.

A large circular scraper.

A large triangular end-scraper.

Two rod scrapers.

Five cores.

Two burins.

Two graver-like tools.

A flake “écaillée.”

Three end-scrapers.

Medium circular scraper.

A dozen flakes.

198 Annals of the South African Museum.

Smithfield B to C.

Two fabricators.

A large disc.

Three end-scrapers.

A redirecting flake.

A circular scraper in jasper.

A crudely worked triangular end-scraper which may be a point.

Some sixteen pottery sherds may belong to this or to the previous

Group.

Home Site VI.

This site is about 50 yards south-west of Home V and 100 yards

west-north-west of Big Rock. It is in a similar pan-like depression,

and again no petroglyphs are to be found in immediate association.

One or two specimens are assignable to the Alexandersfontein

Variation. Several others, including an excellent burin which has been

frequently retrimmed (text-fig. 12,15), belong to the Vosburg Industry.

The remainder consists of Smithfield B material, and the usual thick

pottery, with design and grass inclusions. With these were a lower

and an upper grindstone (Plate LIX).

Alexandersfontein Type.

Three points on thick flakes.

Vosburg Type.

A burin and a couple of flakes.

Smithfield B.

A crude end-scraper.

A grindstone (upper and lower).

A few flakes.

A dozen potsherds.

Home Sire VII.

This lies in the valley between Burkitt’s Hill and Big Rock Hill,

150 yards south-west of Home II and on the talus of Big Rock Hill.

Once again there are no petroglyphs near. A hundred yards farther

south of this site was found a single Fauresmith cowp-de-poing,

unassociated and alone. Apart from this individual heavily rolled

specimen of lydianite, the site shows the same series as the others.

Vosburg: Its Petroglyphs. 199

Weathered Alexandersfontein, patinated Vosburg, early Smithfield

B, and the clean, unpatinated late Smithfield types all recur.

Alexandersfontein Type.

A point with a slightly faceted butt.

Vosburg Type.

Four nondescript flakes.

Early Smithfield B Type.

An Alexandersfontein flake retrimmed.

A small core-face used as a chisel.

A large end-scraper.

A tabular shale end-scraper, triangular in section.

A flake “écaillée.”

A pointed flake showing use.

A small end-scraper.

A dozen flakes.

Late Smithfield B Type.

Two used graver-like tools.

Two medium end-scrapers.

One small end-scraper.

Two circular scrapers.

One side-scraper.

Two notched scrapers, one on a freak flake.

A fragment of jasper.

Some forty flakes.

Hicgu Rock Home (Burxirt’s HI).

This is the “shelter” referred to by Burkitt (see above) and little

is left there now. Smithfield B artefacts are still to be found, together

with the less patinated B-C phase material which again includes the

later pottery, and a few scraps of disintegrating bone (Plate LVII, B,

and LVIII, A).

(Additional to Burkitt’s material.)

Vosburg Type.

Two pointed end-scrapers, plain butts.

Two cores, plain platforms.

A refacing flake.

A dozen flakes.

200 Annals of the South African Museum.

Early Smithfield B Type.

A large core.

Two small cores.

A dozen flakes.

Late Smithfield B Type.

Side-scraper.

Rod scraper.

A fragment of thin pottery.

A few flakes.

Bic Rock Home.

This lies about the base of Big Rock (Plate LVIII, B). It is fairly

scattered, extending some 20 yards in all directions, but is rich.

Some fifty artefacts found in this area belong to the Alexanders-

fontein Variation and do not differ from the material described above.

A hundred or so which were found immediately about the base of

Big Rock belong to the Vosburg Industry. These, though small,

show themselves to have closer morphological affinities with the

Capsian of North Africa than with any South African culture. Tools

of a recognisable conventionalised type are unfortunately rare. Most

common are a number of batons or core-like tools, possibly gravers

of a sort (text-fig. 13, 12). One tool shows serration (text-fig. 13, 17),

and two burins are represented (text-fig. 13, 1o and 11). Here

Smithfield B and B-C tools are represented but meagrely.

Alexandersfontein Type.

Six points, some broken.

Eight nosed-points.

Nine discs, usually with one face flattened, and generally 2-5 cm.

diameter.

Seven end-scrapers.

Three side-scrapets.

A possible burin.

About thirty flakes.

Vosburg Type.

Seven cores.

Two long cores.

A triangular point.

201

Vosburg: Its Petroglyphs.

g@ Rock.

o Rock.

Fic. 13.—Flakes of Alexandersfontein and Vosburg types.

y Rock.

Big

1-9. Flakes of Alexandersfontein type.

10-18. Flakes of Vosburg type.

202 Annals of the South African Museum.

A point used for pecking.

An end-scraper.

Five redirecting flakes.

Two discs (2-5 cm. diameter).

Five burins.

Two used gravers.

Sixty flakes.

Early Smithfield B.

An end-scraper and four flakes.

Home VIII.

An unimportant Home situated 200 yards east of Distant Rock

on the slope of the hill.

Alexandersfontein Type.

One large burin.

Early Smithfield B.

One end-scraper.

A crude scraper, broken.

Five flakes.

Late Smithfield B.

Two end-scrapers.

An early Smithfield B flake, reused.

Two flakes.

PAINTED SHELTER.

The little Painted Shelter, measuring only a few feet across and

as few in depth, and almost completely unprotected from rain, has

nevertheless yielded a fair number of implements. The floor deposit

is less than an inch in depth, and hes directly upon dolerite blocks.

From the talus below the shelter come a few older implements,

though whether they were ever in the shelter it is impossible to say.

One is a flake of Alexandersfontein type, a few others are of Vosburg

type.

Within the shelter, and immediately about it are numbers of

Smithfield B implements. The usual associable pottery is found,

and a single fragment of broken, water-worn lydianite, very similar

Vosburg: Its Petroglyphs. 203

to the anvil at Site V. There are no signs of any earlier or later

inhabitants of the shelter.

In the general region of the Painted Shelter a number of additional

implements were found.

Alexandersfontein Type.

One point.

Vosburg Type.

A dise and two flakes.

Early Smithfield B (mainly from the shelter).

Four fragments of thick pottery with grass inclusions.

A crude side-scraper.

Eight cores.

Five fabricators.

A battered hammer-stone.

One naturally polished stone, slightly used.

Two end-scrapers.

A dozen flakes.

Late Smithfield B.

Four cores.

A large end-scraper on an earlier flake.

Five small end-scrapers.

A jasper side-scraper.

A jasper thumbnail scraper.

A disc.

A trimmed flake.

Two unworked flakes.

THe MOTIVES FOR THE ART.

There are various approaches to the general subject of prehistoric

and primitive art. We may adopt the “Art for art’s sake” attitude,

divorce the piece of art entirely from any social or historical surround-

ings, and judge it by our own conventional standards. We may

perhaps seek an ulterior and sublime motive behind the art, isolating

religious from profane art. Prehistoric art as seen by the civilised

eyes is labelled “‘sacred object—object of cult—magic,” etc., while

the places where such arts are practised become sacred spots.

204 Annals of the South African Museum.

Our own approach is based upon an attempt to be scientific. It

is intended to prove historical sequences between styles of art, and

to date these relative to deposits. From the petroglyphs themselves

certain deductions have been made; they have been analysed into

styles and techniques; an attempt has been made to analyse the

animals and objects depicted, and attention drawn to those peculiar

physical characteristics shown on Big Rock, and probably typical

of the artists of that period.

Each of these approaches has its uses. In some of the petro-

glyphs a true artistry can be seen, most notably in the pair of heads

on Big Rock. It would also be interesting to know why the artists

drew at all, but our knowledge is insufficient to justify our labelling

all these petroglyphs as religious or magical.

In a survey such as this, where a single site is taken into considera-

tion, certain questions must arise, perhaps more or less particular to

the site, perhaps of general application. We have already dealt

with some of the environmental factors, we have seen, for instance,

that fences affect the modern environment fundamentally, by pre-

venting the normal game and human migrations of this region. We

have already noted that the presence of a population too great for

the region has sucked up the underground waters and lowered the

water-table. Environment works in many other more subtle ways,

and with some such effects we may deal here.

Let us look at these petroglyphs in their environment. We have

an arid stretch of country, 60 miles from the Orange River, and thus

60 miles from the nearest regular water. In a wide plain we have

the Keurfontein, a perennial spring. Immediately overlooking the

Keurfontein are a few small hills, but without any petroglyphs or

home sites; quarry sites, showing unworked flakes, crusts of lydianite

and so on, all belonging to Smithfield B patination, but without a

single implement, are relatively common.

Farther back are other hills, still overlooking the water-hole, but

separated from it by a mile of plain. Here we have a number of

homes, generally with a view of the spring. This is the region of

petroglyphs.

Among the living Bushmen we know that the household or horde

never camped near water. To do so would frighten away game and

defeat the hunter’s ends. A not-too-frequent visit to the water-hole

or a nearby quarry site would be safe enough, but the camp must be

some distance away. What better spot could be chosen than a

raised plateau or hillock within view of water? From here the game

Vosburg : Its Petroglyphs. 205

could be observed as it made its migrations annually to the Orange

River. The camp would not be permanent, but it would be used

annually during the migratory season over a period of years. The

Keurfontein is set in a region of aridity, which had a sparse animal

population, supporting a few hunters whose main emotion was

hunger. Prehistoric man was hungry, the hunger gnawed at him

and became more and more insistent. He found that his hunger

demanded expression in action. Set as he was in dolerite hills,

watching over the plains for game, he had time for action. He needed

food, and if game was scarce he drew food on the rocks, and so vented

his feelings. Exactly the same primitive and simple reaction is

seen in the pornographic scribblings which occur in our own society—

though here food-hunger is not the predominating emotion.

This emotional satisfaction of a need must have been very basic

to primitive art in its early stages. It may be followed by a second

step, an attempt to rationalise this apparently irrational action,

“Why do we draw when we are hungry ?’’—‘‘ Because we want to

gain power over our food supply.” Once this belief is accepted and

socialised we have reached sympathetic magic. From this belief

may arise other things. The morale of the hunter may be increased;

or the belief may arise that to depict a thing is to give power over it,

therefore it is “sorcery” to depict a man. In such a development

there are three clear steps: wish fulfilment, rationalisation, and finally

a fear reaction. No religious element has entered the field. There

is no personification. The power is there to be evoked by the act

of drawing ; it is not “given” by a higher being. |

While this simple series of psychological stages will account quite

happily for the greater number of the petroglyphs at Vosburg, it is

certainly not to be extended to all of these, nor to much of the pre-

historic art in Africa and elsewhere. There are other developments

present.

On Big Rock ideas are being drawn. The snakes, the grids, the

concentric circles, the queer mythical beaver-like animal, and the

devil himself, do not fit in with this theory. This art is more sophisti-

cated, though often more crude, than the pure animal designs. We

may draw a parallel between the figures on Big Rock and the animal

and symbolic figures shown at initiation ceremonies all the world

over. The “devil” is certainly a dancer, with animal horns and a

tail, similar to dancers described by Miss Bleek (D. Bleek, The Naron,

Cambridge, 1928, p. 23) among the living Bushmen :—

“When a girl reaches maturity the eland-bull dance is held in

206 Annals of the South African Museum.

her honour. All men and boys leave the village, save two grand-

fathers, who tie elands’ horns, or wooden imitations thereof, to their

heads. . . . The Auen hold this dance in the same manner. The

Nusan recognised the tune. A similar ceremony takes place among

them, a man with a bird’s beak on his head taking the eland bull’s

place.”’

Quite apart from the action of any tribal or transmitted tradition

which bound these generations of artists together, there is a very

potent and interesting factor which has affected these sites at Vosburg,

and almost certainly had its effects at other sites. This we may

call a “locality tradition.” It manifests itself here as a strong

desire on the part of every visitor to the site to imitate and emulate

the older artists. Directly a herdboy, or even the more educated

farmer, sees the petroglyphs, he will, without a moment’s thought,

attempt to produce something of the same type. The drawings

themselves thus form a basis for tradition. How this has been met

by various individuals can be seen in the palimpsests, especially the

two ostriches on the Paintings Hill superposition. Modern drawings

of farmers on horseback (on horses, be it noted, with the traditional

neck of the thoroughbred and hence probably recent), some with

guns, the ostrich hunts, the Victorian clothing, and the final phase

of the lovers’ hearts and initials, are all confined to just those areas

in which the engravings are found. A hundred yards away they do

not exist.

This locality tradition is thus strong to-day. It is a form of

direct suggestion, and while the more educated generally repress it,

most visitors, when I was at work, amused themselves by adding to

the series.

How far did such suggestion create or foster native “tradition”?

It has easily bridged the gap between the native mind and that of

the European—what other gaps between earlier races were similarly

bridged? It is obvious that tradition acting through transmitted

word of mouth is not an essential. The practical example of the

many designs already present is sufficient to call out any innate

artistry in the newcomer.

This would seem to suggest very strongly that the peoples who

were responsible for these various petroglyphs were not necessarily

bound by a common tribal or cultural tradition, and also that they

were not necessarily inspired by the same motives.

Vosburg: Its Petroglyphs. 207

DEFORMATIONS, TECHNIQUE, AND ExtTINcT SPECIES.

From time to time attempts are made to use the fauna depicted

in petroglyphs as fossil evidence from which the art and the dying

fauna can be correlated. A tertiary fossil is known, and the bones

form the basis of a reconstruction which is similar to the form of an

animal pecked on a rock-surface. The main source of error is the

failure to realise how crude and rudimentary the various techniques

employed in petroglyphs may be. It is only in the hands of a few

outstanding artists that species can be made recognisable, and the

measurable variation in two similar animals in like styles on adjacent

rocks may be very considerable.

Another peculiar element at Vosburg is the presence of the quagga

petroglyphs (of Style 34), which have had elephant-like trunks added

in the same technique by an artist of the same period. The quagga,

being a striped animal, is easily recognisable, but the addition of a

trunk throws an interesting light on several petroglyphs which

supposedly represent Tertiary fossils. Some, at first sight, certainly

suggest hippopotamus or rhinoceros with trunks added. The likeness

to accepted reconstructions of prehistoric animals may have been quite

unconsciously increased by the process of adding oil or graphite to

“bring out’ the petroglyph, and to conceal chippings or scratches

which were “not intended” by the original artist.

It seems important in a question of this nature, where an extinct

species is concerned, to retain the surface of the rock in its natural

condition without any sort of preparation for photography. A petro-

glyph ceases to have any scientific value once it has been tampered

with, and the patina, technique, and form have been touched up.

The acceptance of a petroglyph as depicting an extinct species

becomes less and less justifiable as the time-gap which has elapsed

since the fossil lived increases. No zoologist would accept a new

species described from the evidence of a petroglyph alone: why should

he accept such proof as post-dating the existence of an extinct form ?

The question of technique is important. Any artist will admit to

the relative inaccuracy of oils compared with water colours. Any

sculptor will admit to the distortion of texture and form shown in a

piece of work executed with coarse tools. In dolerite other factors

come in: the textural inequalities of the rock and the unevenness of

the rock-surface. Of the crudeness of the stone tools with which the

work was done, little need be said, as workers in stone tools have

often risen to heights hardly equalled among our own craftsmen.

208 Annals of the South African Museum.

SUMMARY.

If we exclude the single find of a Fauresmith couwp-de-poing of

lydianite, in the region of Home V and some distance from the

engraving, we are here presented with at least four cultural phases

evidenced by stone implements, and with eleven stylistic periods

shown in the petroglyphs.

The earliest of the implement phases is the Alexandersfontein

Variation, an ugly development on a Moustierian theme. Apart

from the neighbourhood of Big Rock, specimens are relatively

scattered and seem not to form an integral part of any one congeries.

They are spread more or less homogeneously over the whole region,

so that within any area the size of a home site one or two flakes of this

type will be found. If this distribution can be taken as evidence of

disturbance by slow water-action, a long lapse of time is again

suggested since the making of these tools. Much the same may be

said of the material relegated to the Vosburg Culture, which is less

evenly distributed, though quantities of these implements only occur

together about Big Rock and Home V.

At all the home sites, save Big Rock, occur localised groups of

Smithfield B implements, or of Smithfield B-C type and technique,

or, In some cases, both together. At Big Rock both periods are

relatively rare, and do not suggest that the makers of these imple-

ments lived near this rock. The general distribution of Smithfield

implements shows that these hills were regular stopping-places for

these folk. The hills command a clear view of the Keurfontein, and

are sufficiently far away not to scare game from making use of that

spring.

From these facts we must presume that Smithfield man made the

greater number of these petroglyphs. It must be remembered that

the makers of the Alexandersfontein and Vosburg types of tool were

here before them, and that the Vosburg congeries include burins and

also stones which show signs of having been used as gravers. It is

not possible to link the Alexandersfontein implements with any type

of engraving, but the localisation of the Vosburg artefacts in the

region of Big Rock strongly suggests that these people made some of

the earlier Big Rock engravings. To these Vosburg folk may be

tentatively attributed my Style 2 (Burkitt’s first phase), represented

by the incised eland.

The next clear association is given us at the Painted Shelter, where

the lightly pecked petroglyphs of Style 6 are associated with paint-

Vosburg: Its Petroglyphs. 209

ings and with Smithfield B implements as certainly as any elements of

culture can be.

After this period the Smithfield B—C folk take up the task of decor-

ating the hills, until the time marked by the appearance of metal

spearheads (Style 10) and finally of our own facetious efforts.

My thanks are due to the Research Grant Board for their help on

my second visit to Vosburg, to my wife, to J. G. Taylor, to B. D.

Malan, and to Mr. and Mrs. Burkitt, for their help on the various

visits. Throughout my work I became more and more indebted to

the various members of the Vos family, most especially to Mrs. Vos

and to Mr. Henry Vos; to them, therefore, I dedicate this paper.

210

XXXII.

XXXITI.

XXXIV.

XXXV.

XXXVI.

XXXVII.

XXXVIII.

XXXIX.

XL.

XLI.

XLII.

XLIV.

XLV.

XLVI.

XLVII.

XLVIII.

XLIX.

. Rhinoceros, Style 64, Eastern Outlyers.

. The same, showing mode of occurrence on a cloven rock.

. An eland, Style 9, Eastern Outlyers.

. Buffalo, followed by a man.

Annals of the South African Museum.

EXPLANATION OF PLATES.

Petroglyphs on Big Rock, flattened to show the relationship. Re-

construction from tracings.

Eastern face of Big Rock. Photograph.

North-eastern portion of Big Rock. Photograph.

Northern portion of Big Rock. Photograph.

Two heads, buck and giraffe, north face of Big Rock. Photograph.

“The Devil,” south-western face of Big Rock. Photograph.

Palimpsest on Paintings Hill. Photograph and tracing.

Palimpsest on Rock 6. Photograph and tracing.

A. General view on Big Rock Hill.

B. Palimpsest on Rock AK. Tracing.

Two views of design in Style 1, Rock AL. Photograph.

Animals in Style 3, from Rock 30. Compare Burkitt.

and tracings.

A. Ostriches in Style 3, Rock AO. Photograph.

B. Quagga with trunk added, Style 34. Photograph.

Quaggas with additions, Rock AF, and a neighbouring rock.

Young quagga, bearing light (and later) graffiti on the body. Probably

Style 3B.

Deeply pecked quagga, 100 feet west of Big Rock. Roveused tracing.

Eland on Rock 14, Style 3s.

A “Dancing Buck,” Style 3c, Rock AN.

Buck in Style 6. Human figure above, probably in Style 44, Rock AE.

Photograph

The lower figures are later. The buffalo

is in Style 7, or more probably 9.

. General view of the Painted Shelter from below.

. The Painted Shelter, showing ruiniform rocks.

. The paintings on the rock.

. A. A typical Home Site (Home V).

B. General view of High Rock, Burkitt’s Hill, from the east.

. A. View of High Rock.

B. View of Big Rock and site.

. View of Home VI, showing upper and lower grindstones.

. Panorama from Painted Shelter, including Keurfontein homestead,

the Eastern Outlyers, Burkitt’s Hill, and Big Rock Hill.

XX XIT.

Plate

XXIV.

Mus., Vol.

Afr.

Ann.

‘dIHSNOLLVTHY MOHS OL GANALLV IT “MOOU DIA NO S

WWW

dX'T

VOULAd

Lid.

Neill & Co.,

Goodwin.

Ann. S. Afr. Mus., Vol. XXIV.

Goodwin,

Plate XXXII.

Neill & Co., Ltd.

BIG ROCK.

FACE OF

EASTERN

Plate XX XI

Naa

iea

—

_—

—

~—

fal

—

NORTH-EASTE

r UX

Neill & Co.,

Ann. 8. Afr. Mus., Vol. XXIV.

Goodwin.

Plate XXXY.

Neill & Co., Lid.

NORTHERN PORTION OF BIG ROCI

Plate XXXVI.

Ann. S. Afr. Mus., Vol. XXIV.

exerts |

fo) ke!

dl O

YON

‘

GANV MONG

TH OMA

eill & Co., Ltd.

Goodwin.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XX XVII.

Goodwin. Neill & Co., Lid.

SOUTH-WESTERN FACE ROCK.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XX XVIII.

Goodwin. Neill & Co., Ltd.

PALIMPSEST ON PAINTINGS HILL.

Plate XX XIX.

Ann. S. Afr. Mus., Vol. XXIV.

MOOUW

NO LSUSHNITVd

Neill & Co., Ltd.

Goodwin.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XL.

te,

ea SS vod

A. GENERAL VIEW ON BIG ROCK HILL.

Goodwin. Neill & Co., Ltd.

B. PALIMPSEST ON ROCK AK.

Plate XLI.

Ann. S. Afr. Mus., Vol. XXIV.

Neill & Co., Lid.

oodwin.

IN STYLE 1. ROCK AL.

SIG

OF DE

VIEWS

TWO

|} Ann. S. Afr. Mus., Vol. XXIV.

Goodwin.

Plate XLII.

Neill & Co., Ltd.

30.

‘IK

FROM ROC

ni)

STY Lk

ANIMALS

Plate XLIIT. ©

Neill & Co., Ltd.

ROCK

IN STYLE 3.

)

ce)

OSTRICHE

Ann. S. Afr. Mus., Vol. XXIV.

Goodwin,

Ann. S. Afr. Mus., Vol. XXIV.

Goodwin.

QUAGGA

WITH

ADDITIONS.

Plate XLIV.

Neill & Co., Lid.

Ann. 8. Afr. Mus., Vol. XXIV.

Goodwin.

SM.

oe ~. .

SENS

Plate XLV.

NESS

Neill & Co., Ltd.

3B.

PROBABLY STYLE

THE BODY.

GRAFFITI ON

LIGHT (AND LATER)

QUAGGA BEARING

Plate XLVI.

Ann. S. Afr. Mus., Vol. XXIV.

“VE

OUTALS

(MOON

ora

ALSO M

NCE Rul

OOT)

Voovad

Lid.

Neill & Co.,

Goodwin.

Ann. 8S. Afr. Mus., Vol. XXIV. Plate XLVII.

STYLE 3B.

Goodwin. Neill & Co., Ltd.

Plate XLVIII.

Ann. 8S. Afr. Mus., Vol. XXIV.

MOO'U

« MONA

YNIONV,,

& Co., Lid.

Neill

Goodwin.

Ann. S. Afr. Mus., Vol. XXIV. Plate XLI

Goodwin. Veill & Co., Lid.

BUCK IN STYLE 6.

‘VO TALS YOATLAO NULSVA ‘SOUMOONTHY

win.

Good

Ann. 8. Afr. Mus., Vol.

Goodwin.

XXIV.

Neill & Co., Ltd.

SAME AS PLATE L, SHOWING OCCURRENCE.

S. Afr. Mus., Vol. XXIV.

Ut TF

>» 3B

VERT NG

Ke .

ies

ATG.

RN OUTLY

KASTI

STS GR:

Bn a

< NG AN

Gooduin.

Ann. S. Afr. Mus., Vol. XXIV. Plate LITT.

STYLE 9 OR 7.

jae

ica

jaa)

[ea

jae)

ica}

)

FOLLOWED BY A MAN.

BUFFALO

Goodwin. Neill & Co., Lid.

Ann.S. Afr. Mus., Vol. XXIV.

“UChLLTHHS GUoNIVd

| H iL

Goodwin.

Ann. 8. Afr. Mus., Vol. XXIV. Plate LY.

Gaon ; 7 7 Neill 4: Co., Lid.

THE PAINTED SHELTER.

SCLHONT NI WPVOS

. —

Goodwin.

Ann. 8. Afr. Mus., Vol. XXIV. Plate LVII.

A. A TYPICAL HOME SITE (HOME YV).

Goodwin. Neill & Co., Lid.

B. GENERAL:- VIEW OF HIGH ROCK, BURKITT'S HILL,

FROM THE EAST.

Ann. 8. Afr. Mus., Vol. XXIV. Plate LVIII.

A. VIEW OF HIGH ROCK.

Goodwin. Neill & Co., Ltd.

B. VIEW OF BIG ROCK AND SITE.

SHUNOLSANTAD GHMOT UNV WHddOo ONIMOHS “IA WNOH

aosae — - 2 - a oer: ?

Plate LIX,

oll & Co., Ltd.

—

fo)

f=}

—f

Goodwin.

Ann. 8. Afr. Mus., Vol. XXIV. Plate LX.

Wie Irae

Goodwin. Neill & Co., Lid.

Ann. §. Afr. Mus., Vol. XXIV. Plate LX

Goodiein,

Neill & Co, Ltd.

PANORAMA FROM PAINTED SHELTER.

ANNALS

OF THE

SOUTH, AFRICAN MUSEUM

VOLUME XXIV.

PART V, containing :—

9. Khp Kop Cave, Hermanus.—By A. J. H. Goopwin, M.A.,

¥.R.S.8.A., Senior Lecturer in Ethnology and Archeology,

University of Cape Town. (With Plates LXI, LXII, and

7 Text-figures.)

10. South African Natwe Snuff-Boxes.—By Miss M. Suaw, B.A.,

Assistant in Ethnology. (With Plates LXITI-LXXV.)

11. Ovambo Kmves.— By Miss M. Suaw, B.A., Assistant in

Ethnology. (With Plates LXXVI-LXXXIV.)

12. Native Pipes and Smoking in South Africa.—By Miss M. Saw,

B.A., Assistant in Ethnology. (With Plates LXXXV—

XCIX and 2 Text-figures.)

Title Page and Index to Volume XXIV.

ISSUED AUGUST 1938. PRICE 16s.

PRINTED FOR THE

TRUSTEES OF THE SOUTH AFRICAN MUSEUM, CAPE TOWN

BY NEILL AND CO., LTD.,

212 CAUSEWAYSIDE, EDINBURGH.

( 211 )

9. Klip Kop Cave, Hermanus.—By A. J. H. Goopwiy, M.A.,

F.R.S.8.A., Senior Lecturer in Ethnology and Archeology,

University of Cape Town.

(With Plates LXI, LXII, and 1 Text-figure.)

HIstTory.

Kure Kop Cave is situated about half a mile or more from the coast,

directly through Hermanus itself. The hill, which is just above the

local school, is isolated from the surrounding hills, and has been

made into a small nature reserve. The cave is a long shallow open-

ing, about 100 feet long by 80 feet deep, and faces southward. It

was cut almost certainly by water action in the Table Mountain

Sandstone, and stands about 300 feet above sea-level. It is on the

same level as a marked terrace or peneplain which is visible on the

surrounding hills, and probably is a remnant of the 300-foot raised

beach. This cutting would certainly have been long prehuman.

The first attempt at any sort of scientific excavation of the deposit

was undertaken by Mr. James Drury of the South African Museum,

and we may draw fully on his report of the cave as he found it in

December 1925 (Plate LXI).

At that time the opening was in some places only a few inches

above the floor level (see Section C). There were two entrances to

the cavern, and these led into a fairly lofty cave seven feet high.

The cave had been inhabited, some 40 or 50 years before his visit, by

a poor-white or coloured family, and had been divided by them into

two rooms, by means of a wall built from front to back of the cavern

between the two entrances. The cave wall had also been colour-

washed by these folk. This family must have cleared out nearly

all the rear portion of the original deposit to make the rooms more

lofty. They did, however, leave the front portion of the deposit

untouched. The back of the cave showed marks suggesting that

the deposit had at that point been two feet deeper than Drury found

it. This is also partly to be accounted for by the removal of some

of the material by farmers for guano.

The side and back walls of the cave are smooth vertical joints in

VOL. XXIV, PART 5. 15

212 Annals of the South African Museum.

the rock, and the roof is similar, consisting of a plane, here and

there cracked, but otherwise almost unbroken. At the back of the

longest of three natural passages leading into the hill the roof rose

to a dome, some 15 to 18 feet high in Drury’s time. The floor was

quite flat, and here and there shell, etc. from the deposit had been

thrown aside to level the floor.

Mr. Drury sank a few test holes in the deposit, and found from

these that in most of the cave there was only a few inches of deposit,

overlying a deep layer of sand. At the one entrance his natives

started a trench to enlarge the entrance and make access easier.

The section at the entrance (see plan) was found to consist only

of shell, ash and stones, about three feet thick, overlying white sand

and stones. Meanwhile he commenced a test at the western end

of the cave, and found that the deposit was somewhat thicker. A

proper entrance was made from outside, and excavation started.

The natives now cut a deep entrance at the western end, and

found that the occupational layer was deeper as it went back into

the cave (Section B). The top layer consisted of a very dark sandy

soil intermixed with a few shells, stones and animal bones. At its

greatest depth, towards the centre of the section, this layer amounted

to about three feet. The next layer was not quite so dark, and con-

tained a quantity of crushed and broken shell mixed with stones

and a few bones. Below this layer, and almost in the centre of the

section, was a band of white burnt shell ash, about three inches thick,

and covering a circle about three feet in diameter.

The next layer was not unlike the previous one, but contained

more shell, bones and stones. It was in this deposit that almost

all the finds of interest were made.

This layer started immediately below the lip of the cave, and dipped

steeply to form a bowl-like deposit. In the bottom of this was a

mound of loose shells. In this mound was found a lower jaw and

fragments of the cranial portion of a skull. The human jaw was in

the front part of the mound, and about ten feet from the cave lip.

The skull was farther back, and more on top of the mound. Under-

lying the mound was a band of fire-ash extending over most of the

floor of the bowl. In this Mr. Drury found a bored stone, about 15

feet from the entrance of the cave. Under this fire-ash was a layer

of red sand which lined the bowl, thickening towards the back of

the cave.

A hole was sunk from the bottom of the bowl, and it was found

that the underlying floor consisted of soft white sand, similar to that

Klip Kop Cave, Hermanus. 213

encountered in the higher parts of the cave. No indication of rock

was found at a depth of 2 feet. After further excavation in the bowl,

“”

= 3

DE VLLL DRURY. 1925.

§ i) x Li RXQQ YY GooPwin. /935-

2 =x SV i

a =

a Ee 4 i EE RECENT WALL

pei

A Vg aN ot

tie N VBS Zo |

ae ir

Hy — - -N- ~ gh wh 3 _---SH

ZG 2 FS

2 m

5 SCALE OLAS eNO. 20.Ft.

SAND

SECTIONS A&C

Oo°O O —

2002 LEAK os3 = =,

gost dS =” SECTION B

ae srt.

| restr \ 1 vest

i} ‘

another shaft, 6 feet deep, was sunk with the same result. The

white sand was soft and showed hollows, as though water had

214 Annals of the South African Museum.

percolated through the deposit. The test hole was in danger of

caving in and was abandoned.

Drury’s finds were few. He was surprised to find so little material

showing use. Nearly all the more interesting finds came from the

bottom of the deposit. There were some half-dozen grindstones,

showing little use: flakes and sections of round pebbles; four bone

bodkins or arrow-points; a bored stone; and a long, slender, pointed

bone from towards the surface immediately under the top black

soil towards the cave rear. Animal bones were uncommon. He

recognised hippo, eland and some small buck, seal, a few fragments

of tortoise-shell, a few albatross and duyker bones, and a single

fragment of ostrich egg-shell.

His report ends: ‘These people must have lived almost entirely

upon shellfish, as the quantity of animal matter was exceedingly

small for the amount of shell deposit. There were a few pieces of

whalebone, and what might have been elephant. Practically all

these finds were in the vicinity of the loose shells in the centre of the

cave at the lower level.

“The front bank of talus or overflow consisted of shell, stones and

ash, and covered the whole width of the cave front right down to

the foot of the slope forty feet below. This deposit was up to three

feet in depth in places, but thinned to nothing at the bottom.

“T put in four days excavating Section B, which measured roughly

30 feet in length, along the side of the cavern, and 20 feet wide

at its greatest width. The greatest depth was 12 feet, and the base

of the human deposit was just over 8 feet from the surface.

“This cave has certainly been of interest, as I have never yet.

found a cave containing so much deposit but so few used tools. It

reminds me more of an outside midden deposit than of a cave. The

small quantity of animal remains makes me wonder if there was any

scarcity of animals at this place, or were they just too lazy to hunt?”

Since 1925 excavation has been carried on by Mr. W. H. Patterson,

of Hermanus, who has been in charge of the nature reserve. His

work has been carried on slowly through the years, and he has had

every opportunity of watching for material in the deposit. All the

material he has excavated has been passed through a sieve with

+-inch mesh, and debris has been removed from the cave, so that

all that remains is stratified human deposit, overlying disintegrated

sandstone (Plate LXII). This careful removal of useless shell and

bone material has made recent work very much more simple. In

the course of his excavations Mr. Patterson has discovered a number

Klip Kop Cave, Hermanus. 215

of useful objects, none of very remarkable value, but all located with

certainty.

Mr. Patterson discovered a few human bones, disturbed and in

a bad state of preservation. I am led to suspect that these were

parts of the same skeleton found in a disturbed state by Drury.

This seems to have been the only interment in the cave.

Stone implements, apart from the usual broken and split pebbles

used for breaking shellfish, etc., were rare for the size of the cave.

There seems little reason to believe that specimens were missed by

Mr. Patterson, as the portion of the deposit dug by us was as poor.

In his collection are three grooved stones, all with multiple

eroovings. One specimen is conical, and shows fine regular bead-

eroovings. All seem to have been bead stones. Three oblong

rectangular “palettes” of slate, probably 4 or 5 inches long when

intact, were found. All are broken. One measures 2 inches across,

the other two measure 2} inches across. Judging from our knowledge

of the present Bushmen, these are very probably shovels used to

remove earth from edible roots. A fourth palette is intact, and is

of interest, as one end has been tapered slightly by crude chipping

on the edge of the slate to form a bat-like shovel or scraper, the

thinned end forming the handle.

In addition to the bored stone found by Drury, Patterson found

three other specimens. Two are flattened ovates, the third is a true

ovate, the boring being through the greatest length. With these is

a long bored-stone rimer. It is of rough couwp-de-poing form, and

shows signs of usage towards the point. (A similar specimen from

a cave above Wellington is in the Stellenbosch University collection.)

A few nondescript flakes, probably chance spalls from pebbles and

hammer stones, make up the total of stone implements.

Bone material is present. A bone tube, made from a suitable

section of bone, measures 24 inches long by 4-inch in diameter.

With this is an ivory tube, 4 inches long, down which a ;3,-inch hole

has been bored, apparently with a hot metal rod, a technique inviting

considerable speculation. Four bone arrow-points, a bone awl and

a few fragments of used bone complete the collection.

In addition Mr. Patterson noted the following raw material—two

hippo tusks, a few balls of soft white clay containing no sand grains,

possibly kaolin from a disintegrated granite deposit, or perhaps a

sedimentary white clay from shale deposits. Among the faunal

material are several lower jaws of porcupine. The remainder of the

bones suggest small buck, sea-birds, ete.

216 Annals of the South African Museum.

PRESENT EXCAVATION.

The present excavation was carried out at the invitation of the

Bureau of Archeology, University of the Witwatersrand. Mr.

A. J. H. Goodwin was approached, and he and Dr. K. H. Barnard

of the South African Museum undertook to excavate and analyse

the midden deposit. The cave was visited on Monday, November 18,

and excavation was carried on until Wednesday, November 20, 1935.

A section of midden, measuring about 6 feet by 4 feet in length and

breadth by 6 feet in depth, had been left standing by Mr. Patterson

(Plate LXII). The faces were examined and measurements taken,

and it was realised that the standing section was contiguous to

Drury’s original section, the N.E. corner of Drury’s trench coinciding

with the S.W. corner of the standing portion. It was therefore

thought advisable to start work on the eastern face of this section.

A section, 2 feet wide, was dug from front to back, following the steeply

falling layers of midden. In order to follow the stratification the

section was dug in such a manner that the depth of each layer was

9 inches at the forward end of the section, rising to nothing at the

back. The depths given therefore refer to the front of the section.

All material was passed over two sieves, the upper with $-inch

mesh, the lower with }-inch mesh.

0-8 inches. Two fragments of crude circular scrapers, one 2 inches

in diameter of sandstone, the other 14-inch in diameter of quartz.

One unworked quartz crystal, 2 unworked surface quartzite flakes,

one bead. Fish.

8-16 inches. Core-like fragment of surface quartzite, one double

crescent, 6 untrimmed stone flakes, 2 fragments of red ochre, 20

ostrich egg-shell beads, a few fragments of egg-shell, some chips of

crystal, a single shell crescent, upper and lower jaws of one or two

small buck. Fish.

16-24 inches. Five shell crescents, 3 fragments of ochre, 5 ostrich

ege-shell beads, a few fragments of egg-shell. Fish.

24-32 inches. Four shell crescents, 6 unworked flakes, a Wilton

core, crystals, ochres, 3 small shells (? Nassa), possibly bored. Three

ostrich egg-shell beads. Fish.

32-40 inches. One bone arrow-point. Three flakes. Crystal.

Ochre. Two beads. One bored fish-shell. Fish.

The whole section overlay wind-blown sand, which was completely

sterile. This had heaped up towards the rear of the cave, and over-

lay white sand, mixed with disintegrating fragments of Table Mountain

Klip Kop Cave, Hermanus. 217

Sandstone, showing that this was the original floor of the cave,

which had disintegrated very considerably. The forward portion

of the shell deposit lay directly upon this sandstone. Following

Drury’s observations, it was deemed unnecessary to dig farther into

this layer.

The absence of Middle Stone Age implements in this cave is peculiar,

as Dr. W. E. Hardy of Cape Town has found Middle Stone Age im-

plements a little behind the Klip Kop. These are of Table Mountain

Sandstone, wind-worn, and badly made. They suggest a rather crude

Mossel Bay type. The evidence is thus sufficient to show that the

region was inhabited during Middle Stone Age times.

SHELL CONTENT.

Dr. K. H. Barnard very kindly accompanied me on the excavation

and kept a careful note of the species and general quantities of shells

discovered in the various layers. There is no marked difference in

the types of shell throughout the deposit. What isolated forms occur

are noted. It is important to realise that no exact ratio between

the shells was possible to ascertain, as most of the shell material

was considerably broken. The shells have therefore been given a

rough ratio:—

Mollusca—

Gasteropods:

Patella cochlear moderately abundant.

5 patriacha ‘ abundant.

» barbara (plicata) . very abundant.

» oculus a few.

5 umbella one.

», longicosta moderately abundant.

» argenviller very abundant.

Turbo cidaris abundant.

5 sarmaticus very abundant.

Oxystele merula abundant.

Nassa kraussiana a few.

Bullhia laevissima a few.

» vrhodostoma 5 : aU ay hee

Burnupena cincta (olim. Cominella

porcata)

= limbosa abundant.

lagenaria

Purpura squamosa . two.

cingulata . two.

218 Annals of the South African Museum.

Mollusca (continued)—

Gasteropods:

Turritella knysnaensis . : . afew.

Haliotis midae 4 : ; . abundant.

is sanguinea ; ‘ . moderately abundant.

Lamellibranchs:

Mytilus edulis (meridionalis) . . very abundant.

Donaz serra. : : , . moderately abundant.

Ostrea sp.

The following are probably all adventitious, having been brought

into the cave attached to larger shells like Haliotis and Patella:—

Patella spp. juveniles, Helcion pectinatus (one), Mytilus crenatus

(one), Gibbula spp., apical whorls of Vermetus, and Crepidula

hepatica, the last-mentioned being fairly abundant. Coral

(one). Pulmonate (land snail), Trachycystis tollinw (six), probably

wandered in from the surrounding bush, not collected by humans.

(Layers 8-16 and 16-24 inches.)

Crustacea. Barnacles p F : . abundant.

Vertebrates. Fishes of genera Dentex, Pagrus,

Sparus, Diplodus, ete. (all be-

longing to the family Sparide) abundant.

Tortoise . 2 ; : . fairly abundant.

Shrew (Insectivore) . : . a few.

Vlei-mouse : : : . a few.

lahyb< : A : . afew.

Small buck. : . afew, upper and

lower jaws.

Inorganic. Pisolites. Concretions from the recent Limestone,

2 in layer 24—32 inches.

CONCLUSIONS.

1. Klip Kop Cave has been inhabited by a people with a single

mode of life. No change in the culture, climate or food habits

throughout the period of occupation was observed.

2. The culture is typified by the eating of shellfish, the catching

of fish, the use of bone awls and points similar to arrow-points used

by the living Bushmen; the making of stone crescents, shell crescents,

ostrich egg-shell beads, bored stones, slate “palettes”; the use of

ochres, white clays, etc., and is also marked by the complete absence

of pottery.

Klip Kop Cave, Hermanus. 219

3. The great bulk of the used stones found have been broken in

use rather than to any plan. They consist of chipped and split

sea pebbles.

4. The fish enumerated show no change in the temperature of the

sea-water during the period of occupation; nor do they present any

species extinct or unusual in these waters.

My thanks are due to Mr. James Drury for the use of his photo-

graphs and manuscript; to Mr. W. H. Patterson for allowing me to

make full use of the material excavated from this cave by him;

to Dr. K. H. Barnard for the valuable list of shells and bones com-

piled by him; to the Commission for the Preservation of Natural

and Historical Monuments, Relics and Antiques for a grant towards

the excavation of the cave; and to the Municipality of Hermanus

for permission to excavate.

EXPLANATION OF PLATES.

Puate LXI.

1. View outside the Cave, 1925 (Photo, J. Drury).

2. Mouth of the Cave when excavated by Mr. Drury (Photo, J. Drury, 1925).

PLiate LXII.

Interior of the Cave after excavation by Mr. Patterson, showing section of

original midden (Photos, A. J. H. Goodwin, 1935).

XI.

Plate L

Ann. S. Afr. Mus., Vol. X XIV.

j nS

Vj =

ij MW) Le

s/H mn

iS)

Yes é

W fii, ay

baw

—

oS}

yy Y)

Wy, y J y T

“oy y yy Vy

Wey 2 ye Yj

Yy Zi Yj

Cj VMs y is j

yj 1 Ys 4

Yj ) Yi

7 iY j

4 5 Yj

WM M////V) yy, Yi

- |

Wi g Gj

W .

La | N

3 s

Ss =

By my

Ann. S. Afr. Mus., Vol. XXIV. Plate LXII.

7. Way,

Goodwin. Neill &: Co., Lid.

( 221 )

10. South African Native Snuff-Bozes.

By Miss M. Suaw, B.A., Assistant in Ethnology.

(Plates LXIII-LXXV.)

THE following supplement to my previous paper (Ann. 8S. Afr. Mus.,

vol. xxiv, p. 141) has been added for the purpose of describing a

very outstanding collection of over a hundred South African native

snufi-boxes which belong to Mr. R. C. Camp, of Plumstead, Cape

Province. As post-rider, and later as campaigner, Mr. Camp has

had great opportunities, extending over a period of forty years, for

making his collection.

A few newly acquired specimens in the South African Museum

Collection have been added to the list.

TECHNOLOGICAL NOTES.

1. In all cases a polish is produced by frequent handling and

rubbing, by occasional deliberate and more frequent accidental

application of fat or grease, and sometimes by rubbing with fine clay.

2. Gourds.—To hollow out a ripe gourd a small hole is made at

the end where the stalk has been, as much of the flesh as can be

reached is taken out with a piece of wire, and the cavity is filled

with water to help rot the remainder of the flesh. The gourd is

hung up, and from time to time is shaken vigorously to loosen the

rotting flesh, which finally comes away quite cleanly from the rind,

and can be shaken out through the hole, leaving the hollow shell,

which is then dried. In small gourds, those generally used for snuff-

boxes, the majority of the flesh can be scraped out with the wire,

and the remainder is cleaned out by putting in stones and water

and shaking them about.*

3. Tough-shelled fruits such as Strychnos are treated and used in

a similar manner to the gourds.

4. Nut-shells have the flesh scraped out with an awl through a

small opening.

* Mr. F. Cornner of St. Cuthbert’s, Tsolo.

222 Annals of the South African Museum.

5. Moth cocoons are prepared by cutting off one end and removing

the grub, and fitting the end with a small stopper. They harden

with use.

6. Bone.—

a. A shin bone from which the marrow has been removed is

usually chosen for snufi-boxes. The original shape is

generally kept, but the surface may be scraped down

with a knife or spear-blade to form a rim.

b. A solid piece of bone is carved to the desired shape with

a knife or spear-blade, and hollowed out with an awl.*

c. Pieces of bone are carved into snufi-spoons.

7. Horn.—The original outer surface of the horn is seldom left

untouched, but usually scraped down with a knife and polished.

a. The natural hollow of the horn, after the removal of the soft

matter, is used as a container, with the solid tip frequently

carved to a conventional or naturalistic shape.

a, 1. In this case the bottom may have to be filled in, and wood,

calabash, or horn is used. A fairly thick disc is cut to

fit exactly, forced in, and held with brass, copper, or

iron tacks, or wooden or horn pegs, rubbed flush with the

sides, and varying in number from two upwards. Holes

of the necessary size are first made in the horn with an

awl for the tacks or pegs to be hammered in.

a, 2. In some cases both ends of the horn are cut off and have to

be filled in as described above.

b. A solid piece of horn is chosen for the snuff-box and carved to

the desired shape with a knife or spear-blade. The hollow

is gouged out with an awl.*

c. Pieces of horn are carved into snuff-spoons.

8. Bamboo.—A suitable length of bamboo is cut and the ends are

filled in with discs of gourd rind. At one end the disc is cut to fit

exactly and remains immovable, and at the other it forms a movable

lid.

9. Wood is carved with a knife or spear-blade. The boxes are

cut out of a single piece without joins. The hollowing is done with

an awl or gouge of some sort, frequently made red-hot to burn out

the hollow. Sometimes a central column is burnt from which to

work outwards.

a. In some cases a hole is made in a convenient place in the wall

of the box to assist in the hollowing. This is filled in later

* Casalis, The Basuto, p. 140.

South African Native Snuff-Bozes. 223

with a round of calabash rind or of wood, cut to fit exactly

and forced in.

b. To make small holes in wood, a red-hot awl can be passed through,

burning a passage as it goes.

10. Stoppers are made of—

a, 1. A round dise of gourd rind cut to fit the hole.

2. The same, with a thong knotted through a hole in the

centre to attach it to the neck of the box.

6, 1. A wad of rag wound into a roll and sewn with cotton.

2. The same, decorated on top with beads.

c. A piece of hide.

1. The stopper is carved out of a thick piece of hide.

2. A piece of leather is rolled and sewn down one side or

on top, with sinew.

d. Wood is carved to various shapes.

e. Bone is carved to various shapes.

jf. Horn is carved to various shapes.

11. As an alternative to fitting a wooden bottom in bone or horn

snuff-boxes, a piece of skin is thoroughly wetted, placed over the

hole and bound down with string to reach a certain distance up the

side. When dry the string is removed and the skin remains fixed

firmly to the box.

a. The same method is used to make a cap, instead of having a

stopper. By putting a little grease under the skin before

binding, and by removing the string and easing the cap while

it is setting, it can be kept movable.

6b. The same method is used for the strengthening bands of skin

which are often placed round the mouth or the neck of a

snuff-box to prevent splitting, or as a mend after splitting

has occurred, or simply as a decoration. In this case a

jointless ring is cut either from the tail skin or a narrow leg

skin, drawn over while wet, and moulded into position while

it sets and contracts.

12. Carving—A sharp knife or spear-blade is used to carve con-

ventional or naturalistic forms on wood and horn.

13. Branding is done with a red-hot iron.

1. The pattern is first outlined with a sharp point and then

filled in with branding.

2. The outline is branded and then filled in with branding.

3. The required surface is branded without any preliminary

outlines.

224 Annals of the South African Museum.

4. A sharp point is heated to make branded punching, cross-

hatching, and other types of fine line decoration.

14. Incising is done with a sharp point, usually iron.

a. On materials where it does not show up sufficiently it is

rubbed with some black substance, generally charcoal

and fat mixed to a paste, which sticks in the incisions

and shows up the pattern.

15. Wire is used as a decoration—

a. By insertion into gourd shells. Holes are made with an

iron point the same diameter as the wire to be used,

which is generally brass or copper. Short pieces of

wire are inserted at each end into the holes, close

together to form a motif, frequently a diamond or

triangle. The ends are left unfinished inside as the

aperture in the gourd is generally too small for the

maker to be able to fasten them in any way.

b. By sewing, on leather straps; the holes are bored with an

awl and the wire is passed through.

c. By twining, round objects of any material, particularly

round projecting portions, to form a band of varying

depth.

d. By winding simply, to form a band round objects.

e. Several strands of wire are plaited to form a ring round

the object.

16. Lead can be inlaid by carving the required pattern on the

object to be decorated and pouring molten lead into it. If it is to

be used as a repair a corresponding hollow is carved on each side of.

the crack and the lead poured in. It is generally seen on pipes.*

17. Beads are inserted into the skin of a gourd by making tiny

holes in it while it is yet soft, and inserting the beads, which are

fixed in quite firmly as the gourd dries.t+

18. Two parallel rows of holes are bored in a gourd shell. A length

of thread is passed through one of the end holes, from the inside,

and sufficient beads are threaded on it to cover the distance to the

corresponding hole in the next row. The thread is passed through

and down inside the gourd to come out again through the next hole

of the commencing row, when more beads are threaded on it, and

so on.

19. Beadwork over Gourds.—The work commences at the top or

* Fleming, Southern Africa, pp. 202-203.

+ Junod, Life of a South African Tribe, vol. ii, pp. 132-133.

South African Native Snuff-Bozes. 225

bottom of the gourd, and is taken down or up, increasing and decreas-

ing to fit the gourd exactly.

1. Diamond Network.—When finished the work gives the effect of

a series of interlocking diamonds. The points of the diamonds

are one or sometimes two beads, placed with their bores

vertical. They are connected by one or more beads which

have their bores diagonal left and right. Two sinew threads

pass through the points, dividing for the sides, and joining

again at opposite vertical points. Sometimes the diamonds

are distorted by two of the sides having one bead and two

having two.

2. Diamond-strung Rows.—The stringing is exactly the same as

above, but only the points of the diamonds are there, which

gives the effect of horizontal rows of beads, opposite in each

alternate row. All the beads have their bores vertical. In

this type the colour arrangement determines whether the

diagonal or horizontal aspect shows up.

3. Wound Rows—

a. On straight objects like bottles or bamboos the beads are

strung in rows with their bores horizontal. On first

sight it appears as if the beads were wound round the

box on a long string and then had a thread run down

at one point to hold them together. Actually, as each

coil is completed the string is passed through the first

bead of the coil for a second time, and then the beads

of the next coil are threaded on. To keep the line

vertical the string can sometimes be taken forward and

sometimes backward.

6. Similar to the above in every respect, but that each coil,

instead of being attached by the thread to the preced-

ing one, is sewn to a single vertical thread, which passes

down the box.

4. Hanging Rows.—The same type as above, modified for use on

gourds, which are curved, so that one support only would not

hold the beads up. Several strands of sinew, from three

upwards, are stretched tightly down the sides, attached at

the top to a single coil of beads and at the bottom to each

other, to a single bead, to a ring of beads, or to a wad of cloth.

The stringing commences horizontally, and each time a

vertical string is reached the horizontal string is passed once

round it before putting on the next bead. More often than

226 Annals of the South African Museum.

not the colour pattern is based on this arrangement, by chang-

ing the colour each time a vertical strand is met.

20. Bands of metal are fastened in the following manner. A hole

is bored through each end, and when the band is in position a short

piece of thick wire—brass, copper, or iron—is passed through the

holes, and the ends are bent once and hooked round each other.

21. “Blood and Clay” Technique-—When cleaning a fresh skin

(usually that of an ox) the blood and remaining tissues that are

scraped off are mixed with ochre or powdered clay, and a little more

blood added to form a stiff paste. A clay model of the desired shape

has been prepared and baked in the sun. The paste is spread over

this and given any further modelling that is desired (ears, tails,

eyes, etc.), and when nearly dry a rough nap is pricked up with an

awl. When quite dry a round hole is cut for the opening, and through

it the clay model is picked out.* This is a dying art.

22. Brass, copper, or iron hoops are inserted in the side of a box to

secure the strap. Two holes, exactly the same diameter as the wire,

are bored with an awl, at about # inch distance, and both ends of a short

piece of thick wire (about ;;-inch diameter) are inserted in the holes.

23. A single hole or two holes are bored in the side of a snuff-box

for a leather strap to pass through. The holes are usually near the

mouth, so that the thong can be passed through from both sides,

the ends knotted at the mouth, and then pulled up inside.

24, Straps.—Materials used—

1. Plain thin thong.

2. Plain flat piece of leather of varying width and thickness

(,; inch to } inch).

3. Two-ply-twisted cord, 7.e. two thin thongs twisted together.

4. Three thongs plaited together—generally for bands.

5. Beads strung on sinew or, less frequently, cotton:

a. single; 6. two single; c. double (two twisted to-

gether); d. treble (three plaited together).

. Cloth—a single strand or a two-ply twist.

7. Strands of wire, twisted or coiled to make a loop.

Attachment—

A. End passed through a hole and knotted.

B. End passed through a hole and a bead and knotted.

C. Both ends passed through the same hole and knotted

inside.

for)

* Fleming, Southern Africa, pp. 204-205.

t Mr. F. Cornner.

South African Native Snuff-Bozes. 227

D. Both ends passed through the same bead, tightly enough

to hold.

E. Ends knotted together.

F. Ends plaited together by threading one through holes in

the other.

G. Ends plaited together by threading each through holes

in the other.

H. In the case of two-ply cord, the single thong is passed

through a loop or hoop as far as its centre, before

making the twist.

I. In the case of bead cords they are generally fastened

to the other beadwork, either (1) direct; (2) through

a white porcelain or brass button; or (3) through a

large bead, generally of outstanding colour.

Ends sewn with sinew.

. Ends passed and held down under a band of metal.

Ends passed and held under a band of leather.

M. A band of leather passed through slits in the ends.

Sess!

(The figures T. 1, etc., beside each description refer to the paragraphs

in the Technological Notes.)

ZULU.

There are twelve snuff-boxes from the Natal-Pondoland border and

Zululand, of which five typical examples are illustrated on Plate

LXIII. They are all made of gourds, but with three forms of

ornamentation—branding, beadwork, and wirework.

Of the first there aretwo. One isa bottle-necked gourd, decorated

with four rings of incised and branded triangles, and having a wooden

stopper, and a leather strap to attach it to the waistband. The

other (fig. 6) is a plain round gourd, reddish brown and highly polished,

with a design like a spider web on one side. It has a round of gourd

rind as a stopper. (ANS Pay WO rans Leyes)

Height, 2,9, inches.

Greatest diameter, 32 inches.

Diameter of stopper, $ inch.

Of the second there are three, all bottle-necked gourds covered

with a network of beads and practically identical with fig. 1, Plate

XXIV, of the previous article. (LT. 2, 19), 24¢55 ana #)-)

All the others are decorated with conventional designs in brass

and copper wire. Tiny holes are bored in the rind in the required

VOL. XXIV, PART DO. 16

228 Annals of the South African Museum.

pattern, and short pieces of wire are inserted, each end in a hole.

They are not turned over or fastened off in any way inside. The

majority are plain round gourds, but two are bottle-necked.

(it Le 2, 10(a1); 15(q)-)

Fig. 5 is a dark brown gourd decorated with triangular motifs in

a spiral arrangement, alternately of brass and of copper wire. A

round of gourd rind serves as a stopper.

Height, 3% inches.

Diameter of lower bowl, 2, inches.

Diameter of upper bowl, 1? inch.

Diameter of stopper, 2 inch.

Fig. 2 is a dark brown leathery gourd, with four long diamonds,

half of brass and half copper wire, arranged alternately. It has a

gourd rind stopper.

Height, 3 inches.

Diameter of bowl, 213 inches.

Diameter of stopper, $ inch by {% inch.

Fig. 3 is a reddish brown gourd with a triangle design placed

between two horizontal bands. Again copper and brass are used

alternately. The stopper is missing.

Height, 2,9, inches.

Diameter of bowl, 23 inches.

Diameter of mouth, 3 inch.

Fig. 4 is a dark brown gourd with a lumpy surface to the skin.

The triangle design is of brass and copper wire alternately. The

stopper is missing.

Height, 3 inches.

Diameter, 44 inches.

Fig. 1 (S.A.M., 5480,,)) is a dark brown gourd decorated with brass

wire. Four sets of concentric rings are placed between bands top

and bottom. The stopper is missing.

Height, 3% inches.

Diameter, 33 inches.

TRANSKEI.

A very fine example from Pondoland (Plate LXIV, fig. 7) is a highly

polished light brown bottle gourd, decorated extensively with brass

wire, in the same manner as the Zulu snuff-boxes. On the nozzle

of the upper bowl the wire is twined. The gourd has broken off at

the neck and has been mended by boring a series of holes each side

of the crack, through which to pass twisted fibre, which is knotted

South African Natwe Snuff-Bozes. 229

at each hole and passed twice round the neck. It also holds in posi-

tion the leather strap, which is passed under it.at each side and held

by a yellow glass bead. The mouth is reinforced with a band of skin,

put on wet, and has a very small leather stopper. There is a small

hole in the base filled with a round of calabash. It has been suggested

that this was used to fill the snuff-box, as it is considerably larger than

the mouth and would be quicker to work with.

(T. T; 2, 10(¢2); 11), 15(q and o)s 24 (5 and B):)

Height, 64 inches.

Diameter of lower bowl, 23 inches.

Diameter of upper bowl, 1? inches.

Diameter of mouth, 5%, inch.

Diameter of stopper, 3°; inch.

Of the remaining eight examples from the Transkei, six are illus-

trated on Plate LXIV.

Fig. 1, from Butterworth, is made out of light-coloured, hard, fine-

grained wood, and is cut so that the grain shows very prettily on

one side. It is oval in horizontal section. Two holes are bored in

one side for the passage of a short hide loop, which still has the hairs

on. The wooden stopper is mushroom-shaped.

(Te 9, I)» 10,4), 23, 244 and A):)

Height, 23 inches.

Diameters of bowl, 143 inch by 1? inch.

Diameter of stopper, ;{; inch.

Depth of stopper visible, + inch.

Fig. 2, from Butterworth, is the top portion of a light-coloured

bottle-necked gourd, which has probably broken off at the neck and

been converted thus. The hole at the base has been filled in with a

round of gourd rind, and a similar piece acts as a stopper for the

mouth. Two holes are bored in one side and a short leather strap

is passed through. (T. 1, 2, Tax), 10(a1), 23; 244 ana ay-)

Height, 213 inches.

Diameter, 1,8, inch.

Fig. 3, from Tembuland, is made of horn, light brownish in colour,

with patches of darker brown and of grey. The base is filled in

with a round of wood held in place by tiny copper tacks. The stopper

is cut out of a thick piece of hide. A two-ply twisted leather cord

is attached to a double leather band round the neck.

(I, Ta)» Tat) 10(c1), 12, 24,3, and E):)

Height, 54 inches.

Diameter of base, 142 inch.

230 Annals of the South African Museum.

Diameter of mouth, 5%, inch.

Greatest diameter of, stopper, 3 inch.

Fig. 4 is an example of the “blood and clay” snuff-boxes described

on p. 148 of the previous article (Ann. 8.A.M., xxiv, 3). It is a

well-worn specimen, khaki-coloured, and made in the shape of a

ram. It has a wooden stopper in the opening in the head.

(T. 10(4), 21.)

Total length, 43 inches.

Greatest height, 22 inches.

Greatest width, 14} inch.

Total length of stopper, 7°; inch.

Diameters of stopper, 38, by 3%, inch.

Fig. 5, from Kingwilliamstown, is made of greyish horn, with two

white rings at the base and a white streak down one side. The

base is filled in with a very thick convex piece of brown wood, held

in by a few small horn pegs. It has a leather stopper, attached by

a thong to the plain leather strap which is fastened round the neck.

(T. Tq and @1)> 10(¢9); 12, 24, 2,and E):)

Total height, 52 inches.

Length over outer curve, 63 inches.

Depth of wooden base, $ inch.

Diameters of wooden base, 13 inch by 1{ inch.

Length of stopper visible, ;5, inch.

Diameter of stopper, $ inch.

Fig. 6 is an insect powder tin, decorated with twined zinc and

copper wire in alternate bands. It belonged to a Fingo.

(T. 15(¢)-)

Height, 2%, inches.

Diameter, 13 inch.

Plate LXV illustrates some recently acquired specimens in the

Museum collection, from the Pondomisi, of Tsolo District, Transkei.

Fig. 1 (S.A.M., 5557) is the lower part of a necklet. Four bead-

work flaps are suspended from a beaded cloth necklet, which is

fastened by a white bead loop over a conical brass button. The

flaps are white, with sky-blue edging, and a double vertical row of

continuous diamonds outlined in sky-blue down the centre. One

flap is considerably larger than the others, and has on its lower edge

three small flaps of alternate rows of white and sky-blue beads. To

each of these flaps is attached a round snuff-tin, bearing the head

of King George V. The tins are purchased at the trading stores, and

are used either for snuff or to hold the little paper receipts given to

South African Native Snuff-Boxes. 231

the natives by traders for grain brought to the store, and which

entitle them to the same value of articles from the store. (T. 19,).)

Length of flap (including small flaps), 9% inches.

Width of flap, 3} inches.

Diameter of tins, 12 inch.

Depth of tins, ? inch.

Fig. 2 (S.A.M., 5495) is a small round gourd entirely covered with

china-blue and white beads (} inch), arranged in alternate rectangles

and triangles, with a little knob at the base. The beads are strung

on sinew and hang loosely between vertical strands of sinew, to which

they are attached at each change of colour. Round the mouth is

a tiny vandyked frill of blue and white beads, with white-centred

red at the upper points. The top of the cloth stopper is beaded

in alternate rows of blue and white. There is a short handle of two

strands of beads, one blue and one white, twisted together and

attached on each side just below the frill.

(UE. 2, 10(2); 19,4), 24 (50 and 1:)

Height, 23 inches.

Diameter of bowl, 2,°, inches.

Diameter of stopper, 2 inch.

Depth of stopper visible, ;5, inch.

Fig. 3 (S.A.M., 5558) is a white glass bottle covered with black

cloth, over which alternate rows of sky-blue and white beads are

wound. Round the edge of the base is a frill of loops—white in the

centre, with white, dark blue, red, dark blue, and white beads top

and bottom. The bottle is attached to a short necklet of three strands

of sky-blue and white beads, in alternate pairs, plaited and fastened

at one end to the top row of the beading, and at the other by a white

porcelain button through a white bead loop. The stopper is missing.

(e 19 (sa)5 24 54 and I1and 2+)

Height, 34 inches, plus §-inch fringe.

Diameter of bowl, 1, inch.

Diameter of bottle neck, 2 inch.

Fig. 5 (S.A.M., 5509) shows a light-coloured horn snuff-spoon

(shghtly shortened to fit the plate). The handle is semi-circular in

horizontal section, the flat portion being in front. It is decorated on

the back with incised cross-hatching, filled in with black. The bowl

is set at a slightly obtuse angle, is widely oval and rather flat. The

tip of the handle is black. (Dae)

Length of handle, 153 inches.

Width of handle, } inch.

232 Annals of the South African Museum.

Length of bowl, 1 inch.

Width of bowl, ? inch.

Outer depth of bowl, 4 inch.

Fig. 6 (S.A.M., 5514) is a light-coloured horn spoon, the handle of

which is entirely covered with rows of china-blue and white beads,

arranged more or less in alternate pairs of rows. In the centre are

two groups of three rows—dark blue between two white-centred red.

The leaf-shaped bowl is curved at an obtuse angle. A hole is bored

through the projecting horn tip of the handle for a double stranded

loop of alternate china-blue and white beads.

(ite Te), 19 (sa)s 24 (55)-)

Length of handle, 10} inches.

Length of bowl, 1 inch.

Length of loop, 2? inches.

Basuto.

Of fourteen snuff-boxes from Basutoland two are bottle-necked

gourds, and the remainder carved of horn in various shapes. Typical

examples are shown at Plates LXVI and LXVII.

Plate LXVITI, fig. 1, comes from the Maluti Mountains. It is a light-

coloured highly polished gourd with a design of punched dots, made

with a red-hot iron point. It has a wooden stopper carved to hour-

glass shape, and a plain thong strap fastened round the neck.

(ie r 2, 13 (4), 245 and p):)

Height, 42 inches.

Diameter of lower bowl, 243 inches.

Diameter of upper bowl, 113 inch.

Diameter of stopper, $3 inch.

The other gourd snuff-box is decorated with a branded conventional

design of diamonds and triangles, and has a similar stopper.

Of the horn snufi-boxes the most outstanding are fig. 3, Plate LX VI,

and fig. 3, Plate LX VII. Each is carved from a single horn.

Fig. 3, Plate LXVI, is carved out of dark greeny-brown horn, and

highly polished. Its ornamental value far exceeds its utility. The

actual container is a conventionalised human head, with white bone

eyes. The head is hollowed from the opening at the top, and has not

a very great capacity. It is set on a long neck and shoulders, from

which extend two long curved prongs by which it was probably stuck

South African Native Snuff-Bozes. 233

through the hair or the belt. (They are slightly shortened in the

illustration to fit the plate.) (LT. 7), 10;,), 12.)

Length of head, 14% inch.

Width ear to ear, 1} inch.

Diameter of stopper, 2 inch.

Length of neck, 1,8 inch.

Length of prongs, 11? inches.

Width across prongs, ? inch.

Fig. 3a is a front view of the same snuft-box.

Fig. 3, Plate LX VII, is the same type, but is plainer, and the con-

tainer holds more snuff. It is made of light-coloured horn, with a

few dark brown streaks in it and not very highly polished. It has

an hour-glass-shaped wooden stopper, with a hole through the waist,

probably meant for a strap. (T. 7%), 10¢4g), 12.)

Total length, 214 inches.

Length of container, 4 inches (neck, 1,}; inch).

Diameter of bowl, 14 inch.

Diameter of mouth, 4} inch.

Length of stopper visible, 13 inch.

Greatest width across prongs, 1} inch.

(It is drawn to half the scale of the others, as it was too long for

the plate.)

Plate LXVI, fig. 1, shows a highly polished double snuff-box, carved

out of the solid part of a black horn to an hour-glass shape. The

hollow does not go right through the centre. One bowl contains

snuff and the other lead shot. The mushroom-shaped stoppers are

of the same black horn. The box is suspended from a necklet of

black porcelain beads, with large blue- or pink-centred white spots,

and two red glass beads with the same spots, threaded on a thong

which passes round the centre of the snuff-box and is held with a

plain black porcelain bead. The beads do not cover the whole

thing, which is plain at the back and has slits at the ends to fasten.

(Ee T()> 10,4), 244 and q):)

Length of left-hand bowl, 1% inch; of right-hand bowl, 142 inch.

Diameter of bowls, 1 inch.

Depth of stopper visible in both, + inch.

Diameter of stoppers, 7; inch.

Diameter of waist, } inch.

Beads: length, 7; inch; average diameter, 2 inch.

234 Annals of the South African Museum.

Plate LX VII, fig. 4, shows a snuff-box hollowed out of the solid part

of a light-coloured horn, with quite thin walls, so that it is shghtly

translucent. There is a small flat rim round the mouth. The

leather stopper has a ring of double-plaited leather at its base, and

a loop of copper wire coiled on a copper wire foundation passed

through the centre. (T. 7), 10(c2), 244 ana >)

Length, 34 inches.

Greatest diameter, 1,'; inch.

Diameter of mouth, 2 inch.

Height of stopper visible, 5’; inch.

Length of wire loop, # inch.

Platé LXVI, fig. 4,1s a ight brown horn snuff-box, with patches

of dark brown. It is carved to imitate a winged fruit, with a long

nozzle with flanging lip at the fore end, and a ring for a thong at

the base of the bowl (to hang upside down). The wooden stopper

has a large hole through the centre, and through it a copper ring.

(iu 7): 10,4); 12, 24(7).)

Height, 412 inches.

Greatest width of bowl, 12 inch.

Diameter of mouth, } inch.

These five last described snuff-boxes are carved out of the solid

part of the horn, and have their cavities artificially produced by

the maker. The two next to be described make use of the natural

cavity of the horn after the soft matter has been removed.

Plate LXVI, fig. 2, is a dark greeny-black polished horn snuff-box,

in the mouth fitted with a leather stopper. The base is oval, and

is filled in with a dise of wood, held by four brass tacks. One of the

horns has apparently cracked and has been mended with a piece of

skin wetted before binding on, so as to look as if jointless. A thong

is knotted round the neck and looped.

(ihe Ife U 10(c9), 11), 12, 249 and k):)

(a and al)?

Height, 64 inches.

Length over outer curve, 8} inches.

Diameters of base, 12 inch by 22 inches.

Diameters of mouth, 3% inch by 4 inch.

Total length of head and horns, 34 inches.

Width across horns, 2 inch.

Plate LX VII, fig. 2,is carved out of black horn to a very convention-

alised, single-horned, animal’s head. Itis highly polished. The base

South African Native Snuff-Bozes. 235

is filled in with a round of fine-grained, very highly polished wood,

held in by brass tacks. The back of the horn is carved to a raised

edge, and on this edge, near the base, is inserted a brass hoop for

the attachment of a two-ply twisted leather cord, the other end

of which is fastened to a plaited leather band round the neck. The

opening is in the mouth and has a leather stopper.

: (i 1, Ceenaicite 10(c2), 12, 24/5. 4, A and )*)

Height, 74 inches.

Length over outer curve, 1111 inches.

Diameters of base, 2,°, inches by 24 inches.

Depth from back to front of head, 14 inch.

Width of head, +3 inch.

Diameter of mouth, $ inch.

Length of horn, 4? inches.

Diameter of top ring of horn, 3 inch.

There are two more snufi-boxes, variations of this form, with

nozzles instead of carved heads.

BECHUANA.

Bechuanaland is very well represented in the collection, and as

all the specimens were collected by Mr. Camp personally, he can in

most cases give the exact locality from which they were obtained.

The Bechuana use horn in a similar way to the Basuto, but their

snuff-boxes are carved to nozzles rather than heads, and seldom

highly decorated. There are fourteen of these specimens, of which

seven are illustrated on Plate LX VIII.

Fig. 1, from the Batlapin at Daniel’s Kuil, Griqualand West, is

a light horn shading to black at the top, where the mouth has a

small flanging lip and a rag stopper. The base is filled in with a

flat disc of assegaibosch wood, held in by a single row of widely

spaced brass tacks. The lower part of the horn is decorated with

a rough conventional design, incised and rubbed with some black

substance. (lo Ueemacion esis 1eqaya)

Total height, 34+ inches.

Length over outer curve, 414 inches.

Diameters of base, 12 inch by 13 inch.

Diameters of mouth, ;% inch by $ inch.

Depth of stopper visible, $ inch.

Fig. 2 is a similar type, from the same tribe and locality. It is

made of black horn, and the base is filled in with a convex disc of

236 Annals of the South African Museum.

polished gourd rind, held in by small wooden pegs. The nozzle is

separated from the bowl by an inverted shoulder. The mouth has

a small flat rim, and the leather stopper is rolled and sewn down

one side. (DT aianaiary Oceans)

Height, 32 inches.

Diameters of base, 12 inch by 1? inch.

Diameter of Pe z¢5 inch.

Length of nozzle, 13 inch.

Depth of stopper vasible; 3 7s Inch.

Diameter of stopper, 35, inch.

216

Fig. 3, collected from the Batlaro in 1897, is made on the same

principle but rather more elaborately carved. The lower portion

is light coloured, with one long dark streak. The nozzle, which comes

nearer the centre of the horn, is quite black. Below the nozzle is

a grooved flanging rim, and immediately below that a raised lug

with a round horizontal bore, through which a plain thong is looped.

Another short thong which is attached to the leather stopper is held

by a knot through the same hole. This part is divided from the rest

of the bowl by a narrow sloped shoulder. The round wooden base

is held by three wooden pegs. (T. 1, Vary, 10(cg), 12, 24¢ ana B)-)

Total height, 4;°5 inches.

Diameter of aes 1% inch.

Height of lower pent of bowl, 2,7, inches.

Length of nozzle, ;°, inch.

Greatest diameter of nozzle, ;°; inch.

Length of stopper visible, $ inch.

Diameter of PUT 2 inch.

Length of lug, +3 inch.

Height of lug, 1 fon

Fig. 4, from the Barolong of Mafeking District, is made of black

and fawn horn cut off at both ends. The base is filled in with a

convex wooden disc, and the top with a conically sloped round of

wood which has a raised rim round the mouth. Both are held in

with wooden pegs. Two brass rings are inserted horizontally in

holes on the inner curve of the horn. To the lower one a two-ply

twisted leather cord is attached by passing the cord through singly

before commencing the twist. The double-twisted cord is then

passed through a brass bead, under the second brass ring and through

the bead again, by which it is held quite firmly, and the free end is

attached to a rolled leather stopper.

(i. Tq and a2)» 10 (ca), 22, 24 (3D and H)-)

South African Native Snuff-Bozes. 237

Length, 4,3; inches.

Diameter of base, 255; inches.

Diameter at top of horn, 14 inch.

Diameter of mouth, 4 inch.

Length of stopper visible, 3% inch.

Diameter of stopper, 3°; inch.

Fig. 5, collected from the Batlaro in 1897, is made of a polished

black twisted goat horn, oval in section, with an oval wooden base

held in with wooden pegs. The neck ends in an inverted shoulder,

below which there is a wide band of leather, sewn down the front

and projecting at the back in a folded tongue, all in one piece of

leather. The fold is sewn with three rows of chain-stitch in brass

wire. In the centre of the stitching is a small hole through which

is knotted the thin strap of leather which attaches the oval leather

stopper. (ils 1, UG and a1)? 10,2), 15%), 244,92, and yz)

Total height, 64 inches.

Diameters of base, 143 inch by 14'5 inch.

Diameters of neck, 3 inch by 33; inch.

Depth of leather band, 1 inch.

Length of leather tongue, 23 inches.

Width of leather tongue, | inch.

Diameters of stopper, 2 inch by 53, inch.

Fig. 6, from the Bakxatla, of Mochudi, is made of a section of a

light greyish white horn, oval in horizontal section, and cut off at

both ends, which are filled in with oval discs of wood, held by copper

tacks. A conical wooden stopper is inserted in a hole in the top

side, and in the same line are inserted two copper rings, through

which a plain leather strap is passed and fastened by threading each

end through slits in the strap itself.

(Ak 1, Ta and a2)» 10,4), 22, 245 and F):)

Length (centre), 22 inches.

Diameter (centre), 24 inches.

Height of stopper visible, ;%, inch.

Diameter of base of stopper, °, inch.

Diameter of mouth, ; inch.

Fig. 7 is a more elaborate snufi-box from the Batlaro of the Taungs

District. It is made out of a light-coloured horn which has a high

polish and an irridescent sheen. The base is filled in with a flat

238 Annals of the South African Museum.

disc of horn of the same colour, held with iron tacks. A deep flat

shoulder separates the bowl from the nozzle, which narrows towards

the flanging lip round the mouth. There is a highly polished conical

wooden stopper. The bowl is decorated with copper wire inlaid in

the sides. (E. 1, Teanaary> LO, Alaa)

Total height, 4 inches.

Diameter of base, 2°, inches.

Diameter of top, 1,’ inch.

Diameter of nozzle below lip, 3 inch.

Length of stopper visible, +} inch.

Greatest diameter of stopper, 7% inch.

Plate LXIX illustrates some of the Bechuana snuff-boxes made of

materials other than horn.

Figs. 1 and 2 show two snuff-boxes from the Batlaro made of shin

bones. Fig. 1 comes from Gamasep, Langberg. The bottom is

covered with a piece of skin wetted before being drawn over, and then

bound round with string so that when dry it fits quite tightly almost

half-way up the side. Above it there is a further band of hide

similarly drawn on wet. The uncovered part of the bone is decorated

with deeply incised lines filled in with black. The mouth is fitted

with an oval wooden stopper. (ZL. G(ay5 LO, 11S Wey Ae)

Height, 245 inches.

Diameters of base, 1,5, inch by 1,%; inch.

Diameters of mouth, § inch by +4 inch.

Depth of stopper visible, 3, inch.

Fig. 2 is a similar example from Pudduhusche, Langberg. It

differs from the above in that it has no second band of leather, the

mouth has a lip, and the incised decoration is in a triangular design.

The string marks where the skin has been bound are plainly visible.

Ge 6(a)» 10,4), 1k, ea)

Height, 44 inches.

Diameters of base, 1? inch by 1} inch.

Diameters of mouth, +3 inch by 2 inch.

Depth of stopper visible, 3% inch.

Diameters of stopper, # inch by 5% inch.

There is a third example, only differing from the above in size and

decorative design.

Fig. 4 is an entirely different type of snufi-box from Kolobeng. It

is carved out of light-coloured wood and forms the head of a stick

South African Natwe Snuff-Boxes. 239

or ornamental club, the shaft of which tapers towards the grip. The

greater part of the bowl, and a band on the shaft just below it, are

branded brown; the neck is left plain and light coloured. The bowl

is further decorated with four vertical columns of three deep tri-

angular grooves. The stopper is a round of gourd shell with notches

round the edge. (A Sh, NOjeaiy, LA Shey)

Height of box, 23 inches.

Greatest diameter, 2 inches.

Diameter of mouth, ? inch.

Height of stopper visible, ;°, inch.

Diameter of stopper, 2 inch.

Length of shaft, 272 inches.

Diameter at top, § inch.

Fig. 3 is another wooden snuff-box, from the Bakxatla, of Mochudi.

It is made of fairly soft ight wood, branded brown all over, and has

a flanging base. The mouth is fitted with a leather stopper, to the

top of which is sewn a short wide leather strap.

Gi 9, 10(c2), 13,3), 24, and J):

Height, 34 inches.

Greatest diameter, 14 inch.

Depth of stopper visible, + inch.

Length of strap, 1? inch.

Of seven gourd snuff-boxes from Bechuanaland three are shown.

Fig. 5 is a specimen collected in Griqualand West, but not of Bechuana

type. It is a light-coloured round gourd decorated in a diamond

design with brass and copper wire, in the same manner as the Zulu

snuff-boxes illustrated in Plate LXIII. As well as this, and probably

a later addition, there are three groups of five rows of porcelain

beads, alternate sky-blue (;); inch) and black (,4 inch), threaded

on sinew and fastened through holes in the gourd shell. There is

a band of branding round the mouth, with a ring of dots beyond it.

(T. 2; 134 and 4)? (14,); 18.)

Height, 24 inches.

Diameter, 2? inches.

Diameter of mouth, 57, inch.

Fig. 6 is a reddish bottle-necked gourd, from the Barolong of

Mafeking. Most of the lower bowl is covered with a jointless piece

of hide, drawn on wet and moulded to the bowl. The stopper is a

round of gourd shell. A small hole in the upper bowl has been

240 Annals of the South African Museum.

filled in exactly with a similar round of gourd shell. A leather strap

is fastened round the neck and knotted to a loop.

(ar, 1; 2; 10(a1), 11%), 245 mand F):)

Height, 4 inches.

Diameter of lower bowl, 243 inches.

Diameter of neck, 14 inch.

Diameter of upper bowl, 12 inch.

Height of stopper visible, § inch.

Diameter of stopper, $ inch.

Fig. 7 is a narrow-necked gourd, collected from the Batlapin of

Barkley West. It is light coloured, and has incised and branded

decoration in a triangle and diamond design. The mouth is strength-

ened by a jointless band of skin, and below it there is a leather strap

which passes round the neck and through the rather elaborately

carved wooden stopper. The stopper can slide up and down on the

strap, but is prevented by the knot from coming off.

(a 1, 2, 10,4), 11%), 13(), 244 and ¥):)

Height, 444 inches.

Greatest diameter, 2% inches.

Diameter of mouth, +4 inch.

Height of stopper visible, +3 inch.

Width across wings of stopper, 14 inch.

Two other bottle-necked gourds are quite plain, with leather straps.

A third is decorated with a single row of branded triangles, and has

a crack at the neck mended with giraffe tail hair. A fourth is covered

entirely with twined brass and zinc wire, in Zulu style, and looks like

a relic of the Matebele invasion, rather than a Bechuana product.

MASHONALAND.

From Mashonaland there are twenty-eight snufi-boxes of various

styles. Typical examples are illustrated on Plates LXX and LXXI.

Two specimens from Salisbury District, one of which is shown

on Plate LXX, fig. 1, are made of hard shiny nut-shell, reddish brown

in colour, and fitted with a calabash stopper. A leather loop, which

is made of a single rectangular piece of leather slit down the centre,

is inserted in a single hole in the side, and held by a knot.

(T. IE 4, 10(a1); 23, 240 anac):)

Length, 13 inch.

Width, § inch.

South African Native Snuff-Bozes. 241

Diameter of stopper, } inch.

Length of strap, 2 inches.

Two specimens are converted brass cartridge cases. Plate LXX,

fig. 2, is an “ Express” cartridge case with a faintly incised triangular

decoration round the mouth, which is fitted with a conical wooden

stopper. A knotted strap is inserted in a single hole in the side.

(T. 10(g), 14, 23, 244, gana zy:)

Length, 213 inches.

Diameter of base, ? inch.

Diameter of stopper, 3 inch.

Height of stopper visible, $ inch.

Plate LXX, fig. 3, is a “‘Martini Henry” cartridge case (made by

G. Kynock of Birmingham). It has a leather strap attached in the

same way as described above, and below the hole for the strap a

band of plaited copper wire. The conical stopper is made of reddish

wood. (T. 10(a), 15¢¢), 23, 244 anacy:)

Length, 243 inches.

Diameter of base, ? inch.

Height of stopper visible, 3 inch.

Diameter of stopper, $ inch.

A third type, of which there are three specimens, is illustrated on

Plate LXX, fig. 5. It is hollowed out of a solid piece of black horn,

lozenge shaped, with a flanging round base, which is decorated with

triangular notches all round the edge. Round the mouth there is

a band of leather, put on wet, and moulded to fit. The stopper is

made of leather, and from the top of it there is a short leather loop.

(i T()» 10,2); 11%), 244 and p>)

Length of box, 1% inch.

Greatest diameter, ? inch.

Diameter of stopper, 3‘; inch.

Length of loop, +3 inch.

A fourth type, of which there are seven specimens all from Salisbury

District, is made of a length of bamboo, fitted with a round of calabash

at each end, one movable, to act as a stopper, and the other fixed.

The bamboo is decorated with incised cross-hatched lines in con-

ventional patterns, and has a leather strap, by which it could be

suspended from the waistband or necklace. None of these would

242 Annals of the South African Museum.

be stuck through the ear lobe as in Zululand. Three examples are

illustrated on Plate LX XI.

Fig. 3 is quite a small specimen, of rather reddish colour, with

five bands of cross-hatched triangles, separated by double lies.

The cross-hatching in the triangles is further enhanced by filling in

with branding the top half of the diamonds formed by the crossing

lines. (T. 8, 10(a1), 13(4), 23, 24¢, cana x)-)

Length, 21% inches.

Diameter, ? inch.

Diameter of stoppers, 5%; inch.

Fig. 1 differs only in size and in slightly different technique of

decoration (though the effect is much the same), in which the incised

lines run parallel to all three sides of the triangle instead of only

two sides, and of the tiny triangles thus formed, alternate ones are

cut out, and enhanced by rubbing with soot. There are six bands,

divided by three lines. A two-ply twisted cord is looped through

a hole and knotted on the inside.

Length, 7} inches.

Diameter, % inch.

Diameter of movable stopper, 2 inch.

Diameter of fixed stopper, } inch.

Fig. 2 is a shorter stouter reed, reddish in colour. The leather

loop and the fixed stopper are attached in the same manner as

described above. The movable stopper is attached by a short

length of two-ply twisted leather cord, one end of which is knotted

through a hole in the centre of the stopper, and the other end through

a hole in the wall of the reed near the end. Round each end is a

band of single continuous chevron, or zig-zag, carved in the rind.

The remainder is divided into three wide bands, each containing a

large plain single continuous chevron obtained as a result of carving,

in the manner described above, series of triangles with their apices

interlocking and fairly widely spaced.

Length, 4 inches.

Diameter, 13 inch.

Diameter of stoppers, 1} inch.

All of the remainder are carved out of various hard woods. Plate

LXX, fig. 4, from Salisbury District, is of hard polished black wood,

apparently ebony; at its widest circumference it has a reddish tinge.

It is decorated with sparse regularly spaced triangles, enhanced by

rows of triangular grooves and ridges, which run parallel to the base

South African Natwe Snuff-Bozes. 243

of the triangle. The neck, with the exception of two triangular

ridges at its base, is covered with twined brass wire. On one side

a rounded lug is bored horizontally for the passage of a two-ply

twisted leather cord. There is a small calabash rind stopper in the

mouth, and another larger one in a hole in the base, fastened in

flush with the surface. This may have been used to fill the box

from the bottom, as this hole is larger than the mouth, but it was

more probably originally intended to facilitate the hollowing out

of the bowl. (aS Dy ey Mais Ws Aer vee)

Total height, 34 inches.

Depth of neck, 1,1; inch.

Diameter of bowl, 3 inches.

Diameter of top of neck, 13 inch.

Diameter of stopper in base, +} inch.

Length of lug, § inch.

Height of lug, $ inch.

Width of lug, 23 inch.

Figs. 8 and 8a are two views of a similarly shaped snufi-box from

the same district. It is made of dull light-brown wood, and its

only decoration is a semi-circular groove round the circumference

and a circle of diamonds carved round the base of the nozzle. The

mouth is fitted with a small calabash stopper, and a larger hole in

the base has a similar stopper, as in fig. 4. The snuff-box is sus-

pended by a raised lug on the circumference from a necklet of light

wooden beads, which are carved to an hour-glass shape and branded

dark brown. They are strung on two-ply twisted fibre and fit into

each other. (Ae NS Ds iri WOpeays 1s CH)

Height of bowl, Ms inch.

Length of nozzle, % inch.

Diameter of bowl, 2 inches.

Diameter of nozzle, +3 ee to 3°; inch.

Diameter of stopper, inch.

Diameter of stopper in T baee: = inch.

Height of lug, 55 ts inch.

Width of lug, 5, inch.

Length of a oLIse 163 inches (thirty-six beads, eighteen on each

side).

Length of beads, 2 meh to = inch.

Diameter of beads, 5%, inch.

> 16

VOL. XXIV, PART 5. 17

244 Annals of the South African Museum.

Fig. 7, from the same district, is made of dark reddish wood. The

nozzle is covered with skin, drawn over while wet, turned over about

% inch inside the mouth, and bound round with string until dry.

The string marks are clearly visible. The mouth has a brown wooden

stopper, so regular in shape that it might have been turned on a

wheel. In addition to a lug carved at the base in the same piece

of wood, by which the box could be suspended upside down, there

is a second lug carved out of thick hide and inserted in a hole in

the side of the snuff-box, near the nozzle. The bowl is decorated

with vertical grooves on each section, divided by a plain horizontal

band and covering the whole surface, with the exception of a plain

vertical band. SS Ove Ibe. 112.)

Total height, 53 inches.

Greatest diameter, 14 inch.

Height of basal lug, } inch.

Diameter of lug near nozzle, ;% inch.

Length of nozzle, +2 inch.

Diameter of nozzle, 3 inch.

Length of stopper visible, 7’; inch.

Diameter of stopper, } inch.

“

The following three examples are the typical pear-shaped snuff-

boxes of the Mashona. They are all from Salisbury District.

Plate LXX, fig. 6, is of dark greyish brown wood with longitudinal

triangular ridges and grooves over the whole of the bowl. The

flanging nozzle is covered with twined fairly thick brass wire. The

mouth has a calabash stopper attached by a two-ply twisted leather

cord to a raised lug on the side of the bowl. The same sort of cord

is used for a long loop, which would go round the waist or over one

shoulder, and is also passed through the lug and held in position

by three cone-shaped ivory beads of graduated sizes.

(ik 9, 10 (a9); 12, 15(.); 24 (5 4, C, Dana E):)

Height, 6,8 inches.

Greatest diameter, +3 inches.

Length of nozzle, § inch.

Diameter at top, + nn inch.

Diameter of ual and stopper, +3 inch.

Length of lug, ? inch.

Height of lug, 4 inch.

Length of strap, 134 inches.

South African Natiwe Snuff-Bozes. 245

Beads: (1) 4% inch by 33, inch; (2) 2 inch by 4 inch; (3) 7 inch

by 33, inch.

Plate LXXI, fig. 4,1s made of hard dark brown wood decorated in

a diamond design in flat ridges and triangular grooves. On one

side there is a large patch of lead, inlaid into the wood, possibly to

cover a crack. The plain rounded lug is bored horizontally and has

a long leather strap. The flanging nozzle is covered with twined

brass wire, and the mouth has a rag stopper.

(T. 9, 10(1), 12, 15(¢), 16, 246 ana ay:)

Height, 4% inch.

Greatest diameter, 12 inch.

Height of nozzle, ? inch.

Diameter at top, +2 inch.

Height of stopper visible, 5%, inch.

Diameter of mouth, 4 inch.

Length of lug, # inch.

Height of lug, 53, inch.

Plate LX XI, fig. 5, is outstanding for its size. It is made of wood,

light both in weight and colour, carved in longitudinal triangular

grooves and ridges, with small groups of horizontal and diagonal

grooves and ridges at irregular intervals. The large angular lug is

plain. The whole box is branded black. The round stopper is of

brown wood with a hole through the centre and a row of circular

depressions along the top. A plain leather strap is passed through

the lug. (T. 9, 10(a), 12, 13,3), 24, s ana)

Total height, 84 inches.

Greatest diameter, 24 inches.

Length of nozzle, 3 inch.

Diameter of nozzle at top, § inch.

Diameter of mouth, 4 inch.

Depth of stopper visible, } inch.

Diameter at top, 3 inch.

Length of lug, 13 inch.

Height of lug, 2 inch.

Plate LXXI, fig. 6, also from the Salisbury District, is a very slight

variant of the preceding type. It is made of fairly heavy, highly

polished dark reddish wood, with a carved stopper of the same. The

base of the bowl is plain, and cut off bluntly instead of diminishing

to a pot. The upper part is carved into flat ridges. The nozzle

is covered with twined copper and zine wire. A two-ply twisted

246 Annals of the South African Museum.

leather cord is passed and knotted through a hole in the stopper

and attached to the plain raised lug.

(ite 9, 10(a), 12, 15 (¢); 243s and H):)

Height, 6 inches.

Greatest diameter of bowl, 1} inch.

Length of nozzle, +3 inch.

Diameter of nozzle at base, 34 inch; at top, ? inch.

Length of lug, 2 inch.

Width of lug, 2 inch.

Length of stopper, | inch.

Greatest diameter of stopper, $ inch.

OVAMBO.

There are in the collection five Ovambo snuff-boxes, all of which

are illustrated on Plates LX XII and LXXIII, together with three

recently identified specimens in the Museum collection.

There appear to be two main types of snufi-box made by the

Ovambo; one of horn or dark wood, decorated in very much the

same way as the last described Mashona type, and the other of

light wood decorated with branded designs.

On Plate LX XII.

Fig. 1 is made of black polished horn, plain but for a slight bulb

at the base. There is a plain copper band round the top, fastened

over the leather strap which attaches the stopper, thereby holding

it in place. The band is fastened by a piece of copper wire passed

through a hole in each end and hooked to itself. (This is the typical

Ovambo method of fastening bands of metal.) The other end of the

strap passes through a hole in the centre of the gourd rind stopper.

(i. 1, T(a)> 10 (a2), 20, 244 and K):)

Length, 5 inches.

Diameter at mouth, ? inch.

Diameter of stopper, 5%, inch.

Depth of stopper, 53, inch.

Depth of band, ;*, inch.

Length of strap, 64 inches.

Fig. 2 is outstanding in that two snuff-boxes are suspended from

one strap, which is folded in half and divided at each end, the front

ends being knotted through the gourd stoppers and the back ends

South African Native Snuff-Bozes. 247

fastened under the band of tin round each mouth—the one on the

left is fastened with brass wire, and the other with copper wire.

The boxes themselves are made of dull black horn, decorated in a

triangle pattern of flat ridges and triangular grooves. The peduncles

are plain. (T. Ta), 10 (a9), 12, 20, 246 ana x)-)

Left. Right.

Length . 51, Inches. 5,3, inches.

Greatest diameter . 13 imeh. +3 inch.

Length of peduncle 13 inch. 18) Saket

Diameter of base of peduncle . 7, inch. ze Inch.

Depth of bands. A . 4°, Inch.

Length of strap (folded) . : . 42 inches.

Width of strap at fold. : euch,

Fig. 3 1s a single box of the same type, made of light-coloured

wood branded black. The neck flanges slightly and has a band of

leather sewn down one side with sinew, and from it, in the same

piece of leather, runs the strap. The strap narrows towards its

end, and after passing through two holes in the gourd rind stopper,

is knotted through a hole in the handle of a light-coloured wooden

spoon, which is decorated with incised branded lines.

(le 9, 10(a2); 12, 13,5 and 4) 242, A and A)

Length, 8? inches.

Greatest diameter, 1,°, inch.

Diameter of mouth, 1,), inch.

Length of peduncle, 1,), inch.

Diameter of peduncle at base, ;°,; inch.

Depth of stopper, $ inch.

Diameter of stopper, +2 inch.

Depth of band, 5% inch.

Total length of strap, 16 inches.

Total length of spoon, 6,3, inches.

Fig. 4 is carved out of dull black horn, and decorated with inter-

locking triangles enhanced by flat ridges. The peduncle at the base

is quite plain. Round the mouth is a brass band, fastened, as

described above, over the leather strap which attaches the gourd

rind stopper. (LT. Tq), 10(a2), 12, 20, 24 ana x);)

Length, 62 inches.

Greatest diameter, ~ inch by 12 inch.

248 Annals of the South African Museum.

Length of peduncle, 1,4 inch.

Diameter of peduncle at base, 5%; inch.

Depth of stopper, 5°; inch.

Diameter of stopper, $ inch.

Depth of band, 5%, inch.

Length of strap, 64 inches.

Three further examples of the same type, from the Museum

collection, are shown on Plate LX XIII.

Fig. 1 (S.A.M., 3337) is carved out of hard black wood decorated

in a triangle design enhanced with flat topped ridges. The peduncle,

as well as the neck, is covered with a band of tin, fastened as de-

scribed above. The band round the neck holds in place a thick

leather strap, the end of which passes through the centre of the

gourd rind stopper. An ivory spoon, which is decorated on the back

of the handle with incised vertical cross-hatching rubbed with black,

is passed through a hole in the strap, near the box.

(i 9, 10(a9); 12, 144), 20, 240 and K):)

Length, 6,°, inches.

Greatest diameter, % inch.

Diameter of mouth, 2 inch.

Length of peduncle, 1} inch.

Diameter of peduncle, 3 inch.

Depth of bands, ? inch.

Depth of stopper, 53; inch.

Length of spoon, 4,°, inch.

Length of strap, 144 inches.

Fig. 2 (S.A.M., 3338) is made of black horn carved in longitudinal

triangular grooves and ridges. The pedunele is plain, with a rounded

base. The band of leather round the neck is in the same piece as

the strap which attaches the gourd rind stopper, and is cut from a

tail piece and drawn over while wet. (T. Tay. 10(q2), 12, 24,2.)

Length, 64} inches.

Greatest diameter, 14 inch.

Length of peduncle, 14 inch.

Greatest diameter of peduncle,

Depth of stopper, } inch.

Diameter of stopper, +} inch.

Depth of band, 2 inch.

Length of strap, 6? inch.

inch.

bo}

South African Natwe Snuff-Bozes. 249

Fig. 3 (S.A.M., 3339) is made of plain light-coloured horn. The

peduncle has a band of tin, fastened as described above, with a single

ring of flat, native made, copper wire at the top of the peduncle.

A thin thong is bound round the mouth and tied firmly at the back,

by passing through slits in itself, and of the two ends, one is attached

to the gourd rind stopper and the other left loose. The part of the

thong which goes round the mouth is bound with twisted sinew string.

(ie Ta), 10 (a9); 20, 244 and E):)

Length, 7 inches.

Greatest diameter, ~ inch.

Diameter at mouth, ? inch.

Length of peduncle, ? inch.

Diameter of peduncle, ? inch.

Depth of stopper, $ inch.

Diameter of stopper, 2 inch.

Length of strap, 224 inches.

Fig. 4 illustrates the only example of the second type. It is made

of light brown, hard, fine-grained unpolished wood, and shaped like

a barrel, with slightly concave base. The motif of the incised and

branded decoration is a series of single chevrons, in vertical columns

divided by a plain cross-hatched band. There is a rectangular lug

at the mouth, bored vertically for a strap, which is held at the base

by a knot and a pale green glass bead, and divided at about two-thirds

its length into two two-ply twisted cords, one of which is knotted

through the centre of the gourd rind stopper and the other through

the top of a wooden spoon. The spoon is of the same wood and

slightly decorated at the top with incised branding.

(i 9, 10(a2), 134 and 2)9 240 a and B):)

Height, 34 inches.

Greatest diameter, 14} inch.

Diameter of base, 14 inch.

Diameter of mouth, 1} inch plus 53, inch (lug).

Length of lug, 2 inch.

Length of strap, 5 inches.

Length of spoon, 44 inches.

Bowl, 14 inch by 2 inch.

BATHONGA.

Finally there are two snufi-boxes from the Bathonga of Delagoa

Bay, both illustrated on Plate LX XIV.

250 Annals of the South African Museum.

Fig. 1 is a light brown bottle-necked gourd, highly polished, and

decorated with rows of white porcelain beads (,!, inch), inserted in

the rind, and enclosing alternate bands of cross-hatching and plain

black, all branded. A two-ply twisted leather cord wound several

times round the neck, holds the plain leather loop strap—the lowest

coil passing through a sht in the strap. The stopper is a round of

gourd rind. (ie le, MO eri, Ske anda 0) 24 Gls naenye)

Height, 64 inches.

Diameter of lower bowl, 3 inches.

Diameter of upper bowl, 1,75 inch.

Diameter of stopper, $ inch.

Depth of stopper visible, =; inch.

Width of strap, +3 inch.

Depth of neck binding, 42 inch.

Fig. 2 is a dark brown highly polished leathery looking gourd,

decorated with branded triangles and alternating bands of plain

branding and branded cross-hatching, all outlined with white porcelain

beads (,!; inch). A band of leather round the neck passes through

slits in a plain leather strap with fringed ends. The dark wooden

stopper is carved to an hour-glass shape.

(Ue 1, 2, 10(q); 13¢2 ana 4) Ns 24 2 and ty:)

Height, 7% inches.

Diameter of lower bowl, 4% inches.

Diameter of upper bowl, 23 inches.

Greatest diameter of stopper, 1 inch by ,%, inch.

BAVENDA.

The remaining three figures on Plate LX XIV are three snuff-boxes

in the collection of the University of Cape Town, which I have been

permitted to reproduce here, as they illustrate three stages in the

making of the lead-covered snuff-boxes which are peculiar to the

Bavenda.*

Small gourds or the hard-shelled fruits of the mutasa tree are chosen

for the snufi-boxes, and hollowed out. Hight divisions are made

on the outside of the shell, and alternate ones are scraped down to

the depth of about ;4, inch. Lead, bought at the trader’s store, is

inlaid in these hollows, sufficiently thickly to stand up about 54, inch

above the shell. Then with a stone as a tool, it is worked over the

* Stayt, The Bavenda, pp. 50, 53, and 159.

South African Native Snuff-Bozes. 251

uncovered portions of the bowl until the latter is entirely encased

in lead, and only very faint lines show the junctions.

Fig. 3 (U.C.T., 29:101) is a mutasa fruit snufft-box in the first stage,

with four hollows scraped out. It has a wooden stopper.

(D3; 10(5), 125 16")

Height, 143 inch.

Greatest diameter, 113 inch.

Length of stopper, $ inch (,°, inch visible).

Diameter of stopper, 2 inch.

Fig. 5 (U.C.T., 36:4) is a bottle-necked gourd snufi-box at the second

stage, when the lead has been inlaid into the four hollows. This

box has been fitted with a strap of native spun two-ply cotton twine.

A skein is passed round the neck, held together by a sewn ring of the

same material, then divided into four groups and plaited, thereby

making a long loop of two plaits each side, joined at the ends by a

similar sewn ring. The wooden stopper is missing.

(ag A MO yea, 45 AG).))

Height, 412 inches.

Diameter of lower bowl, 1? inch.

Diameter of mouth, 4 inch.

Length of loop, 16 inches (approx.).

Fig. 4 (U.C.T., 29:101) shows the final stage, when the fruit shell

is entirely covered with lead. It has a wooden stopper.

(ere3 wlOV sy alae eOe)

Height, 24 inches.

Diameters, 1% inch by 133 inch.

Length of stopper, $ inch (3% inch visible).

Diameter of stopper, 3°, inch.

The seven snufi-spoons in Mr. Camp’s collection are illustrated

on Plate LXXV. They are all from the Transkei, with the exception

of fig. 6, which comes from Bechuanaland. They are made of bone,

with incised decoration rubbed with some black substance. In figs.

2 and 3 the decoration is the same back and front, and in fig. 4 the

same on both sides. (Abs Oras WANS)

[ TABLE.

252 Annals of the South African Museum.

Length Width Length Width | Length Width

of across of of of of

Prongs. Prongs. Neck. Neck. Bowl. Band.

Inches. Inch. Inch. Inch. Inches. Inch.

Fig. 1 43 16 To To Ls $

Fig. 2 37's 3 it 3 lige 3

Fig. 3 4 Sn t t g 35

Fig. 4 54 4 2 plus 3 4 1} 3

Fig. 5 Oaey + 13 3

Fig. 6 | Total len|gth, 53 inches 34 3

Hig: 7 Ate 3 lis 32

I should like to thank Mr. Camp for allowing me to publish his

collection, and for the assistance he has given me during the work.

My thanks are also due to Mr. A. J. H. Goodwin for permission to

reproduce the Bavenda snuff-boxes; to Professor Lestrade for infor-

mation about the latter; and to Mr. Frank Cornner for technological

information.

Ann. S. Afr. Mus., Vol. XXIV. Plate LXIII.

M. Shaw. Neill & Co., Ltd.

Ann. S. Afr. Mus., Vol. XXIV. Plate LXIV.

SEES

Neill & Co., Lid.

M. Shaw.

Ann. §. Afr. Mus., Vol. XXIV. Plate LXV.

Za oh OF beyn

are == (Ao,

1s,

raat

pare

M. Shar. Neill & Co., Lid.

Plate LXVI.

Ann. S. Afr. Mus., Vol. XXIV.

julia

Neill & Co., Lid.

Shaw.

Ann. 8S. Afr. Mus., Vol. XXIV. Plate LX VII.

M. Shaw. Neill & Co., Lid.

Ann. 8. Afr. Mus., Vol. XXIV. Plate LX VIII.

—

M. Shaw. Neill & Co., Ltd.

Plate LXIX.

. XXIV

Ann. S. Afr. Mus., Vol

Neill & Co., Ltd.

M. Shaw.

late LXX.

Vol. XXIV.

’

Ann. 8. Afr. Mus.

Sra

Neill & Co., Ltd.

M. Shaw.

Vol. XXIV. Plate LX XI.

Ann 8. Afr. Mus.,

> Co., Ltd.

Neill &

M. Shaw.

Ann, 8S. Afr. Mus., Vol. XXIV. Plate LX XII.

Derr:

M. Shaw. Neill & Co.. Ltd.

Plate LX XIII,

Mus., Vol. XXIV.

Ann. 8. Afr.

Neill & Co., Ltd.

M. Shaw.

Ann. S. Afr. Mus., Vol. XXIV. Plate LX XIV.

LEAKY

M, Shaw. Neill & Co., Lid.

Ann. S. Afr. Mus., Vol. XXIV. Plate LXXYV.

M. Shaw. Neill & Co., Lid.

—

11. Ovambo Knives.

By Miss M. Suaw, B.A., Assistant in Ethnology.

(With Plates LXXVI-LXXXIV.)

THE knives made and used by the people of Ovamboland and their

neighbours to the north and east are in many ways unique, and are

restricted to a comparatively small area. They are so outstanding

in shape that it is somewhat strange that they should have been so

little remarked on by travellers in that region. Galton, one of the

earliest travellers in Ovamboland (1851), says of the Ondonga tribe:

“Their dagger-knives were creditably made and very pretty. The

knife was set into a wooden handle . . . in a wooden sheath, but

both handle and sheath were in part covered with copper plating and

in part wound round with copper wire beaten square.” (8). He does

not remark on the shape, but the description might apply to any

of the knives on Plate LXXVI. A better description comes from

Schinz (11), who travelled in South-West Africa in 1884-87. Speak-

ing of the dress and outfit of a man, he includes “the long dagger-knife,

which is stuck in a sheath that has a window-like opening on the

upper side... . The grip of the knife, which is used sometimes as

a weapon, and sometimes as an implement or plaything, is made of

wood and is often wound round with copper wire for ornament and

for greater strength. Very often the wooden end is cut sloping out,

and then has the strange form that is said to represent ox-horns.”

Passarge, in an article on the Mambukushu (10, p. 297), has two

illustrations, reproduced on Plate LXX XIII, figs. 1 and 2, showing

knives very similar to 8.A.M., 1735 (,) (Plate LXXXIII, fig. 3). He

writes: “‘The knife, without which one seldom sees a Mambukushu,

consists of a lance-shaped blade with a wooden hilt. The most out-

standing part is the wooden sheath, the bottom of which is T-shaped.

This type of knife has a wide distribution in the north Kalahari and

Ovamboland.” Toénjes (18) describes among the weapons used in hand

to hand fighting, ““the two-edged dagger, which varies in length from

20 cm. to 75 cm., stuck in a wooden sheath. Particularly artistic

examples are found among the Ondonga, neighbours of the Ouku-

anyama. Outstanding is one with the big triangular sheath, covered

with copper wire or copper plating.’ Finally, Hambly (7, p. 175)

254 Annals of the South African Museum.

describes the knives of the Vakuanyama of South Angola as being

unique among the knives of Angola, and as “‘varying in length from

48 cm. to 73 cm., with a breadth of 5 cm. to 7 cm. across the scabbard.

The general outline is a well-balanced ellipse. The wood is hard in

texture, and dark red in colour. On one side the scabbard is left

open in such a way as to display the blade, which is long, keen, and

tapering. A leather thong is attached to the back . . . there is a

crescentic-shaped expansion at the tip of the scabbard.”

There are two sorts of knives, the ordinary ones with wooden

handle and wooden sheath, and the special knives with handle and

sheath covered with copper wire or copper plating. The copper-

covered knives were a sign of great importance and the highest

order among the Ovambo. They were the property of the chief,

who gave them as a special mark of distinction to his body-guard and

military leaders (18, p. 59), and were worn hanging on a belt on the

left hip. It has been stated (8) that this type of knife is not made

for ordinary use, but that during a war when the chief leads his

warriors he wears one knife in front and one at the back. If he is

in need of anything he gives the knife to a messenger, who, when

he shows it, is given whatever he asks for, “‘a sort of king’s signet,

having authority to commandeer anything in the name of the chief.”

The ordinary wooden knives are worn on a thin leather strap, over

the shoulder or on the left hip, towards the back, and are used for

all everyday purposes—for hunting, as a tool, or as a weapon in

fighting. One special function is mentioned by Hahn (6, p. 10),

who says “the most favoured mode of ‘smelling out’ the possessors

of evil spirits, wrong-doers and other culprits, is with the aid of an

ordinary Ovambo knife. . . . It is placed in the fire until it is red-

hot, and then, with the edge downwards, drawn across the palm of

his (the witch-doctor’s) hand. While doing this the witch-doctor

utters the names of several suspects, and as long as the knife runs

smoothly after a name has been mentioned and does not stick or

blister his hand the person concerned is innocent. When, however,

the progress of the knife across the palm is interrupted, causing his

hand to be burnt, he jumps up, and pointing at the last-mentioned

man declares him guilty. His finding is accepted as final.” Another

special function is mentioned by Ténjes (14) in speaking of a witch-

doctor, ‘““who has power to cause the death by magic of someone

living at a distance. If, for example, certain people want to kill

someone whom they cannot get hold of, they go to an omutilali so

that the latter can assist them in the attainment of their object.

Ovambo Knives. 255

The witch-doctor asks the name of the man who is to be removed.

Then he takes a pot of water and places it before him, holding his

knife in his right hand. He looks steadfastly into the water and

suddenly stabs the knife into it with a lightning stroke. Then he

takes the vessel, covers it carefully, and puts it on one side. At this

point the people . . . are taken into another place and given some

food. The omutikili takes a little blood from some animai and puts

it into the water in the vessel. Then he recalls the people and, care-

fully uncovering the pot, shows them the blood-stained water—a

certain sign for them that the omutikili has got their man. They

believe, in fact, that he can see in the water a picture of the man

they want killed, and can murder him directly.”

The main features of the Ovambo knives (omukonda) which

distinguish them from other South African knives are the winged

foot, the triangular opening down the front, the hole in the base of

the foot, the knob-shaped top of the handle, and the shape of the

opening at the top of the sheath for the blade, and of the raised

projection on the back of the sheath for the attachment of a leather

strap. The general outline is elliptical, the widest part being either

at the mouth of the sheath or, more frequently, just below it. The

horizontal section of the main part of the sheath and handle is a

rather thin oval, while the horizontal section of the foot is an ellipse,

widest at the centre but generally more curved on one edge than the

other—most frequently the back. The two-edged blade is tanged,

and the tang runs the whole length of the handle, and is visible,

slightly, or bent over and beaten down, at the top. This is a

common feature in South African knives. Unlike so many South

African blades, the blade is not ogee in section—nor are the

Ovambo iron arrow blades or assegai blades. The opening at the

top of the sheath varies in shape from a plain ellipse to an ellipse

with a circular hole in the centre, corresponding to the hole in the

foot. The projection for the thong at the back of the sheath is

generally cylindrical, sometimes quite round, and sometimes inclined

to be square in section. S.A.M., 5058 (,, (Plate LX XIX, fig. 1) and

5368 (Plate LX XX, fig. 1) are exceptions to this, both being saddle-

shaped. In nearly every case the projection is unbranded, and longer

than the part above the opening in front; in a few it is the same

length, and even less frequently, shorter. It is bored vertically,

horizontally, or in some specimens both ways.

The decoration is done mostly with a branding iron, and is based

on the general outline of the scabbard, the edges of which are branded

256 Annals of the South African Museum.

quite black, leaving an unbranded portion in the centre, raised and

well-defined in front, but in the majority of specimens less regular

on the back. Some have a further slight decoration on the front,

carved and then branded, always some arrangement of triangles

which are the basis of nearly all the branded decoration on wooden

articles manufactured by the Ovambo; the only other form used,

as far as is exemplified in the Museum collection, is a ridged

design.

The ornamentation is in the shape of the foot itself, and in bands

of metal or wire. Many South African knife sheaths have a projecting

and ornamental foot, but, with the exception of those from Ovambo-

land, the sheaths are in two pieces bound together, and the foot serves

to keep the lower binding from slipping off. In the Ovambo knives

it has no such function to serve, as the sheath is in one piece. It is

possible that this form may have developed from the two-piece form,

and that the foot, remaining as an ornament after its use was for-

gotten, may have developed as such from the reasonable dimensions of

fig. 4, Plate LX XIX, to the very unwieldy size of fig. 2, Plate LX XVI.

According to Schinz (11, p. 279), the foot is said to represent ox-horns,

but this could not apply to some of the smaller feet. Mr. A. Macdonald

(8) suggests that the shape of the foot of the ordinary knives is copied

from the chief’s knife, which represents the large golden hawk, which

appears to have some mythical significance among the Ovambo

(5, p. 3). It is an interesting suggestion for which I have not yet

been able to find any further corroboration. Hambly states that

the foot is often held between the feet to facilitate the withdrawal

of the knife (7, p. 175), but it would hardly be developed for that

purpose. The shape of the head, too, is often ornamental. The

original metal ornamentation was undoubtedly copper, which is

mined in the country, but contact with Europeans has brought brass

(Plate LX XXIII, fig. 3) and hoop iron (Plate LXXXII, figs. 1, 2,

and 3) into use. Besides being ornamental, these bands serve at

the same time to prevent the sheath from splitting, as, slender

to begin with, the continual friction of the knife blade tends to

wear away the edges inside. The wire is fastened by inserting

each end into the wood of the sheath, and the metal plating is

fastened at the back, either by nailing with small copper nails down

each edge (Plate LX XVI, fig. 4, handle)—seldom do they overlap—

or by hooking a short piece of thick copper wire through holes in

each side (Plate LX XVI, fig. 3), or even by tying with string (Plate

LXXVI, fig. 4, sheath).

Ovambo Knives. 257

The wood used for the handle and sheath is taken from the Omuandi

tree (Diospyros mespiliformis, Hochst.), which is also valued as an

edible fruit tree. Other woods are sometimes used, but not as

commonly. Several specimens, particularly in Class I, look as if the

whole scabbard has been made in one piece, then cut in two, hori-

zontally, for the handle and sheath, and then hollowed out for the

tang and blade. The first step in the hollowing appears to have

been to pass a red-hot wire through the sheath and revolve it, thus

producing the hole in the base of the foot and the circular centre

of the elliptical opening at the top of the sheath. The same would

be done for the handle, or if handle and sheath were carved in one

the wire might be passed through the whole scabbard before separat-

ing the two parts. The method described by Mr. A. Macdonald (8)

is that “first they scoop out as much as is possible with a knife,

and then burn the rest out with the blade of the knife.” He says

nothing about the hole, but it might quite possibly act as a starting-

point. The opening in the sheath is probably cut at the same time

to facilitate the work. Schinz (11, p. 279) gives it as his opinion

that this was the whole function of the opening, “because the native

not only lacks the tools to hollow out a thin or narrow object, but also

lacks the skill to make a wooden object out of two pieces fastened

together.” In contradiction to the first point, the Museum has two

specimens, 3134 4 ana) (Plate LXAXVII, fig. 3), with no opening down

the front, nor have any of the handles an opening down the front. (It

is interesting to note that the only specimens in the Museum collection

with no opening down the front have a hole in the foot, and the only

specimen without the hole (S.A.M., 4846 (2), Plate LX XVIII, fig. 3) has

an opening.) The second point seems to be too unlikely to be worth

consideration, particularly as all the other tribes in South Africa

make their knife sheaths in two pieces. On the other hand, the

copper-covered specimens, even those made for covering, all have

the V-shaped opening in the wood under the copper, which would

seem to support Schinz’s theory.

The copper was mined in the country (2). It was smelted from

ore obtained in the vicinity of the present mines in the Grootfontein-

Otawi district, south-east of Ondonga. The Ondonga tribesmen were

the most noted copper-smiths, and obtained their ore half as tribute,

half as barter, from the Bushmen who collected it. It was all taken

to the chief, who had it smelted and sold in ingots. The Ondonga

were the only tribe that exported copper, trading it a great distance

for pottery and tobacco (11, p. 293; 4; 5, p. 35). The copper plate

258 Annals of the South African, Museum.

in the knives is beaten quite thin, and the wire is flat rectangular in

section, and in every case but one 1s twisted spirally before binding on

the knife. The exception is 8.A.M., 3310 ,,) (Plate LX XVI, fig. 3), on

which it is flat, with diagonal incisions on the edges. On most of the

specimens, particularly 8.A.M., 5059 (Plate LX XVI, fig. 1), the wire

appears from the sharpness of the edges to have been cut in thin strips

rather than drawn.

Iron was also formerly obtained from native mines in the country,

and a great deal was and still is obtained in South Angola, north

of Evale. “In former years iron was obtained from native ores

mined in Angola, but since access to the mines has been cut off”

(presumably by European political divisions) “scrap-iron of every

description has taken the place of native iron” (5, p. 35). Ténjes

(13, p. 59) writes that each smith fetched his own iron from the

mines north of Uukuanyama. These are probably the Angolan mines

mentioned above, as the northern portion of the Uukuanyama tribe

is situated in South Angola. According to custom, they could only

fetch it once a year at the time of the great national festival (Osimanja).

They took helpers with them, and if possible smelted the iron ore on

the spot, so as to make it easier to carry. Hstermann (2) contradicts

Hahn’s statement that access to the Angolan mines has been cut off,

and says the yearly pilgrimage still takes place. Only certain tribes

make iron articles and trade them to the others. The Uukuanyama

are the most noted iron-smiths (11, p. 293; 5, p. 36), and their knife

blades are much better than those of the Ondonga. Native smiths

undergo a long period of training and show great skill, considering

their primitive implements, which consist of a hammer (ofilua), tongs

(onuato), anvil (osikalo), and bellows (omupepo) (5, p. 36; 18, p. 59).

Some smiths specialise in certain articles. A large number are still

to be seen at work in certain districts, particularly in Uukuanyama,

but European contact and competition is fast causing their dis-

appearance (3; 13, p. 59).

The knives are distributed firstly over the whole of Ovamboland,

spreading from there north into Angola, where they are found among

those of the Uukuanyama tribe who live over the border (7, p. 175),

and among the Humbe natives, their neighbours to the west, on the

north and west banks of the Kunene. (Baum, pp. 55 and 70, has

photographs of natives, presumably Ovahumbe, wearing these

knives.) In the east they are found among the Ovakuangari

in the bend of the Okavango River (Baum, p. 67, photograph of

the Kuangari chief, ““whose subjects belong to the Ovambo tribe,”

Ovambo Knives. 259

wearing this type of knife), and farther down the river among

the Mambukushu (10, pp. 296 f.), from where they have had a

definite influence on those of the Barotse tribes living on the

western bank of the Zambesi. Presuming that the form was

evolved in Ovamboland, it is very easy, in view of the constant

trading, and the presence of three big rivers, Kunene, Okavango,

and Zambesi, to see how it could have spread north and east.

It must not be forgotten, however, that the Ovambo travelled

across Africa from the east, so that the present form of the Ovambo

knife may be a local development of the Barotse form, rather than

vice versa, or both may have acommon ancestor. But it seems fairly

evident that in its present form it has travelled eastwards again.

Miller (9) shows two knives identical with those figured on

Plate LX XXI of this volume. They were collected in “‘Zambesia,”’ but

obviously do not belong there, and must have travelled down the

Zambesi, after reaching it via the Ovakuangari and the Mambukushu.

In Angola the Vakuanyama are only another section of an Ovambo

tribe, and it is in their territory that the mines which supplied most

of the iron ore are situated. Trade between the various tribes as

far as those north of the Kunene has always been brisk (13, pp. 84

and 85). The Ovambo tribes also sent expeditions to the Mambu-

kushu (18, p. 85), who in their turn traded back along the Okavango

River, as far into Ovamboland as the Ondonga. According to

Passarge (10, p. 299) knives made from raw iron obtained from the

Barotse, were among the articles traded to the Ovambo for corn

and copper by the Mambukushu. It is quite possible that they

adopted the pattern as a better trading proposition, as no other type

of knife appears to be used in Ovamboland. The Mambukushu also

traded extensively with the Barotse Zambesi tribes on the east

(12, p. 210), from whom they obtained not only raw iron but finished

knives. Culturally the Mambukushu have much in common with

both the Ovambo and the Barotse, though they claim to be descended

from the latter. Passarge thinks they must have learned smithing

while still in Barotseland, before settling in their present district,

because iron ore is not found there and they work only in imported

iron(11). All these tribes too were visited by the Mambari (Portuguese

half-caste traders) until the ivory which attracted them was exhausted.

There is an interesting mixture of types in the example figured by

Passarge, and reproduced on Plate LX XXIII, fig. 1, in which the

handle inclines to the Barotse style and the sheath to the

Ovambo.

VOL. XXIV, PART 5D. 18

260 Annals of the South African Museum.

Most unfortunately very few of the specimens in the Museum have

been labelled as more than ““Ovambo,” without any details as to the

district or tribe from which they were procured, which, together with

the lack of literature on the subject, makes it impossible to correlate

style and tribe, if such a relation exists, which is doubtful. The

only thing that is apparent from the literature is that the best blades

were made by the Uukuanyama (5), and the most artistic sheaths

by the Ondonga (18, p. 59).

The ordinary knives may be classified according to the shape of

. the scabbard. I have made three main divisions.

I. Made of hard lght-coloured wood, rather yellowish in the new

and unused specimens, but toned down to a dull brown in the older

ones; slender, with narrow opening in the front, and small or medium

winged foot; edges branded, with raised portion in the centre, back

and front, left unbranded and outlining the shape of the scabbard;

handles narrowing and then flanging again to a knob, some shouldered,

some not; the blade varies from lancet to V-shape, and has a slight

thickening on both faces for the midrib.

(a) 8.A.M., 3133, 3135 (Plate LX XVII, fig. 1), 4846 (4), and two

exceptions, 3134 ( ana 9) (Plate LX XVII, fig. 3).

The foot is an ellipse both in front view and in horizontal

section; in the latter the greatest curve is at the back. The

projection is bored horizontally (except 3134 (,), vertical). The

raised portion in front is rounded and decorated. 8S.A.M.,

3134 ( ana) are exceptions to the whole collection in that they

have no opening down the front of the sheath. The blade is

lancet-shaped, with ordinary midrib and sharpened edges.

The handle is unshouldered, with narrow neck and _ short

flanging knob.

(b) S.A.M., 4122, 4123 «) (Plate LX XVII, fig. 2), and 4846 ().

The foot is similar to (a), but more curved along the bottom

and less on top. The projection is bored horizontally. The

raised portion in front rises to a central ridge, and is decorated.

The blade is straight to V-shaped (possibly lancet when new)

with ordinary midrib, very slight in two specimens, but more

definite in 4123 (,), and ogee in parts (probably accidentally).

The handle has slight shoulders, a long neck, and square-oval

knob, except 4123 ,), which has definite shoulders, a narrow

neck (no length), and round-oval knob.

Ovambo Knives. 261

Foot straight along the top and very curved at the bottom;

it is exceptional in that it has no hole in the base. The pro-

jection is bored horizontally. The raised portion in front is

flat and decorated. The blade is straight, with ordinary fairly

pronounced midrib. The handle is shouldered with medium

neck and round knob.

S.A.M., 361, 1736 G ana 3) (Plate LX XVIII, figs.1 and 2), 1736 (),

4846 (,) (Plate LX XIX, fig..4), 5058 (.) (Plate LX XIX, fig. 2),

and 5058 4) (Plate LX XIX, fig. 1), which is slightly variant.

The foot has a pronounced curve along the bottom, while

the top is straight in three specimens (1736 @ ang3) and 4846 (,)),

a straight slope in two specimens (5058 q anq2)), and curved like

(a) in two specimens (361 and 1736 (,)). The projection is bored

vertically (except 1736 ,) horizontal, and 1736 () not at all).

The raised portion in front is well raised, rises to a central

ridge with slightly concave wings, and is undecorated. The

blade is straight to V-shaped in four of the specimens (possibly

wear from very slight lancet shape), and lancet in three (4846(,,

and 5058 (anaa)). There is a slight, ordinary midrib, except

in 5058 (9), where it is pronounced, and in 5058 (,), which has

none at all. The handle is unshouldered with slender neck

and flanging knob, round in four specimens, and round-oval in

three (361 and 5058 ( ana2)-

(e) S.A.M., 2101 (4), 2159, 2270 (Plate LX XIX, fig. 3), and 5185

(Plate LX XX, fig. 3).

The horns of the foot are very pointed, the bottom has a

pronounced curve, and the top has a straight slant or a slight

curve. The projection is bored vertically. The raised portion

in front is undecorated and slightly rounded (except 5185, which

has a small conical knob on the sheath and lacks the raised ridge

round the V-shaped opening). The widest part of the raised

portion is at the top of the opening (except 5185), instead of the

top of the sheath as in the other specimens. The blade is lancet-

shaped in two specimens (5185 and 2101 4), and V-shaped in

the other two. There is an ordinary midrib—shght to medium.

The handle is shouldered, with a narrow neck (no length), and

a round knob.

8.A.M., 5294 (Plate LX XX, fig. 2) and 5368 (Plate LX XX, fig. 1)

belong to the same class but are rather exceptional. In 5294 the blade

—

262 Annals of the South African Museum.

is V-shaped with medium ordinary midrib, and in 5368 tends to lancet

shape with very much raised ordinary midrib.

Il. Made of hard light-coloured wood toned down to a dull brown;

large, with medium opening in the front of the sheath, a large decorated

foot, and a handle with a large flat knob. The branding is similar

to Class I.

S.A.M., 50 (fig. 4), 2101 G2 ana3y (Figs. 1, 3, and 5), 3315 (fig. 2),

and 5358 (fig. 6)—all on Plate LXXXI.

The plate shows the major differences in the scabbard, which

do not appear to be correlated at all. The projection is bored

vertically (except 2101 (,), both ways). The blade is lancet-shaped

(except 21015), which is V-shaped with ordinary midrib). S.A.M.,

2101(.) (fig. 3) is the best example of the type of blade—very

curved lancet-shape, with flat wings, and a raised, almost semi-

circular midrib on the front surface, and a slightly raised ordinary

midrib on the back. This type of midrib is not found on Class I

knife blades nor on the arrow blades, but is found on Class III

knife blades, and on all the assegai blades in the Museum collection,

the majority of the latter having it on both surfaces.

To this class belongs the knife figured by Miiller and Snelleman

(9), collected in “ Zambesia.”

III. Made of hard red wood capable of a high polish; short, with

wide openings in the front of the sheath, a funnel-like knob on the

squat handle, and a rather squat triangular foot. They are not

branded nor have they a raised portion back and front, but all have

a band of leather or tin round the top of the sheath.

S.A.M., 3309 (fig. 3), 3313 (fig. 1), and 3314 ganas) (fig. 2)—all

on Plate LX XXII.

The projection is bored vertically, and the scabbards are practi-

cally identical except for the extra width of 3313. The blade is

the same type as Class II, though in 3314 « ,,42) the midrib is so

worn as to appear almost ordinary.

The knife figured by Hambly (7, Plate xvi, fig. 10), from the

Vakuanyama of Angola, is like Class I, a; S.A.M., 361, an example

from Humpata, Angola, is like Plate LX XVIII, fig. 1, Class I, d; and

those shown in Baum’s photographs of the Humbe natives (1, pp. 55

and 79) are like Plate LX XX, fig. 3, Class I, e, and Plate LX XXIII,

fig. 3.

The copper-covered knives are in two classes: those made expressly

for covering with copper, of which there are two types, illustrated on

Ovambo Kniwes. 263

Plate LX XVI, fig. 1 (S.A.M., 5059), made of soft light-coloured wood,

and fig. 2 (S.A.M., 2102(,), made of hard light-coloured wood; and

secondly, those which appear to have been made as ordinary knives

and covered later with copper or brass—Plate LX XVI, figs. 3 and 4

(S.A.M., 3310 G ana2, and Plate LXXXIII, fig. 3 (S.A.M., 1735 (,)).

It is difficult to say whether these latter belong to the same class as

the former, even though all copper is the perquisite of the chief (11,

p- 293), because he sells it. The blades appear to follow the type of

scabbard to which the knives belong.

S.A.M., 1735 (,) (Plate LX XXIII, fig. 3), has a worn lancet blade

like Class I above, which it most resembles in shape of scabbard.

8.A.M., 2102 (4) (Plate LX XVI, fig. 2), has a worn V-shaped blade,

with a semicircular midrib on one side, like Class II, to which it

belongs in shape of scabbard, but S.A.M., 1735 (.), which is

identical in shape, has a V-shaped blade with an ordinary midrib.

S.A.M., 5059 (Plate LX XVI, fig. 1) and 1735(,) belong to none of

the above-mentioned classes. 5059 has a slight but definite ogee

section (the only one in the collection), and 1735 (3) is almost

flat with a scarcely perceptible ordinary midrib.

S.A.M., 3310 (,) (Plate LX XVI, fig. 3), like Class III, has a lancet-

shaped blade and semicircular midrib, but 3310 (9) (Plate LX XVI,

fig. 4) of the same class, has a V-shaped blade and ordinary

midrib.

The information available shows three types of Mambukushu

knife (simenti) :—

1. The knives illustrated by Passarge (10, p. 297) (Plate LX XXIII,

figs. 1 and 2). These are very similar to Ovambo knives in

the Museum.

2. S.A.M., 478 (Plate LX XXIII, fig. 4), a specimen from Sesheke,

Zambesi (not necessarily made there), which might be called

an intermediate type, the shape being Ovambo and the decora-

tion Barotse. There are two knives in one sheath, similar to

Class II, Ovambo, with cross-hatched diamond decoration

that is not characteristic of Ovambo knives, but does appear

on some very modern wooden utensils in the Museum collection.

3. S.A.M., 4861 ana) (Plate LX XXIV, figs. 1 and 4), two specimens

labelled ““Mampukushu, Barotseland,” similar to but much

larger than Class III, Ovambo, with ridged and branded

decoration, which, although not characteristic of Ovambo

knives, is seen on two modern snufi-boxes (S.A.M., 5435

and 5436).

264 Annals of the South African Museum.

Two Mambunda knives, 8.A.M., 2936 and 4860 (Plate LX XXIV, figs.

2 and 3), though different in shape, show relationship to the Ovambo

style in the openings in the sheath, one having them back and front.

The Bushmen tribes which live in scattered groups over most

of this area have adopted many articles from the Ovambo, and

amongst them knives. Two specimens, 8.A.M., 5408, made by the

Bushmen of the Sandveld, and a specimen in the University collection

(U.C.T., 35-108, kindly lent by Mr. A. J. H. Goodwin), made by

the Kung Bushmen in the north-east of 8.W. Africa, though roughly

made, are so like the Ovambo knives of Class I that it is not necessary

to illustrate them. The University specimen differs in that the sides

are not branded. It has a band of the tin off the back of a mirror

round the neck of the handle. The Museum specimen is identical

with the knives in Class I, a.

REFERENCES.

1. Baum.—Kunene-Zambesi Expedition, pp. 55, 66, 67, 79. Berlin, 1903.

2. EsT—ERMANN.—‘Ovambo,” Zeitschrift fiir Ethnologie, vol. lxiii, p. 45. 1931.

3. GALton.—Tropical South Africa, p. 223. 1851.

4, Guricu.—Siidwest Afrika, p. 143. 1891.

5. Haun.—‘The Ovambo,” Native Tribes of S.W. Africa, pp. 35f. Cape

Town, 1928.

6. Hann.—‘Preliminary Notes . .. ,” Journ. S.W.A. Sci. Soc., vol. iii, p. 5.

Windhoek, 1927-28.

7. Hampiy.—‘The Ovimbundu of Angola,’ Field Museum Anthrop. Series,

vol. xxi, No. 2, p. 175. Chicago, 1934.

8. MacponaLp.—St Mary’s Mission, Ovamboland. Private letters. 1935.

9. MULLER and SNELLEMAN.—Industrie des Caffres, pl. x, fig. 1. Leyden, 1892.

10. PassarGe.— ‘Die Mambukushu,” Globus, vol. Ixxxvii, pp. 232, 295 ff.

Brunswick, 1905.

11. Scuinz.—Deutsch Siidwest Afrika, pp. 279, 293. 1891.

12. ScHuLtz and Hammar.—The New Africa, pp. 209f. 1897.

13. ToéNnsES.—Ovamboland, pp. 56, 59, 67, 84, and 88. Berlin, 1911.

14, TéNsES.—WéGrterbuch der Ovambo-Sprache, p. 154 (omutikili). Berlin, 1910.

EXPLANATION OF PLATES.

Puate LXXVI.

1. S.A.M., 5059. Knife with sheath of soft wood, hollowed at the top for the

blade, and having a short V-shaped opening in the front. The lower edge is

covered with copper plating, nailed along the top with copper nails and welded

at the corners. The mouth has a narrow band of copper, which covers the pro-

Ovambo Knives. 265

jection at the back and is hooked together in the centre with copper wire. The

remainder of the sheath is wound round with spirally twisted flat copper wire,

each coil of which fits into a groove made in each edge of the wooden sheath.

The base of the handle is covered with copper plating fastened in the same way,

and the top of the knob is covered with a thin piece of brass hammered down.

The neck is loosely wound with similar wire. The short projection at the back

is bored vertically. The short blunt blade has a raised midrib and, contrary

to the general rule, is slightly ogee in section.

Inches. Inches.

Total length . 73 Width of handle at base . 13

Width at base 103 Width of handle at neck ine

Depth of copper plating at base 3 Diameter of knob . 15

Width of top of sheath 13 Length of blade : 3795

Depth of copper plating at top +% Greatest width of blade . 1,3,

Length of handle

Ondonga.

2. S.A.M., 2102(;). Knife with sheath and handle of hard wood, outstanding

for its breadth across the foot, which is separate from the rest of the sheath. It

is covered with copper plating, which overlaps and is nailed down on the back.

It appears to have had two round discs nailed on the front near the centre. The

stem of the sheath is inserted under the copper in front, and cut to fit at the back

of the foot, but this has not proved strong enough, and it has had to be bound

with fine (European made) brass wire. The upper portion as far as the opening

and including the projection at the back (which is bored vertically) is covered

with copper, overlapping and nailed down at the back. The remainder is wound

with spirally twisted copper wire. The lower part of the handle is covered with

copper plating fastened by hooking together with copper wire. The neck and

knob are wound with spirally twisted copper wire, leaving the flat top uncovered.

The blade is narrow, V-shaped, with raised midrib.

Inches. Inches.

Total length . 9% Width of handle at base . ae, ibe

Width across foot . 174°5 Width of handle at neck F 4

Width at bottom of sheath 1,1, | Diameter of knob . Sn wl

Width at top of sheath 12 Length of blade. ‘ anes

Length of handle 34 Greatest width of blade . : 13

No history.

3. S.A.M., 3310 (4). This specimen appears to have been an ordinary wooden

knife, similar to fig. 2, Plate LX XXII, ornamented with copper. The sheath is

made of hard reddish wood, very flat-oval in section, and has a wide opening in front,

almost the entire length of the sheath. The solid portion at the top is made up

of two rectangular pieces of wood hammered in and reinforced bya band of leather.

It is difficult to say whether it was made this way, or mended after splitting.

The crescentic piece of copper on the foot is seated on grease, each end tapering

to a long point which is passed through the foot and bent down at the back.

Two similar but straight pieces form bands, one round the mouth over the leather,

and the other towards the centre of the sheath. The remainder is wound with

flat copper wire. The handle is covered with two pieces of copper, fastened like

266 Annals of the South African Museum.

figs. 1 and 2. The roughly-made blade is leaf-shaped with raised midrib, which

is inclined to be semicircular in section.

Inches Inches.

Total length . é : eh NOS Width of handle at neck £

Width across foot . : . 32 Diameter of knob . 18

Width at top (without copper) 1,5 Length of blade : 53

Greatest width of blade . 14

Length of handle . : ees.

Width of handle at base 1

No history.

4, §.A.M., 3310 (.). Knife similar to fig. 3. The foot, the top of the sheath

and the lower portion of the handle are covered with copper plating, nailed at

the back of the handle, nailed and hooked on the foot, and lashed with twisted

fibre at the top of the sheath. The remainder of the sheath and the neck of the

handle are wound with twisted copper wire. The blade is V-shaped with a slight

midrib.

Inches Inches

Total length . 10% Width of handle at shoulders . as

Width across foot . : . 44 Diameters of top of knob AE 3

Greatest width of sheath . 2 Length of blade 54

Length of handle . ; . 3,2, Greatest width of blade . +2

Width of handle at base . 13

No history.

Prate LXXVILI.

1. S.A.M., 3135. Large knife with light-coloured wooden sheath, which

terminates in a small winged foot, and has a V-shaped opening down the front.

The opening is outlined by a narrow raised ridge, which continues along the base

of the foot, and at the top of the sheath becomes a platform, which continues in

the same shape on the handle. The back has a similar slightly raised ellipse,

both being left their natural colour while the edges are branded. The projection

at the back of the sheath is bored horizontally. The blade is lancet-shaped with

a definite but very crooked midrib, and sharpened edges.

Inches. Inches.

Total length . : 5 . 252 Diameters of knob 15% X 135

Width across foot . 211 Length of blade. : ales

Greatest width of sheath 24 Greatest width of blade . .

Length of handle 6g Length of projection 455

Width of handle at neck 2

From Karibib, 8.W. Africa.

2. S.A.M., 4123 (;). Knife with wooden sheath and handle, below the knob

of which there is a band of tightly-wound fine copper wire. The edges are branded

and the raised central portion back and front left natural colour. The cylindrical

projection at the back is bored horizontally near the top, and has also a hole in

the top, as if the maker had commenced boring vertically. A short thong is

passed through and knotted in a loop. The narrow sharply-pointed blade has a

distinct midrib.

Ovambo Knives. 267

Inches. Inches.

Total length . 174 Diameters of knob ; 12 x118

Width across foot . 2,5, Length of projection. 2 28

Greatest width of sheath 13 Length of blade. : . 1352

Length of handle 44 Greatest width of blade . 15

Width of neck 3

Uukuanyama, Mafa, Ovamboland.

3. S.A.M., 3134. Knife similar to fig. 1. Unlike all the others, it has no open-

ing in the front of the sheath, and has merely a continuation of the raised ellipse,

left unbranded. There is a similar plain raised ellipse on the back, only the

edges and the triangular decoration in front being black. The projection at the

back is bored vertically. The blade is lancet-shaped, and has a slight midrib.

Inches Inches.

Total length . : 5 . 8% Diameters of knob : Pe xls

Width across foot . : 74 Length of blade. : . 44

Greatest width of sheath 14 Greatest width of blade . : 44

Length of handle 3% Length of projection 4 wll

Width of handle at neck 3

From Karibib, 8.W. Africa.

Puate LX XVIII.

1. S.A.M., 1736 (,).. Knife with wooden handle and sheath, having the edges

branded and a raised unbranded central portion which outlines the V-shaped

opening and the lower half of the foot, and rises concavely to a central ridge at

the top of the sheath and on the handle. The projection is bored diagonally

upwards from the sides near the top, so that the holes meet in the centre, and

a conical hole is bored vertically from the top to meet them. The effect, inten-

tional or not, has been to fix the thong loop more or less immovably. The blade

is V-shaped with raised midrib.

Inches. Inches.

Total length . ; : . 132 Diameters of knob : 1,5, x1}

Width across foot . 24 Length of projection ; - 25.

Greatest width of sheath 13 Length of blade 44

Length of handle 4.5 Width of blade at top g

Width of handle at neck

Northern Ovambo.

toa |

2. S.A.M., 1736 (3). Knife similar to fig. 1, very long and slender, with more

elaborately carved neck and knob. The thin projection at the back is not bored

at all. The slender blade is lancet-shaped with a slight midrib.

Inches. Inches.

Total length . : : . 262 Diameter of knob . 15%

Width across foot . 1 Length of blade : gi

Greatest width of sheath 13 Greatest width of blade . 1s

Length of handle 52 Length of projection 22

Width of handle at neck 4

Northern Ovambo.

268 Annals of the South African Museum.

3. S.A.M., 4846 (2). Knife with light-coloured wooden handle and sheath. In

section the lower part of the sheath is oval, but the upper part and the handle

have had the edges cut square. The sides of handle and sheath are branded,

leaving a raised unbranded central portion back and front. This portion is flat

until half-way up the handle, where it rises to a central ridge. The triangular

decoration is branded. This is the only knife in the Museum collection with no

near the centre, and has a short thong passed through twice, attaching the knife

to a larger thong loop which is fastened by looping one end through a slit in the

other. There is a wide band of hoop iron round the sheath below the projection,

fastened at the back with brass wire. The blade is lancet-shaped with raised

midrib.

Inches. Inches.

Total length . : : . 16% Diameters of knob . . 11x

Width across foot . : wr 8 Length of projection : . 143

Greatest width of sheath 1411 Length of blade. 2 . 10+

Length of handle : eg De Greatest width of blade . =

Width of handle at shoulders . 3

Uukuanyama.

Puate LXXIX.

1. S.A.M., 5058 (q). Knife with wooden handle and sheath. At the back of

the sheath is a saddle-shaped projection which is bored cylindrically for the

passage of a thong. The greater part of the back and front is raised and left

natural colour; the sides are blackened by branding. The raised edges outlining

the opening are continued as an edging to the top of the handle, and the wood

between them rises to a central ridge. The knob of the handle is oval with straight

ends. The blade is lancet-shaped and quite flat in section until } inch before

the sharpened edges.

Inches. Inches.

Total length . : ‘ . 15h Length of handle . ; » 58

Width across foot . : wbyS Diameters of top 27, x15

Greatest width of sheath 24 Length of projection . 23

Greatest width of blade . 1d Length of blade st

Ondonga.

2. S.A.M., 5058 (,). Knife similar to fig. 1, but without the raised edging to

the central light-coloured portion. The opening in the front is small. The

blade is lancet-shaped with raised midrib.

Inches Inches.

Total length . st Diameters of top . 1i x1}

Width across foot . 255 Length of projection : F 1f

Greatest width of sheath 1} Length of blade. : . 42

Length of handle 32 Greatest width of blade . 15

Ondonga.

3. S.A.M., 2270. Knife with wooden handle and sheath. The sides are

blackened back and front by branding, leaving an unbranded portion, flat at the

Ovambo Knives. 269

back and raised in front, where it forms an elongated diamond, the centre being

at the top of the opening in the sheath. The raised portion is slightly rounded.

The handle has very slight shoulders before the flanging, roughly oval knob.

The projection at the back is bored vertically, and has a leather thong held by

three iron beads where it widens at the base, and by a wedge at the top to a strap.

The strap bends over, forming a loop through which the other end of the thong

is passed. The blade is V-shaped with a slight midrib.

Inches. Inches.

Total length . : : =0 122 Diameters of top 14 xit

Width across foot . . 3,3; Length of projection . 25

Greatest width of sheath . 1% Lengthof blade . : . TAB

Length of handle . : . 43 Greatest width of blade . 1,5

Width of handle at shoulders . 3

No history.

4. S.A.M., 4846 (;). Knife with wooden handle and sheath. Near the base

there is a jointless ring of hide to prevent splitting, which has already occurred

higher up. The sides are branded and the centre left plain, rising to a central

ridge and making an ill-defined rim round the opening. The projection is bored

vertically and has a leather strap, split at the top to make a loop and held at the

bottom by a cross-piece of leather passed through the end. The blade is lancet-

shaped with a raised midrib. This is an old and very worn specimen.

Inches. Inches.

Total length . : : . 142 Diameters of top 14 x1}

Width across foot . : 2 ole Length of projection 5 . LS

Greatest width of sheath ee Length of blade. ; . oF

Length of handle . : . 4% Greatest width of blade . cree bell

Width of handle at neck 4

Uukuanyama.

PLATE LXXX.

1. S.A.M., 5368. Knife with handle and sheath, made of reddish wood, long and

slender, rounded oval in section, with unbranded portion raised in front but not at

the back. The knob flanges widely from a narrow-shouldered neck. There is a

binding of thick copper wire near the foot and one of fine copper wire at the base

of the handle; both appear to be native-drawn. The saddle-shaped projection at

the back is bored horizontally for a thin strip of hide long enough to go round the

waist and over one shoulder. The blade is lancet-shaped with raised midrib.

Inches. Inches.

Total length . : ; ae pe Diameters of top. : 13 x1}

Width across foot . : Ede, Length of projection : - 5525

Greatest width of sheath 15 Length of blade. y . 16}

Length of handle 5L Greatest width of blade . 5 il

Width at neck ins

No history.

2. S.A.M., 5294. Knife with wooden handle and sheath. On either side of

the opening the wood is quite flat, the back of the sheath and handle being rounded.

The edges, knob, and back of the scabbard are branded black and the remainder

270 Annals of the South African Museum.

left natural colour. The unbranded part above the opening and on the handle

is not raised, but is separated from the black edges by a deep semicircular groove

on each side. The handle is shouldered and has a thin round neck, which is bound

with fine brass wire and supports a shallow knob. There is a 34-inch binding

of fine brass wire at the bottom of the sheath, and four narrow bands of brass,

two on the handle and two on the sheath. There are three bands of hoop iron on

the sheath, two close together at the top and one lower down. The blade is lancet-

shaped with a distinct midrib and a very sharp edge. The projection is bored

horizontally near the top.

Inches. Inches.

Total length . 263 Diameters of knob 3 18 x14

Width across foot . 43 Length of projection se

Greatest width of sheath 1g Length of blade. 3 . 204

Length of handle . 58 Greatest width of blade . 13

Width of handle at shoulders . 3

No history.

3. S.A.M., 5185. Knife with wooden sheath and handle, the handle being

shouldered below the flanging knob. The edges of the front of the sheath and

the top of the foot are branded, leaving unbranded a narrow flat piece in the

centre below the opening, and a raised portion which has a conical point in the

centre above the opening, the back, and the remainder of the foot. The edges

of the handle are branded, leaving a plain raised central portion back and front.

The projection at the back is saddle-shaped and bored cylindrically for a loop

of thin twisted sinew. The blade is lancet-shaped with raised midrib.

Inches. Inches.

Total length . ; : - 102 Diameters of knob . 12x15

Width across foot . . 3%, Length of projection 3 . Ls

Greatest width of sheath = le Length of blade. : . CF

Length of handle og 4 Greatest width of blade . SRE

Width of handle at shoulders . 3

No history.

Puate LXXXI.

1. S.A.M., 2101 (1). Knife with wooden handle and sheath. The handle is

shouldered half-way and surmounted by a funnel-shaped knob, the flat top of

which is covered with small round depressions. The foot is straight along the

bottom, with the points curving forward. In section both handle and sheath

are flat oval. The cylindrical projection at the back of the sheath is bored verti-

cally and twice horizontally. The base of the handle, the top of the sheath,

and the decoration on the foot have been left natural colour and the rest branded.

The blade is lancet-shaped and has a very definite, almost semicircular midrib.

Inches. Inches.

Total length . < : . 14% Length of handle . : 7

Width across foot . 4% Diameters of knob 13x14

Width at shoulders rs Length of blade. : . 10d

Length of projection 2% Greatest width of blade . 0 Be

Greatest width of sheath 13

No history.

Ovambo Knives. AAAI

2. S.A.M., 3315. Foot of sheath of knife similar to fig. 1, but more slender,

and rounder in section. The foot is smaller and the points do not curve forward.

The raised portion is left natural colour, and the remainder branded. There is

a narrow band of tin at the base of the sheath.

Inches. Inches.

Depth of foot : : oe see Width across foot . 4 . 848

No history.

3. S.A.M., 2101 (2). Knife with wooden handle, deeply shouldered half-way for

a funnel-shaped knob, which is rounded on top and has a brass furniture tack

in the centre. The raised portion in front is unbranded. The blade is lancet-

shaped, with a raised midrib, semicircular in section, and is very finely made.

Inches. Inches.

Length of handle . : . 58 Diameter of knob . : hy ed

Width of handle at base . 143 Length of blade. : owls

Width of handle at shoulders . 1 Greatest width of blade . . IL

No history.

3a. Foot of sheath belonging to fig. 3, similar to fig. 1, but rounded along the

bottom, and not set straight on the sheath. The raised portion is unbranded.

Inches. Inches.

Depth of foot : : Bie Width across foot . 5 tenes

4, §.A.M., 50. Foot of sheath of same type as fig. 1, but the ends are inclined

to curve backward. A raised portion back and front is left unbranded, as is

the entire back of the sheath. There is a band of hoop iron at the base of the

sheath.

Inches. Inches.

Depth of foot : : . Width across foot . : TS

No history.

5. S.A.M., 2101(3). Foot of sheath of the same type as fig. 1, but that the

front is entirely branded, except for the raised portion of the foot. The centre

of the back and the projection are also unbranded. The foot is not set straight

on the sheath.

Inches. Inches.

Depth of foot 4 : aoe Width across foot . : 5

No history.

6. S.A.M., 5358. Knife with wooden handle and sheath, similar in type to the

foregoing, but larger and heavier, with an oval knob. The section of the foot is

the segment of a circle, straight along the front edge and curved at the back.

The edges of the scabbard and of the foot, back and front, are blackened by

branding, and the remainder is left natural colour. The projection at the back

is bored vertically. The blade is lancet-shaped with slightly raised, semicircular

midrib.

272 Annals of the South African. Museum.

Inches. Inches.

Total length . : : . 212 Width of handle at shoulders . 1,

Width across foot . 313 Diameters of knob : 2, x14

Greatest width of sheath 24 Length of blade. : . 10z

Length of projection 3 Greatest width of blade . Ste

Length of handle 57°5

No history.

Puate LXXXII.

1. S.A.M., 3313. Knife with wooden handle and sheath, having a comparatively

small winged foot, and open down the front, about 1 inch from the top of the

sheath, so widely that the edges are only about 4-inch wide, and the opening is

wider than the blade. There is a band of hoop iron round the closed part at

the top, fastened at the back, over a cylindrical projection, by a wire hooked

through holes in each edge. The projection is bored vertically and a double thong

is inserted, making a short loop at the bottom, and cut off short and wedged

with two splints of wood at the top. The handle is shouldered and flanges from

the neck to an oval knob, which is flat on top. The horizontal section of the whole

is a very flat oval. The blade is lancet-shaped, with a raised midrib.

Inches. Inches.

Total length . 13,5, Width of handle at shoulders . 3

Width across foot . 23 Diameters of knob : 13, x1

Greatest width of sheath 24 Length of blade. : : 82

Length of projection g Greatest width of blade . in Ue

Length of handle 4}

Ovamboland, West Coast.

2. S.A.M., 3314. Wooden sheath of the same type as fig. 1, fitted with a knife

like fig. 3, which is too long for it. The bands of hoop iron above the foot and

round the top are fastened at the back with an iron pin, hooked through each

side. The projection is bored vertically and has a looped thong, with the loop

at the top and a knot below the projection.

Inches. Inches.

Length . : 63 Width across foot . : ae bees

Greatest width at top. ; 1g

No history.

3. S.A.M., 3309. Knife with handle and sheath of the same type as figs. 1

and 2, but narrower and with narrower opening. The blade is lancet-shaped, with

raised midrib. A short thong is looped and knotted through the vertical bore

of the projection. The horizontal section of the foot is diamond-shaped.

Inches. Inches.

Total length . ; : - 123 Width of handle at shoulders . 3

Width across foot . 2,%, Diameter of knob . 45

Greatest width of sheath 1%, Length of blade 835,

Length of projection 1d Width of blade 1

Length of handle 33

Ovamboland.

Ovambo Knives. 273

Pratt LXXXIILI.

1. Knife figured by Passarge (Globus, 87, 1905, p. 297, fig. 20), with wooden

handle and sheath, having a crescentic foot, and a V-shaped opening in front.

The edges appear to be branded and the raised portion in front left plain. The

projection is bored vertically. No description or measurements are given.

Mambukushu.

2. Knife figured by Passarge (loc. cit., fig. 19), similar to fig. 1, but covered

with copper plating, which overlaps and is nailed at the back. No description

or measurements given.

Mambukushu.

3. S.A.M., 1735. Knife with wooden handle and sheath. There is a raised

ridge outlining the opening and left unbranded, as is the centre of the back.

Thin bands of brass and hoop iron, arranged more or less alternately, are nailed

on with copper nails, so as to cover the front entirely, but leaving a space down

the back. (Two bands are missing.) The band of tin at the mouth of the sheath

goes right round, covering the cylindrical projection, which is bored vertically.

The front of the foot is covered with thin brass discs of various sizes, nailed on

with copper nails.

Inches.

Total length . : ; eels

Width across foot . : lO

Greatest width of sheath .

Length of projection . . 24

Length of handle 4,

Northern Ovambo.

4. S.A.M., 478.

Two knives with wooden handles in one wooden sheath.

The blade is lancet-shaped with a raised midrib.

Inches.

Width of handle at neck ; 3

Diameters of knob . 15; x14

Length of blade. 5 . 8&%

Greatest width of blade . A 1

The

sheath is half an ellipse, terminating in a winged foot, and having two V-shaped

openings in front.

The handles of the knives continue the line of the ellipse with

their outer edges, and the blades are set in the sheath convergently. The handles

are shouldered, and have narrow-necked flat-topped knobs.

sheath is a single short projection, bored vertically for a thong.

V-shaped with distinct midribs.

At the back of the

The blades are

The sheath and handles are blackened by

branding all over, with the exception of a panel down the front of each handle

and a diamond pattern on the sheath, both cross-hatched.

Inches.

Total length . 3 4 . 84

Width across foot . ; : 4

Greatest width of sheath . 2yPe

Length of projection : el eel:

Length of handles ._ right 37’, left 34

From Sesheke, Barotseland.

Inches.

Greatest width of handles 5 1}

Width at shoulders : ; _9

Diameter of knobs right 1,3, left 14

Length of blades right 43, left 48

Greatest width of blades ; z

Piatt LXXXIV.

1. S.A.M., 4861(1).

Knife with wooden handle and sheath. The handle is

shouldered half-way, and topped with a slightly flanging peg.

shaped foot is flat and rectangular in section.

The diamond-

There is a long cylindrical pro-

274 Annals of the South African Museum.

jection at the back of the sheath, bored twice horizontally. The entire handle and

sheath are blackened by branding, and decorated in front on either side of the opening

by a series of triangles, enhanced with diagonal grooves, and on the remainder

of the front as well as on the back of the handle by bands of diagonal ridges in

alternate directions separated by single horizontal grooves. The rest is plain

but for a ring of half-moons cut out round the top of the handle. The blade is

lancet-shaped with a single continuous chevron incised down the midrib back

and front.

Inches. Inches.

Total length . : : . 204% Width of handle at shoulders . 12

Width across foot . 33 Diameters of top. ‘ 13 x 1,5

Greatest width of sheath 211 Length of blade. : . 113

Length of projection 4, Greatest width of blade . ae

Length of handle 542

Mampukushu, Barotseland.

2. S.A.M., 2936. Knife with wooden handle and sheath, having a short V-shaped

opening in front. The handle is round in section for two-thirds of its length, and

then flanges to a flat oval for the insertion of the blade. The sheath is rectangular in

section. The entire handle and sheath are blackened by branding. The sheath, plain

on the back and edges, has a series of triangles on each side of the front, enhanced

with diagonal grooves, and a raised and grooved band outlining the opening

and continuing down the centre. The lower part of the handle has similar

decoration back and front. There is a short projection at the back, bored twice

horizontally. The blade is V-shaped and slightly ogee in section, with an

incised continuous chevron down the midrib, back and front.

Inches Inches

Total length . : : . 811 Length of handle . : . 4b

Greatest width of sheath 24 Width of handle at base 1%

Width of sheath at waist ly Diameter of neck 41

Width of sheath at mouth 13 Length of blade , 4,5,

Length of projection ra Greatest width of blade . 13

Barotseland.

3. S.A.M., 4860. Knife similar to fig. 2, but there is a V-shaped opening in

the back of the sheath corresponding to the opening in the front. The edges

are grooved and the decoration on the front of the sheath and on both sides of

the handle consists of horizontal grooves, which outline bands of diagonal grooves

arranged in alternate directions. The projection at the back is bored horizontally.

Inches Inches

Total length . : : . 94 Greatest width of handle ale

Width across foot . : 4 28 Width of handle at neck 3

Greatest width of sheath . 2,3, Diameters of top . : 1x?

Length of projection reall Length of blade 4z

Length of handle . : . 4,3 Greatest width of blade . 12

Mambunda, Barotseland.

Ovambo Knives. 275

4, S.A.M., 4861(.). Knife similar to fig. 1, but with rounded foot to the sheath.

The back of the sheath is plain and branded black and the front is decorated with

alternate interlocking triangles of plain light-coloured, and grooved black. The

sides have ridges corresponding with the apices of the plain triangles. The pro-

jection at the back is somewhat below the top of the sheath, and is bored twice

horizontally. In section the sheath is a flat oval with the ends cut off straight.

Inches. Inches

Total length . 5 5 . 204 Diameter of neck . ; .

Width across foot . ‘ WEBB Diameter of top 3

Greatest width of sheath 27°5 Length of projection 2 a

Length of handle Bt Oy Length of blade. : 2 42%

Width of handle at shoulders . 12 Greatest width of blade . 13

Mampukushu, Barotseland.

VOL. XXIV, PART 5. I)

Ann. S. Afr. Mus., Vol. XXIV. Plate LX XVI.

M. Shav. Neill & Co., Lid.

Plate LX XVII.

Ann. S. Afr. Mus., Vol. X XIV.

Lid.

Neill & Co.,

M. Shaw.

Plate LX XVIII.

Ann. S. Afr. Mus., Vol. X XIV.

Neill & Co., Lid.

M. Shaw.

Plate LX XIX.

Ann. S. Afr. Mus., Vol. XXIV.

Neill & Co., Ltd.

M. Shaw.

Plate LXXX.

Ann. 8. Afr. Mus., Vol. XXIV.

Neill & Co., Ltd.

M. Shaw.

Ann. S. Afr. Mus., Vol. XXIV. Plate LXX XI.

M. Shaw. Neill & Co., Lid.

Ann. 8. Afr. Mus. Vol. XXIV. Plate LX XXII.

M. Shaw. Neill & Co., Lid.

Plate LX XXIII.

ft = ——— =x

Ra)

Ann. S. Afr. Mus., Vol. XXIV.

Neill & Co., Ltd.

Shaw.

Plate LXX XIV.

Ann. 8. Afr. Mus., Vol. XXIV.

SLU

RANA RA AAR Na ee tt nN CNN AM INC

i TONAL Be

Cy.

See

UGE

LGA»

OTE

RTA

SLELAA LTA

PAA Si MN AC ET

Healy arna lial Po OHS

| |

.: Ltd

Neill &

Shaw.

(277)

12.—Native Pipes and Smoking in South Africa.—By

Miss M. Suaw, B.A., Assistant in Ethnology.

(With Plates LXXXV-XCIX, and 2 Text-figures.)

THE use of the two narcotics, tobacco and hemp (dagga), seems to

have been well established amongst the native tribes in South Africa

by the time of the first European settlement at the Cape. Neither

plant is indigenous to the country, the former having come from

America, the latter from Asia, and it is supposed that hemp was

introduced on the east coast by the Arabs, and tobacco on the west

and east coasts by the Portuguese. Tobacco must in that case

have been introduced in the fifteenth and sixteenth centuries, and

it is probable that hemp came in earlier. Dos Santos (1), writing in

the sixteenth century, discusses the cultivation and use of hemp by

the people of eastern Mashonaland, but makes no mention at all of

tobacco. Certainly hemp had spread as far as the Cape Peninsula

by the time of van Riebeeck’s arrival. It is doubtful whether

tobacco had got as far, but the Hottentots soon became very fond

of it. Those people who did not cultivate their own narcotics

obtained them by barter from neighbouring people. Tobacco even

more than hemp seems to have been a trade article of the greatest

importance, which reached at times almost to a currency. Its value

in this direction was soon realised by European settlers and travellers,

who found it the most popular article of exchange and generally

preferred by the natives to their own variety.

It seems fairly certain that at first tobacco was used almost

exclusively for snuff, and hemp for smoking, and even to-day when

the Government ban on hemp has given tobacco first place in most

parts of the country for smoking as well as snuff, there are still some

places where it is used mostly as snuff. Hemp is occasionally used

for snuff, while some people are such slaves to their pipes that when

tobacco or hemp are lacking, they will put in them a live coal, a

piece of dry dung, or some other substitute.

Dagga was the name given to hemp by the Cape Hottentots, and

it has become a common name in the country, though each native

278 Annals of the South African Museum.

people has its own name for it. The true hemp is Cannabis sativa.

It has frequently been confused with an indigenous plant of an

entirely different family, Leonotis leonurus, of which the leaves are

very similar and might easily be confused although the flowers are

entirely different, those of the hemp being cream and insignificant,

and those of Leonotis large and briliant red. The Leonotis is often

referred to as wild dagga, but how much it is smoked is difficult to

say. It is probably one of a number of aromatic herbs which are

used as substitutes for tobacco or dagga, and have some narcotic

effect from the natural oils contained in them. Cannabis sativa is

reported to have grown wild in various parts of the country since

its importation, but was also cultivated.

Among the Bushmen the Namib are the only tribe that have

cultivated dagga to any extent and for a considerable length of

time (2). A few men among the Auen and Naron copied their

Bantu neighbours and grew a little (8), but on the whole any dagga

they have is obtained by gift or barter.

There is some discussion as to whether the early Cape Hottentots

cultivated dagga or not. At least six of the tribes are reported to

have done so, but Maingard (4), after considering the evidence in

the early records, concludes that the information is not reliable,

and that actually the western tribes had it by barter from the

eastern, who are known definitely to have obtained their dagga at

a big annual bartering between themselves and the Xosa. The

dagga was made up into little balls for use (6). The Nama Hottentots,

however, did grow their own. It was the only other plant besides

tobacco that they cultivated (6).

The Bergdama of 8.W. Africa grew a fair quantity of dagga at

one time, but latterly, under missionary influence, it has practically

disappeared (7).

The majority of the Bantu tribes must at one time have cultivated

dagga. With them it spread south and west until it reached the

Bushmen and Hottentots as well. There is definite record of the

fact that the Xosa, the southernmost branch, traded it to the Cape

Hottentots, and it was probably in that way that it spread west

and north to Namaqualand, and from the Nama to the Herero

who did not cultivate it. It would seem that dagga had not made

its appearance before the western branch of the Southern Bantu

started on their journey from the Eastern Highlands, as it was

reported unknown among the Ovambo as late as 1887 (8), and when

they did take to it, which was in comparatively few cases, they

Natwe Pipes and Smoking in South Africa. 279

obtained it from their Bushmen neighbours (9). The south-eastern

and central tribes cultivated it quite extensively; generally small

gardens were made at the back of the huts or in an old deserted

cattle kraal. Certainly in the beginning it must have been less

plentiful than latterly, as the earthenware pipe-bowls that have been

excavated in many parts of the eastern side of the country are

quite small compared with the really capacious bowls of modern

times (Plate LXX XVII). The Arabs, if it be they who introduced

dagga, also introduced their own method of smoking it, that is through

water, in order to cool the smoke somewhat. The fumes of dagga

are highly intoxicating and irritating to both lungs and throat.

Taken in moderation it does not seem to have any ill effects, but

the majority of those who smoke it inhale it very deeply into the

lungs, and consider coughing and choking and running at the eyes

necessary for a proper enjoyment, even to the point of becoming

unconscious. When thus smoked to excess it is considered to have

very serious effects on the health and brain. For this reason and

on account of the irresponsible behaviour of individuals under its

influence, its growth and sale have been prohibited by the Government,

but there is a considerable illicit traffic, by Europeans as well as

native and coloured.

Although tobacco since its introduction has always been greatly

sought after by all the peoples of South Africa, its use is not nearly

as old nor as widespread for smoking as for snuffing. The Bushmen,

Hottentots, and South-Eastern Bantu, particularly the Transkeien

tribes, seem to have become the most inveterate smokers of tobacco,

a fact which reflects the routes of the early European penetration

of the country. Smoking of tobacco is practised, however, by all

if to a lesser extent by some, and is gaining ground with the suppres-

sion of dagga. Nowadays, although the HEuropean-grown article

has been introduced everywhere, there are few tribes who do not

grow their own tobacco as well, although several of them when first

visited by Europeans were dependent on barter for their supply.

Generally each household has its own tobacco garden in or near the

village and apart from the other gardens. The leaves when dried

are made up in a variety of forms, generally some sort of roll in the

south and east, and some sort of ball or block in the north and west,

and these are the standard quantities when tobacco is used for trade.

With the exception of the Namib, who have cultivated it for

quite a long time, and a few isolated men among the Auen and Naron,

the Bushmen neither did, nor do they now, grow tobacco, but

280 Annals of the South African Museum.

obtained it by barter from their Bantu neighbours (10). It was the

first thing for which they begged European travellers among them,

and they still prefer it to any other payment for services rendered (11).

The Hottentots did not cultivate tobacco either in the early

days, but got it by trade, labour, and begging from Europeans (12).

In 1661 it was reported that the Nama visited by van Riebeeck’s

expedition had not seen it smoked, but all learned in a short time,

and within recent years they have cultivated it themselves (18).

The Bergdama have also cultivated tobacco for a long time, and

according to tradition had it from the south, where one of the

ancestors bartered a goat for it, long before the arrival of the first

Europeans among them (14). Comparatively few men cultivate it

and trade it to others, and in particular to the Herero, who also

obtained a good deal of theirs from the Ovambo in the early days,

but did not cultivate it on their own until the middle of the last

century, when someone started to do so and made his fortune out

of it (15). Even in Ovamboland the southern tribes hardly cultivate

tobacco at all. Those in the north grow quite a lot (16), but the

majority and the best comes from north of the Cunene River, whence

it is imported in round balls (17). Those who do grow tobacco have

to give a certain proportion of their crops to the chief (18). Tobacco

is the most important trade medium in Ovamboland (19), and is

mostly used as snuff, as on account of its high price only the well-

to-do can afford to smoke it (20).

Among the central tribes tobacco obtained a strong footing

somewhat earher, but certain of them until comparatively recently

did not grow their own. Among the Tswana, the Batlapin when

Burchell first visited them obtained a very inferior brand from the

Bahurutshe (21), and only started to grow their own about 1830 (22).

The northern Tswana have grown tobacco for a considerable time,

and do it up in cone-shaped lumps which they trade to the neighbour-

ing peoples, particularly the Bushmen (23). The Basuto have

cultivated tobacco since long before the appearance of Europeans

among them (24). They do it up im large rolls which are kept rolled

up in roughly made mats. The Venda, too, have grown tobacco

for a very long time, but until quite recently used it only for snuff (25).

After a preliminary grinding they mix it to a paste which they press

into oval pitted moulds, and when set the blocks can be traded or

stored until needed. The Mashona, too, cultivated tobacco almost

entirely for snuff. The Bashankwe tribe, west of Salisbury, grow

and export a great deal of tobacco to their neighbours, who prefer it

Native Pipes and Smoking in South Africa. 281

to any other (26). The Makaranga peoples pound their tobacco up

and mix it with wild honey before pressing it into conical moulds

to dry (27).

It is, however, the south-eastern tribes who are the greatest

smokers of tobacco. Almost every household or village has its

own tobacco garden, at the back of the huts or in an old deserted

cattle kraal. The cultivation is done by men, particularly the

older men, and occasionally old women, never by young women.

The leaves are carefully prepared, stalked, and dried, and are

then rolled up into long cigar-shaped rolls, which are wrapped in

roughly made grass mats for storing, and are often used for

barter (28).

Presuming, as we may, that dagga came into the country before

tobacco, we may take the water pipe to be the earliest form of pipe

in the country. Archeological finds have included very many

earthenware and stone pipe-bowls of the type used with a water pipe

but considerably smaller than the modern examples (Plate LX X XVII,

figs. 1-6). These pipe-bowls have been found over a large area from

Zimbabwe to the Natal coast. It is interesting to compare with

them a modern Indian pipe-bowl obtained in the Indian Market, at

Durban (Plate LXX XVII, fig. 15).

The modern water pipe (Plate LX X XVIII) consists most commonly

of the horn of an ox or one of the larger antelopes, into the side of

which is inserted a length of bamboo reed, or sometimes a hollow

wooden tube, which has fixed on its top end a bowl, which varies in

size, and is made of baked clay, carved stone, or even wood. Some-

times at the junction of reed and horn gum or wax is used to make the

joint water-tight. In making the stone bowls, after the cylinder has

been roughly shaped, the bore is commenced from each end alternately,

until two rounded or cone-shaped hollows have been drilled and scooped

out, and meet at a point, the distance of which from the centre varies.

In Mashonaland Y-shaped soapstone bowls are made with two

containers forming the fork of the Y (29). A bowl of this type was

found in Zimbabwe. The baked-clay bowls can be moulded by hand.

The narcotic is placed in the bowl and ignited with a live coal;

individuals in some tribes have small neat metal tongs for handling

the coal. The horn is half-filled with water, just past the point

where the reed enters its side. The mouth is applied to the large

open end of the horn, and the smoke is drawn through the water,

which cools it. If the open end of the horn is too large for the mouth

to go round, it is placed in such a way that half the rim rests against

282 Annals of the South African Museum.

one cheek and the other half is covered by the mouth (30). This

type of water pipe was used throughout the country, excepting by

some Bush tribes (Naron, O’Kung of Angola, Heikum and Kung of

S.W. Africa) and the Nama and Korana Hottentots, but there are

several local variations in detail. For instance, the most common

way of inserting the reed stem in the side of the horn is diagonally,

at an angle (Plate LX XXIII, figs. 1-4 and 7) (Xosa and Transkeien

tribes generally, Zulu, Basuto, Tswana, Mampukushu, and Herero), but

certain of the Cape Bushmen inserted it vertically, and some Bantu

tribes (Bomvana, Swazi, Thonga, Venda) use a much longer reed and

put its lower end right down into the horn through the large open end,

instead of inserting it in the side (Plate LX XXVIII, fig.5). In this

latter type the smoker places his hands round the top of the horn

including the reed, and makes a small opening between them through

which to inhale (31). Certain Bush tribes, the Cape Hottentots,

and the Herero dispensed with a reed altogether, and placed the bowl

directly in the horn, which the Bushmen and Hottentots covered with

‘skin at its larger end and cut off at the point to form a smaller mouth

(Plate LX X XIX, fig. 5), while the Herero inserted the bowl towards

the small end and left the big end open. Sometimes the large

opening of the horn is filled in with wood or horn and a smaller one is

made in the side higher up, with the reed still inserted diagonally

(Plate LX XXVIII, fig. 4), or the small end of the horn is used (fig. 2).

In another variety of water pipe which is found mainly in the north,

the horn water-vessel is replaced by a long-necked calabash into the

side of the bowl of which the pipe-bowl is inserted, either directly or

or a hole is made in the side of the neck to form the aperture for the

mouth (Plate LXX XIX, figs. 1, 2, and 4) (Bushmen of Angola, Kung

and Heikum of 8.W. Africa (who use no water in it), Mampukushu,

and Makaranga tribes). Sometimes a wooden ring is fastened to the

bottom of the calabash to act as a stand (32). The Mashona and the

tribes near the Zambesi mouth use a plain narrow oval gourd, narrower

at the top end, where is inserted a reed which goes right down to the

bottom of the gourd and has a bowl fixed at its top. The smoker

inhales through a square hole cut in the side of the gourd (88).

Failing any of the above types, makeshift water pipes are made out

of clay by the poorer members of most communities. In some cases

a water-vessel is properly made and baked in the shape of a brick

(Plate XC, fig. 3). In one ingenious example half an actual brick has

been used, and passages bored, one right through the centre and

Natiwe Pipes and Smoking in South Africa. 283

corked at each end, in which water is placed, and two others joining

it from the top, one for the bowl and the other for the mouthpiece

(Plate XC, fig. 4). A still simpler device is used a great deal by youths

and poor people particularly in the Transkei (84), Bechuanaland (35),

and Matabeleland (86). A water pipe is made in wet sand and the

smoker lies down to smoke it. The Tswana and Transkeien method

is to damp the soil and then make a little raised mound with a de-

pression at one end to hold the narcotic. At the other end a reed

mouthpiece is inserted to connect with the bowl. A mouthful of

water is taken and the smoke is inhaled through the water, and

expelled either through the nose or with the water from the mouth.

The reed may be dispensed with and the mouth applied direct to the

aperture. The Matabele ground pipe has a water-vessel dug in the

soil. Two channels run from it, one at each side, making an acute

angle with each other. Where they reach the surface a pipe-bowl is

inserted in one and a reed in the other. The Heichware Bushmen

carry a pipe-stick with them and make a small hollow in the ground

for the narcotic with a small tunnel leading from it into which they

insert the stick and smoke through that (87).

Le Vaillant (38) describes a dagga pipe of the Gonaqua in the

Hastern Province... “... they employ the bamboo reed and

baked earth or soft stone which they cut and scoop out to a con-

siderable depth without breaking it. They generally make them

very large—lI have seen some the shanks of which had an orifice more

than an inch in diameter.’ The description sounds like two parts of

a water pipe but that he adds that “they prefer these to the Huropean

article because the latter is not big enough.”

The water pipe was originally designed for smoking dagga, but

when tobacco first came to be smoked (as well as being used for snuff)

the water pipe was used for it in most parts, notably in the Trans-

kei, Natal, Basutoland, and Vendaland, until the European style of

tobacco-pipe was copied. In most parts of the country dagga smoking

is generally a social activity at which two or more people share a pipe

and pass it from one to another. Amongst Bushmen and Hottentots

both men and women smoke it (39), but among the Bantu generally

men only. An amusing description of the effects of dagga is given by

Schulz (40) . . . [Each takes] “two or three mighty draws between

which the smoker exhales the smoke each time, until he takes the last

draw which he retains in the lung while he passes the pipe to his

neighbour . . . [it] seems to produce intense transient excitement

of the brain, causing the smoker to burst out into vaunting speech,

284 Annals of the South African Museum.

rapidly uttered and interspersed with long-drawn yells of a most

blood-curdling nature. . . . This period over, the native grasps a

thin long reed, through which he ejects the fluid still contained in his

mouth, now of a dull blue colour from contact with the smoke, in a

succession of little pearl-ike bubbles, which he turns into circular

shapes on the floor. All this time the pipe is circulating from one to

another of the company, and it is no unusual occurrence for a smoker,

fired into almost madness by the fumes, to spring up, shouting like

a demon, and go through the evolutions of a war dance, accompanied

by the loud acclamations of his companions who in their turn perform

some equally excitable freak. From this pipe they gather inspiration

in extreme moments, the warrior before battle, the hunter before going

after dangerous game, and the witch-doctor for the execution of the

intricacies of his profession.” A similar idea was found among the

Bergdama—the dagga pipe was passed round among the company

before a collective hunt, as it was believed to give the body greater

power for the struggle; ordinarily only the older men were allowed

to smoke it because of its intoxicating qualities (41).

The Cape Hottentots were very great dagga smokers. According

to Kolb it was a frequent accompaniment to some of their ceremonies

(42). Otherwise among Bushmen and Hottentots and many Bantu

dagga is smoked purely for its own sake, or for conviviality, but among

the South-Eastern Bantu it is frequently accompanied by a spitting

game which has its local variations but of which the Zulu form is

a good example. On dark evenings particularly, the pipe is passed

from one to another, and the strong flow of saliva caused by the smoke

forms the chief requisite for this very favourite game that is played .

between two people or two sides. To the light of torches held by small

boys, each in turn after a good draw at the pipe, inhaled well

into the lungs and held as long as possible, spits the saliva on to the

floor or ground through a reed or wooden tube (Izintshumo) about nine

inches long, which is kept specially for the purpose (Plate LX X XIX,

figs. 6 and 7). Each starting from a given position spits the saliva

out in a string of closely-joimed bubbles, which if properly done should

stand without bursting for four or five minutes. The bubbles might

represent cattle and a cattle kraal, or an army, and the object of the

game is to surround the opponent’s bubbles with a closed ring so that

he cannot get out. If the ring breaks, the opponent has a chance of

breaking through and trying to surround in his turn (48).

The smoking of dagga is by no means confined to the tribal areas,

but is also practised by native and coloured in the towns, where it is

Natwe Pipes and Smoking in South Africa. 285

the object of vigorous police campaigns. ‘Two specimens that were

taken in a police raid in a slum area of Cape Town are figured on

Plate XC (figs. 1 and 2). One is made of a pickle jar and the other of

a coconut shell. Another specimen that featured in a court case in

East London and is now in Hast London Museum has been described

tome by Miss Courtenay Latimer. The water-vessel consists of a round

enamel basin about sixteen inches in diameter, built up on top with

clay into a cone, in the centre of which is a large depression for the

dagga. The reed stem is inserted in the side of the cone. It appears

that the owner of this magnificent specimen made a good thing out of

it by charging threepence a smoke to a large clientele.

The simplest type of pipe found in South Africa and possibly the

earliest tobacco-pipe (though dagga is sometimes smoked in them too)

is a short piece of the shin-bone of some animal. In one end of it the

tobacco is stuffed and in the other end dry grass to keep the ashes from

coming into the mouth (Plate LXXXYV, figs. 1-4). One specimen

(fig. 3) has a round disc of ostrich egg-shell, bored in the centre, to

serve the same purpose. Occasionally they are ornamented with

incised lines. These pipes are still common, particularly among the

Bushmen, Hottentots, Bergdama, and Herero of South West Airica,

all of whom use them for either tobacco or dagga. The Tswana

used them in the early days and in isolated districts do so still, as

anyone might do who had no better pipe. The grass in the end

serves a double purpose—not only does it prevent the tobacco from

coming into the mouth, but also it absorbs the nicotine and is

carefully removed and kept to be smoked itself in time of

need.

These bone pipes were copied in stone; the shape, however, was

modified to that of a European cigar- or cigarette-holder. A number

of different soft stones have been used for pipes and pipe-bowls, but

the majority are some kind of serpentine. They may be cut and

hollowed out with an iron spear-point, a knife, or a small stone or

iron drill, after they have been roughly shaped by rubbing on a

stone. A knife or an iron point is used to make the simple

line ornamentation (44). Pipes of this sort were and are still used

by Bushmen, Hottentots, Herero and Tswana. Among the Korana

Hottentots there is a distinction between men’s and women’s pipes,

women using the cigar shape (Plate LXXXV, fig. 6), and men the

clay-pipe shape (Plate LX XXV, figs. 10 or 15) described below (45).

To this type of pipe belong the joints of bamboo and tubes of wood

described by Stow (46) as being used by the Cape Bushmen, and

286 Annals of the South African Museum.

the latest variety is empty cartridge-cases, which are used alone

or mounted in bone (Plate LXXXYV, fig. 5). Like the water pipe,

this type of pipe may be smoked in private or passed round in

company, with a finger put over the mouthpiece to prevent the

escape of the precious smoke (47).

Wooden pipes of an indigenous pattern do not appear to have been

made in the South and Kast at all, but only in the North-west, among

the Mampukushu and some of the Ovambo tribes, whose connection

with the peoples north of the Zambesi is still comparatively recent.

A pipe from Ovamboland figured on Plate XCIX, fig. 3, is identical

with those used by the southern Angolan tribes and probably came

from there. They consist of a carved wooden bowl and a long iron

or wooden stem. The tobacco-pipes of the Mampukushu are quite

outstanding (Plate XCIX, fig. 5), and they attach great importance

and value to them (48). The stem of the pipe is carved out of

wood, about an inch in diameter and of varying length (quoted

from one metre to four feet, but our specimens are very much

shorter). The mouth of the stem is flanged, but there is no record

of a separable mouthpiece, and the fore-end is carved to represent

a man’s head, conventionalised, and looking rather like the figure-

head of a boat, but that it faces aft instead of fore. In a simpler

and less ornamental type the stem is made of a length of bamboo

about an inch in diameter, with the fore-end closed. The bowl,

which is generally of clay, is inserted a little distance from the fore-

end. In the two specimens of the wooden variety in the South

African Museum the bowl is of iron. Green says “the stem varies

in length, some being at least four feet, and usually composed of

a portion of a gun-barrel, with an additional iron tube of their own

manufacture.” (49).

The type of European tobacco-pipe that was copied in the first

instance was the old-fashioned clay pipe, small and with a little

spur at the base of the bowl (Plate LXX XVI, figs. 9 and 10). It was

copied all over the country almost exactly in stone, and in a modified

form and generally somewhat larger in wood. Its influence can still be

seen to-day, as many of the modern pipes still have the spur. The

influence of later styles of European pipes is to be seen too, both in

wood and stone, though the mouthpiece is seldom copied, as the

natives prefer a widely flanging end into which they insert a separable

mouthpiece of a short length of reed, wood, or bone. With this the

owner does not willingly part, because, coming as it does in direct

contact with his mouth, it is thought to contain something of his

Native Pipes and Smoking in South Africa. 287

spiritual essence, and would therefore be dangerous if it fell into the

hands of an enemy.

The greatest number and variety of wooden pipes are found in the

south-east, particularly in the Transkei, where their carving and

decoration have provideda means of artisticexpression that is foundin

few other things (Plates XCI to XCVIII). The pipes are well made,

Text-rig. 1.—S.A.M., 5513. Pondomisi wooden pipe, covered with dark

blue, light blue, and white beads. Has spare mouthpiece.

and hard woods that are capable of a high polish are generally chosen.

Not only is the wood carved—originally with an assegai blade as tool—

out of one piece into a variety of shapes, but the pipes are further

ornamented by inlaying the bowl with lead, brass, copper, and even

bone. The method of inlaying lead is to carve the desired pattern

into the surface of the bowl, place it in the sand, and pour molten lead

into the grooves thus made (50). The overlapping lead is cut away

and smoothed down with a spear blade, and the bowl is polished well

by greasing and rubbing on the hand and blanket. (In the Transvaal

Museum there is a small piece of lambskin, well greased and rolled up,

with which the former owner polished his pipe from time to time.)

Brass, copper, and bone inlays are generally cut to shape first and then

tacked into position on the bowl. Among the Pondomisi beadwork

is a favourite decoration (text-fig. 1). The inside of the bowl is lined

with tin to protect the wood. Nowadays pipes are usually made by

men specialists, and a great number are imported by traders.

Soapstone pipes of this type also show a great variety in styles of

288 Annals of the South African Museum.

carving, particularly in Griqualand West and the Transkei, and even

they are frequently inlaid, though generally not as elaborately, plain

bands being the most common (Plate LX XXVI, figs. 3 to 7).

Both wood and soapstone pipes frequently have fitted on to the

stem a carved end of another substance, horn, bone, or lead, into which

the mouthpiece is inserted (Plate XCVIII, figs. 5, a and 6, a).

In Basutoland there was common until quite recently a true corn-

cob pipe, for which a portion of a dry mealie cob was rubbed down and

hollowed out for the bowl, and a thin reed stuck in the side near the

Text-Fic. 2.—T.M., 35: 123. Basuto corn-cob pipe.

bottom for the stem. It would be interesting to know how the

the inspiration came (text-fig. 2).

Another type of tobacco-pipe which really belongs north of the

Zambesi, but overlaps slightly, is one with a large black earthenware

bowl, to which is fitted a very long bamboo stem. South of the

Zambesi it is found among the Batonga in the north-west of Southern

Rhodesia (Plate XCIX, figs. 1 and 2).

The Mampukushu have a pipe which consists of an earthen bowl

which may be smoked with or without a bamboo stem (Plate XCIX,.

fig. 4), and it is reported that the Matabele and those of the Bathonga

just in the vicinity of Lourengo Marques used a similar pipe, which

among the latter was smoked only by old women and dandified

young men.

The Ovambo use a fairly elaborate type of black earthenware

pipe-bowl similar to those figured on Plate XCIX, figs. 6 and 7.

They are fitted with a wooden or iron stem, twelve to fourteen

inches long. The examples figured are attributed to the Herero,

and though I can find no reference to them in the literature, the

Herero are known to have obtained pipes from the Ovambo.

All adults and even small children smoke tobacco when they can get

it. The Bushmen prefer it to any other narcotic. The Hottentots,

particularly the Nama, having once been introduced to tobacco,

reached a stage where it was ‘almost a necessity tothem.” A meeting

Nate Pipes and Smoking in South Africa. 289

without a pipe is quite unthinkable, and failing tobacco or dagga, they

will smoke the dried excrement of small animals, or for preference, in

the early days, of an elephant (51).

Both water pipes and the ordinary type seem to have existed side

by side in most parts of the country, excepting among some Bush

and Hottentot tribes as previously mentioned, but nowadays the

water pipe for dagga is disappearing, and ordinary European-made

wooden tobacco-pipes are common throughout the country, and will

probably soon have taken the place of all native-made articles.

ABBREVIATIONS IN DESCRIPTION OF PLATES.

CO: . Camp Collection.

D-C.C. . Duggan-Cronin Collection.

KeIViea. . Kimberley Museum.

SREIMGS %: . Transvaal Museum.

SC ee . Schofield Collection.

S.A.M. . South African Museum.

S.A.M.M.L.C. Malleson Loan Collection, at the S.A. Museum.

DESCRIPTION OF PLATES.

Pirate LXXXYV.

Fig. 1.—S.A.M., 5278. Bushman, N.E. Kalahari. Section of shin-bone of

small animal. Decorated at centre with incised lines rubbed with dark brown

substance. Small end stuffed with grass, which acts as a filter.

Fig. 2.—C.C. Nama Hottentot, Namaqualand. Section of shin-bone of small

animal. Undecorated.

Fig. 3.—C.C. Nama Hottentot, Namaqualand. Section of shin-bone of small

animal. Large end decorated with incised lines rubbed with soot. At small

end is fitted an ostrich egg-shell disc, bored through the centre, to prevent bits

of leaf from coming into the mouth.

Fig. 3, a.—View of end of same pipe showing bored disc.

Fig. 4.—S.A.M., 5060. Bergdama, Zesfontein, S.W. Africa. Section of shin-

bone of small animal, covered, except for a band near the top, with a section of

a small stem of Cotyledon paniculata (Botterboom).

Fig. 5.—S.A.M., 3169. ? Bushman, no history. A brass cartridge-case, into

the narrow end of which a small and short section of bone has been inserted to

serve as a mouthpiece.

Fig. 6.—C.C. Hottentot, Griquatown, C.P. Cylindrical pipe of polished greeny

fawn serpentine with fawn flecks. The mouthpiece is of lead, fused on. The

bowl is octagonal in section.

290 Annals of the South African Museum.

Fig. 7—C.C. Nama Hottentot, Warmbad, 8.W. Africa. Cylindrical pipe of

polished brownish-green serpentine with fawn flecks. Simple incised lines round

mouth. Butt-end is small and has small opening as shown alongside.

Fig. 8.—C.C. Hottentot, from Prieska, C.P. Polished grey serpentine bowl

with patches of dark brown, green, and black. Has a short lead stem fused on,

and vestige of spur at base of bowl.

Fig. 9.—S.A.M., 5808. ? Bushman. Found in mountains of Moordenaar’s

Karoo, north of Laingsburg, C.P. Small bowl of polished light green serpentine.

Mouth decorated with rough incised cross-hatching. Round the base four small

holes to give a grip to the lead which is poured through the base of the bowl to

form a short stem, and visible half-way up the bowl inside.

Fig. 10.—S.A.M., 5444. Hottentot, Griqualand West. Pipe of polished dark

green serpentine with darker streaks. Very short stem and spur at base of bowl.

Undecorated.

Fig. 11.—C.C. Hottentot, Warmbad, 8.W. Africa. Cylindrical pipe of

polished fawn-coloured serpentine. Two bands of incised cross-hatching between

horizontal lines round mouth.

Fig. 12.—C.C. Hottentot, Griquatown, C.P. Cylindrical pipe of khaki-coloured

unpolished clayey stone, with considerable amount of incised decoration.

Fig. 13.—S8.A.M. No history. Cylindrical pipe of polished brown serpentine

with light flecks, darkened with smoke. Slight cross-hatched incised decoration

round mouth. Mouthpiece of lead, inlaid in horizontal and vertical bands for

strength at intersection.

Fig. 14.—C.C. Nama Hottentot, Warmbad, S.W. Africa. Polished olive-

green serpentine with red streak. Style of European clay pipe with spur at base

of bowl. Band of roughly incised diagonal lines round mouth of bowl.

Fig. 15.—C.C. Hottentot, Griquatown. Pipe of polished yellowish-brown

serpentine with fine lighter flecks. Style of European clay pipe. Band of incised

diagonal lines round mouth of bowl.

Fig. 16.—C.C. Nama Hottentot, Warmbad. Pipe in style of European clay

pipe with no spur at base of bowl. Polished greeny-brown serpentine. Decorated

with four bands round top of bowl, one plain, three incised with diagonal lines

in alternate directions.

Fig. 17.—C.C. Hottentot, O’Okiep, Little Namaqualand. Pipe of olive-green

polished serpentine with black flecks and red patches. The base of the bowl is

carved as a conventional jackal face. Different from clay-pipe style.

Fig. 18.—C.C. Hottentot, Namaqualand. Bowl only of khaki-coloured

polished soapstone. The bore is plugged at the back with a slightly carved wooden

stopper. Copied from a different style of European pipe-bowl.

Fig. 19—C.C. Hottentot, Warmbad, 8.W. Africa. Greyish fawn serpentine.

Made with three bowls, each connected with the others by the bore which passes

right through the stem and the base of the bowls. The hole is plugged with wood

where the arrow indicates. The two outer bowls have rough incised cross-hatching

round the mouth. I have seen specimens with as many as seven bowls in two

rows. The idea of two bowls to one pipe may have been of native origin, but

Tam quite sure that a greater number than that has only been conceived under

European influence and for the tourist trade. None of the pipes of this type that

I have seen bear any sign of having been smoked.

Fig. 20.—C.C. Hottentot, Brakkies, C.P. Pipe with bowl and short stem

Native Pipes and Smoking in South Africa. 291

of polished greeny-brown banded shale, with ridge projecting forward and ter-

minating at peg at base of bowl. At the end of the stone stem is a brass pipe

ring, and in it is inserted an unpolished wooden stem or long mouthpiece.

Pratt LXX XVI.

Fig. 1.—S.A.M., 1790. Tswana, Bechuanaland Protectorate. Pipe of polished

greenish-brown banded serpentine. Bowl and stem made in one with a band of

tin at the top of the stem. The base of the bowl and top of the stem are inlaid

with lead. There is a narrow brass band at the end of the stem.

Fig. 2.—C.C. Batlaro, Kuruman, C.P. Polished yellowish-brown banded

serpentine, with red flecks. Large projection at base of bowl. Two bands of

fine incised cross-hatching round mouth.

Fig. 3.—C.C. Hottentot, Griquatown, C.P. Chrome-yellow highly polished

serpentine, with quartz flecks. Over the end of the stone stem is fixed a zinc

band into which a wooden stem is inserted. The zigzag decoration round the

mouth and wing is incised.

Fig. 4.—C.C. Hottentot, Namaqualand. Polished golden brown serpentine

with dark and quartz flecks. The two bowls are connected only at the base,

below which is a carved projection. Into a brass band round the end of the short

stone stem a longer wooden stem is inserted. The bore passes right through the

base of the two bowls, and is plugged at the front of the foremost bowl with a

small grooved bone peg.

Fig. 4, a.—Underside of bowls of above, to show carving.

Fig. 5.—C.C. Hottentot, O’Okiep, Little Namaqualand. Polished light

brownish-yellow serpentine, with spur at base, and very short stem joined by a

flanging brass pipe-ring to long carved black horn stem. The mouth of the bowl

is decorated with two bands of incised diagonal lines, and is fitted with a brass

pipe-cover of European pattern and, probably, manufacture.

Fig. 6.—C.C. Griqua, Griquatown, C.P. Elaborately carved and polished

chrome-yellow clayey stone. A square brass fitting covers the junction of the

brown wooden stem and the short vertical stone stem. The bore is plugged in

three places as shown by arrows, in two flush with the surface and the third with

a large wooden peg. At one place the bore has broken through the stone, and a

wooden lining has been put in. Round the mouth of the bowl is a band of open

cross-hatching.

Fig. 7.—C.C. Griqua, Griquatown, C.P. Polished chrome-coloured clayey

stone, very fine grained. Two bowls joined by a short stem, both decorated with

roughly incised lines round mouth; one has a rim. There is a brass band at the

end of the stone stem, where a rough wooden stem is inserted.

Fig. 8.—C.C. Griqua, Griquatown, C.P. Bowl carved out of banded shale,

reddish chocolate in colour, with bands of chrome. <A band round the mouth

and another round the middle are inlaid with lead. The bowl is set on a base,

one end of which is inserted in a carved and elbowed wooden stem, the insertion

end of which is inlaid with lead. The bore is plugged in three places, one flush

with the surface and two fitted with carved horn pegs. A light wooden mouth-

piece is fitted to the flanging end of the stem.

Fig. 9.—S.A.M., 1420. White clay pipe of European manufacture. Bowl cast

to represent the head of a bearded man. The pipe was found in Cape Town,

VOL. XXIV, PART 5. 20

292 Annals of the South African Museum.

22 feet below surface, during excavations for the present railway station, at a

spot where the step of an old landing-stage and some post office stones were also

found. (These stones were used prior to the first settlement at the Cape in 1652.)

Compare figs. 8, 9, and 10, Plate LXXXV.

Fig. 10.—S.A.M., 932. White clay pipe of European manufacture of the time

of Charles II. Note the angle of the bowl and the spur at its base. Compare

with stone pipes, Plate LX X XV, figs. 14 and 15, and Plate LX X XVI, figs. 1 and 2.

Pirate LXXXVIL.

Bows FOR WATER Pires, ANCIENT AND MODERN.

(Unless otherwise stated, all are round in cross-section.)

Fig. 1.—S.A.M., 520. Old bowl, excavated East London district. Dark grey

baked clay. Very small capacity.

Fig. 2.—8.C. Old bowl, excavated at Beacon site near Durban. Carved out

of black stone. Very small capacity.

Fig. 3.—S.A.M. Old bowl, excavated at Durban in 1882. Reddish-brown

baked clay. Rather larger capacity.

Fig. 4.—S8.C. Old bowl, excavated at University site, Durban. Yellowish clayey

stone, oval in cross-section.

Fig. 5.—S.A.M., 5857. Old bowl, excavated at the Beacon site, near Durban.

Red baked clay. Very small capacity.

Fig. 6.—S8.C. Old bowl, excavated at Tinley Manor site. Rough red baked

clay.

Fig. 7—S.A.M. No history. Black baked clay, slightly burnished on outside.

Decorated in triangle design of deep incisions.

Fig. 8.—S.A.M. No history. Black baked clay, slightly burnished outside.

Undecorated.

Fig. 9.—S.A.M., 5869. Found in grave in Umhloti Dune, Durban. A stone

ink-bottle with a hole knocked in the bottom and used upside down. (A similar ~

bowl was reported as seen in use in Constantia, Cape, within recent years.)

Fig. 10.—C.C. Modern bowl from Transkei. Carved out of grey-brown soap-

stone and polished. Bowl decorated back and front with two bands of diagonal

V-shaped grooves.

Fig. 11.—C.C. Modern bowl from Transkei. Carved out of dark grey stone,

unpolished. Bowl decorated with two horizontal bands of short diagonal V-shaped

grooves.

Fig. 12.—S.A.M., 4769. Modern bowl from Ovamboland, 8.W. Africa. Black

baked clay slightly burnished.

Fig. 13.—S.A.M., 796. Found at Smithfield, O.F.S. Greyish stone, very heavy

in weight. Quite undecorated. Very thick walls. (Compare Plate LX XXIX,

fig. 3.) :

Fig. 14.—S.A.M., 3336. Modern bowl from the Herero, 8.W. Africa. Polished

greeny-grey serpentine with heavily carved base, carving of horizontal and vertical

V-shaped grooves and flat ridges.

Fig. 15.—S.A.M., 5859. Modern Indian pipe-bowl from Indian Market, Durban.

Bright red lightly baked clay. Very fragile.

Native Pipes and Smoking in South Africa. 293

Fig. 16.—S.A.M., 3000. No history. Carved out of light pinky-grey stone.

Undecorated except for raised rings at the centre. (Compare with fig. 15.)

Fig. 17.—S.A.M., 2544. No history. Made of almost white sandstone.

Fig. 18.—C.C. Modern bowl from Transkei. Smooth yellowish-brown serpen-

tine. Plain bowl, base decorated with diagonal V-grooves, and flat ridges.

Fig. 19.—S.A.M., 5888. ? Bushman. Dug up at Graaff Reinet, C.P. Carved

out of chalky stone, dull fawn in colour. Decorated with vertical bands of short

diagonal grooves (semicircular in cross-section), with a vertical band of raised

dots, on two opposite sides.

Fig. 20.—S.A.M., 3333. Zulu, Utrecht, Natal. Polished green soapstone bowl

on long cylindrical base with ring attached by four spokes just below base of bowl.

Prate LX XXVIII.

WATER PIPES, FOR SMOKING DaaGa.

Fig. 1—C.C. Transkei. A light-coloured polished ox-horn water-vessel,

with large end left open for mouth and bound with light-coloured hide, which is

secured at the lower edge by four wooden pegs. A piece of strong reed, strengthened

in the centre with bindings of sinew, is inserted diagonally in the side of the horn

and held in place with thick black wax. On top of the stem is placed a blackened

and polished stone bowl, plain but for a ridge with a diamond cut-out pattern round

the neck. The bowl is round in cross-section.

Fig. 2.—C.C. Hottentot, Namaqualand. Black polished ox-horn water-vessel,

the large end of which is filled in with a round of wood, held in with wooden pegs.

Pointed end cut to nozzle shape, with small opening for the mouth. (This shape

and the ring at the base make it appear that this is a converted snuff-box or

powder-horn.) Just below the nozzle is a small hole. The plain reed stem is

inserted diagonally near the mouth, and held in place with a binding of cloth over

which some waxy substance has been smeared. A similar binding is found at the

top of the stem where it enters the base of the carved and polished red serpentine

bowl, decorated with four vertical bands of diagonal V-grooves and flat ridges,

separated by two V-grooves and a sharp ridge. The bowl is oval in section.

Fig. 3.—S.A.M., 5678. Mampukushu, Okavango River, S.W. Africa. Black

unpolished kudu horn water-vessel, with large end left open for mouth. Wooden

stem shouldered at each end. The lower end is fitted almost at right angles into

a small hole in the horn near the point, with a strip of bark put in to make the

fit tighter. The other end is inserted in the base of a wooden bowl, round in cross-

section, branded black on the outside and unpolished.

Fig. 4.—S.A.M., 2879. Said to be Basuto. Black polished horn water-vessel,

ornamented with round brass furniture tacks. The large end is filled in with wood,

held in by the tacks. The point is left intact, and an oval opening is made in

the side near the point for the mouth. A short length of smooth reed is inserted

diagonally in the side of the horn, and at its other end into a pinkish-mauve soap-

stone bowl, round in cross-section, and elaborately carved on bowl and neck with

vertical and horizontal bands of V-shaped grooves and sharp ridges.

Fig. 5.—S.A.M.M.L.C., 17. Swazi, Eastern Transvaal. Long, dark grey, un-

polished eland horn water-vessel, with large end open for mouth. Near the top

there are two bands and a loop of raw hide, made by slitting two ends of a straight

294 Annals of the South African Museum.

piece of hide, and slipping them both over the end of the horn, leaving the centre

piece to form the loop. A small black clay bowl, round in cross-section, with a

raised ring round the neck and a hide strap put on in the same way as that on

the horn, is placed on the top of a long stout reed, which when the pipe is in use

is placed right down into the horn through the mouth. When not in use the reed

can be put through the loops on bowl and horn for ease in carrying and storing.

Fig. 6.—C.C. No history. Probably Transkei. Light-coloured polished ox-

horn, with large end openfor mouth. A short, thick, polished, light wooden stem

is inserted at right angles to the horn just near the mouth, and at the point of

insertion the horn is bound with a wide band of hide, which is drawn on while wet

and allowed to set, therefore showing no join. On top of the stem is a heavy

sandstone bowl, covered with thin skin, drawn over while wet. Before drawing

on the skin the bowl has been bound with sacking. The bowl is round in section.

Fig. 7.—C.C. Xosa, Komgha, Eastern Province. Large, light-coloured, highly

polished ox-horn water-vessel, with large end left open for mouth. Polished

wooden stem inserted diagonally in the side of the horn about centre, and fixed

in with thick plaster of black wax. On top of the stem is a heavy bowl of polished

red jasper, with raised knobs on it. The bowl is square in section, with rounded

corners.

PLuatEe LXXXIX.

WatTER PIPES FOR SMOKING DaGGa.

Fig. 1—8.A.M., 5677. Mampukushu, Okavango River, S.W. Africa. Long-

necked gourd water-vessel, in which the neck of the gourd serves as the stem

and the bowl as the actual water container. A wooden bowl is set in the side of

the bowl of the gourd and secured with a plaiting of fibre cords over bark.

Fig. 2—C.C. ? Matebele, from Bulawayo, 8S. Rhodesia. Light brown polished

narrow gourd water-vessel, with opening for mouth at narrowest end, and short

reed stem, bound at base to secure fit, and stuck at right angles into fattest part _

of gourd. On top of stem is a dark soapstone bowl, the upper part of which is

carved in long V-grooves and ridges. The bowl is secured to the gourd with

a short piece of bark fibre cord.

Fig. 3—K.M. 1900. Bushman, Blauwater, Graaff Reinet, C.P. Small black

horn water-vessel with large end filled in with wood or horn, and the point cut

off to form a mouth. The wooden stem is inserted in the side of the horn diagon-

ally, and at its other end into the base of a plain greyish sandstone bowl.

Fig. 4.—K.M. 2288. Bushman, Ovamboland. The water-vessel is a calabash

with a long curved neck, in a forward bend of which a hole has been made for

the mouth. The pipe-bowl is set in the side of the bowl of the calabash, and is

made of wood.

Fig. 5.—T.M. 6635. Bushman. Short black horn water-vessel open at its

smaller end for mouth, and closed at the larger end with a cap of hide put on

wet and bound with string until dry, when it had set quite fast. Near this end

is inserted at right angles with the horn a small black clay bowl, decorated with

vertical chevron grooves.

Fig. 6—D-C.C. Zulu. Spitting-tube. The narrower stick on the right hand

Natwe Pipes and Smoking in South Africa. 295

of the sketch is bored down its length for the tube. The other is solid and is

merely a handle. The two are bound together with fibre string.

Fig. 7.—D-C.C. Zulu. Same spitting-tube as fig. 6 rolled up in the little

grass mat in which spitting-tubes are kept when not in use.

PLatTE XC.

DacGa PIPES FROM Towns.

Fig. 1—S.A.M., 4276. District Six, Cape Town. A pickle jar is used as a

water-vessel. The bowl is of wood, and the wooden stem of the bowl and the

reed stem through which the smoke is inhaled are bound together with rags,

which form the large stopper for the jar. Both stems are visible within the jar.

Fig. 2.—S.A.M., 4274. District Six, Cape Town. A coconut shell is used as

a water-vessel, and in the top of it are inserted at different points the reed smoking

stem and the stem of the wooden bowl. There is a small hole between these

two insertions. Both insertion points are built up with wax. (Figs. 1 and 2 were

captured in a police raid.)

Fig. 3.—S.A.M., 2426. No history. A water-vessel of baked red clay, roughly

shaped like a brick with a depression in one end for the leaves of the narcotic,

and a passage leading from it to the other end, where the top of the neck of a glass

bottle serves as a mouthpiece.

Fig. 4.—K.M. 1378. From Kampfersdam, near Pretoria. Half of a hard

whitish-grey brick has a channel bored right through the centre horizontally, and

plugged at each end with cork. From the top of the brick are bored two other

channels to meet the horizontal one. One is enlarged at the top to serve as a

bowl for the dagga, and in the other is inserted a copper fitting for a mouthpiece.

PuateE XCI.

PLAIN WoopeEn Tosacco-PIpPEs.

These pipes are for the most part copied from a later and larger style of European

pipe, though the majority retain the spur at the base of the bowl like the old clay

pipes. In every case the bowl of the pipe is lined with tin. In many cases these

tin linings give the appearance of having been European-inspired if not European-

made, particularly in the vandyke cut-out along the lower edge to make the lining

fit the narrowing of the bowl. Possibly they are obtainable in the stores. In

the majority of cases the bore of the stem emerges on the surface of the bowl

for ease in making, and is plugged with a small round of wood, often so well

matched as to be hardly distinguishable. In every case on this and succeeding

plates the plug is indicated with an arrow.

Fig. 1.—C.C. Transkei. Dark brown wood, with fair polish, very much after

European style. Wooden mouthpiece.

Fig. 2.—C.C. Transkei. Dark brown highly polished wood, with carved horn

end to stem. (The wooden stem is cut down to a thin column over which the

horn piece fits so well that when the wood has a high polish it is frequently hard

to see the join.) The mouthpiece is a short length of the bone of a bird.

296 Annals of the South African Museum.

Fig. 3.—C.C. Xosa, Transkei. Dark unpolished wood. Only the top portion

of the bowl is lined with tin. The horn end to the stem is inlaid with a band of

bone in the centre. Wooden mouthpiece.

Fig. 4.—C.C. Transkei. Plain brown wood, with fair polish. Wooden mouth-

piece.

Fig. 4, a.—Front view of bowl to show tapering.

Fig. 5.—C.C. Transkei. Dark red wood of the karreeboom (a species of Rhus),

with fair polish. Bone mouthpiece.

Fig. 6.—C.C. Transkei. Woman’s pipe. Light-coloured highly polished

wood. The whole of the base of the bowl is inlaid with lead. Wooden mouthpiece.

Fig. 7.—C.C. Transkei. Polished brown wood; bowl and stem all in same

plane. Wooden mouthpiece.

Fig. 8.—C.C. Griquatown (Transkeien type). Dark brown prettily grained

wood, with two bowls set across end of stem. Horn end to stem, similar to fig. 2,

and band of thin brass round fore-end of stem. Wooden mouthpiece.

Fig. 9—C.C. Transkei. Dull brown coarse-grained wood. Flanged end of

stem finished with lead, put on like the horn. Wooden mouthpiece.

Fig. 10.—T.M. Transkei. The bowl is the oil-cap of some machine, inscribed

Justrite Mfg. Co., Chicago, U.S. Originally black enamel paint on iron, but the

paint has worn off. With the mouthpiece of an ordinary European pipe, it makes

a small but efficient smoking outfit. |

Puate XCII.

PxLaIn Topacco-PIPpEs, ALL FROM THE TRANSKEI.

Fig. 1—C.C. Dark brown wood, highly polished. Wooden mouthpiece.

Fig. 2.—C.C. Light brown wood, with fair polish. Bore passes right through

base of all three bowls. Short reed mouthpiece. Plain but for band of lead

inlaid at junction of stem and last bowl, probably for strength.

Fig. 3.—C.C. Plain light brown wood with dark grain. Wooden mouthpiece.

Fig. 4.—C.C. Plain brown highly polished wood, with wooden mouthpiece.

Tin lining of bowl incised where it turns over rim of mouth.

Fig. 5.—C.C. Xosa. Hard brown wood, well polished. Carved bone end to

stem, and wooden mouthpiece secured to stem by a thin thong.

Fig. 6.—C.C. Light coarse-grained unpolished wood, with patch of dark heart-

wood on one side of bowl. Bowl has double lining. Wooden mouthpiece.

Fig. 7.—C.C. Rich brown coarse-grained wood, with fair polish. Mouthpiece

is section of bird-bone.

Fig. 8.—K.M., 1924. Tembu. Plain pipe, carved out of one piece of bone.

PLate XCIIL.

CARVED WoopDEN TosBacco-PIPEs.

Fig. 1—C.C. Xosa, Transkei. Dark brown wood, with carving of fretwork

type on stem. End of stem is carved dark coloured horn. Wooden mouthpiece.

There is a lead inlay to mend part of the carving where it has cracked.

Native Pipes and Smoking in South Africa. 297

Fig. 2.—C.C. Transkei. Brown wood, with good polish. Stem decorated

with fretwork type of carving. A band of lead is inlaid round the end of the

stem. Wooden mouthpiece.

Fig. 3.—C.C. Tambuki, Cala District, Transkei. Dark brown, highly polished,

coarse-grained wood. The tin cover off a European pipe is fixed on with a brass

hinge, and a cylinder of tin is inserted through the base of the bowl. The bore

is plugged at the back with a wooden peg. There is a wooden mouthpiece.

Fig. 4.—C.C. Griquatown. Dark red highly polished wood, with spiral ridge

carved on stem (? to represent snake). Light wooden mouthpiece.

Fig. 5.—C.C. Transkei. Dark brown wood, unpolished. Bowl has double

lining. Wooden mouthpiece.

Fig. 6.—C.C. Xosa, Transkei. Brown wood. Bowl rests on wheel in same

plane as stem. Bowl lined with brass. Mouthpiece is section of a bird-bone.

Fig. 7.—C.C. Xosa. Well-polished stinkwood. Horn end to stem and horn

mouthpiece.

Fig. 7, a.—Front view of bowl of same pipe.

Puate XCIV.

WoopEN ToBacco-PIPES WITH REPRESENTATIONAL CARVING.

Fig. 1.—C.C. Fingo, Transkei. Dark brown unpolished wood. Bowl supported

by small female figure. Wooden mouthpiece.

Fig. 2.—K.M., 1201. Transkei. Made of light-coloured wood to represent an

ox, with spots, hoofs, ears, and horns, and end of stem branded black. The bowl

is in the shoulders of the animal.

Fig. 3.—C.C. Grahamstown. Light brown, very highly polished wood, with

the bowl carved in the shape of a baboon’s head. The eyes are brass, inset, and

there is a band of thin brass round the middle of the stem, fastened by lapping the

edges over each other. The wooden mouthpiece is fairly elaborately carved.

Fig. 4.—C.C. Griquatown. Dark red karreeboom, unpolished. Bowl is carved

to represent a human figure with head projecting in front instead of on top, and

rudimentary arms on sides. The end of the stem is cut in the European style,

instead of flanging.

Fig. 5.—C.C. Transkei. Dark brown coarse-grained wood, with a fair polish.

The bowl is carved to represent a woman’s figure, with sky-blue beads inset for

eyes, and on her back a conventional representation of a child inlaid in lead,

(figured alongside.) Wooden mouthpiece.

Fig. 6.—C.C. Fingo, Butterworth, Transkei. Dark brown highly polished

wood. The bowl and its support are carved to represent the engine of a train with

four wheels. The mouthpiece is a section of a bird’s bone.

Fig. 7.—C.C. Griquatown. Reddish brown highly polished wood, with two

bowls, the first of which is carved to represent a wheel. The mouth end of the

stem is cut in the European style.

PLatE XCV.

WoopEN Topacco-PIrpES CARVED TO REPRESENT HuMAN FEATURES.

Fig. 1.—C.C. Basutoland. Light brown wood, with bowl carved to represent

the seated figure of a woman. The bowl is set in her body. Her head and neck

298 Annals of the South African Museum.

are detachable and make a stopper for the bowl. The stem of the neck has plugs

to make it fit the mouth of the bowl. She has one black and two blue beads

for earrings, and necklace and girdle of alternate black and white. Two black

beads are inset for the eyes. Wooden mouthpiece. The pipe has not been smoked.

Fig. 1, a.—Front view of bowl and figure.

Fig. 2.—C.C. King Williamstown. Light-coloured fine-grained wood, with

bowl carved to represent a girl carrying a pot on her head. The pot is the bowl

of the pipe. Her apron, waistband, and bangles are of lead inlaid in the wood.

In addition she has bent round each wrist a brass bangle of triangular wire. Her

hair is branded black, and black beads are inset for her eyes. A band of lead is

inlaid in the stem and the mouthpiece is of bird-bone.

Fig. 2, a.—Front view of bowl and figure.

Fig. 3.—C.C. Transkei. Dark brown unpolished wood. Bowl carved to re-

present woman’s head and shoulders, with white beads for the eyes. The end of

the stem is of light-coloured horn and the mouthpiece of wood.

Fig. 4.—S.A.M., 4923. Basutoland. Made of dark brown unpolished wood,

very much in the European style, but the bowl is carved to represent a man’s

bearded face, with blue beads for the eyes.

Fig. 5.—C.C. Light-coloured coarse-grained wood, with bowl carved to re-

present human head and neck. White beads inserted for the eyes. The mouth-

piece is wooden.

PuateE XCVI.

WoopEN PIPES WITH METAL ORNAMENTATION WOUND OR FASTENED ON.

Fig. 1—OC.C. Dordrecht, 0.F.S. Light brown polished wood, with three bands

of brass round the stem, fastened at the back with brass tacks. Wooden mouth-

piece.

Fig. 2.—C.C. Transkei. Dark brown wood, slightly polished. The upper

portion of the bowl is covered with a band of thin brass, fastened at the back

by lapping the edges over each other and tacking with copper tacks as well. The -

stem is bound with brass wire in spaced groups of five or six twistings. The end

of the stem is ivory, and the wooden mouthpiece is set in a small ring of tin in

the ivory end.

Fig. 3.—C.C. Transkei. Detail of ornamentation on bowl of dark wooden

pipe similar to fig. 2 in shape and size. The band round the top is brass, the stars

are copper, and the curved bands tin. The brass band is a fitting ring, the other

pieces are nailed on with tiny copper and iron tacks.

Fig. 4.—C.C. Transkei. Dark highly-polished wood, with three bands of

brass round stem, fastened at the back by lapping the edges. Round the mouth

of the bowl is a jointless band of brass, with three rows of triangles cut out to

show wood underneath. Band held on with brass tacks. Wooden mouthpicee.

Fig. 5.—C.C. Butterworth, Transkei. Dark brown highly-polished wood, with

horn end to stem and bird-bone mouthpiece. The bowl is highly ornamented

with plain bands and cut-outs, mainly of zinc, but also of copper and brass. The

bands are fastened in front of the bowl by lapping the edges over, and the small

cut-outs are nailed on with copper and iron tacks. The base of the bowl is inlaid

with plain lead.

Native Pipes and Smoking in South Africa. 299

Fig. 6.—C.C. Transkei. Highly-polished brown wood, with wooden mouth-

piece. Near the end of the stem there is a plain band of brass, fastened at the

back by lapping the edges over each other.

Fig. 7.—C.C. Xosa, Transkei. Detail of ornamentation on dark brown wooden

bowl. The bands round the top and the middle row of rings are zinc, and the top

and bottom rows of rings copper. They are fastened on with copper tacks.

Pirate XCVII.

WoopEN PIPES WITH PATTERNS INLAID ON BowLts IN LEAD.

The process of inlaying lead has been described. It is to be noted that fre-

quently plain bands or patches of lead are inlaid as a reliable means of mending

cracks in the pipe, particularly at the junction of bowl and stem; compare Plates

XCII, fig. 2, and XCIII, fig. 1.

Fig. 1.—C.C. Transkei. Red wood with dark horn end to stem and wooden

mouthpiece. Bowl inlaid with simple line and four dots of lead.

Fig. 1, a.—Front view of bowl.

Fig. 2.—C.C. Transkei. Light brown wood, highly polished. Same design

on each side of bowl inlaid with lead. Horn end to stem and wooden mouthpiece.

Fig. 3.—C.C. Transkei. Dark highly-polished wood. Long narrow bowl inlaid

round mouth with deep band of lead, with jagged lower edge. Carved horn end

to stem and bird-bone mouthpiece.

Fig. 4.—C.C. Transkei. Small dull brown wooden pipe with black horn end

to stem and wooden mouthpiece. Bowl inlaid with heavy plain band and seven

lighter diagonal vandyked bands of lead.

Fig. 5.—C.C. Transkei. Dark brown very highly-polished wood. Bowl and

its support inlaid similarly on each side with lead. Wooden mouthpiece.

Fig. 6.—C.C. Transkei. ?Woman’s pipe. Dark brown unpolished wood.

Long narrow bowl with vandyked band round mouth, and wavy lines on bowl

inlaid with lead. On each side of the bowl is a large rectangle of ivory, let into

the wood and held with lead plugs.

Fig. 7.—C.C. Barkley West. Detail of lead inlay on light brown, very highly-

polished wooden bowl.

Fig. 8.—C.C. Transkei. Detail of lead inlay on brown wooden bowl.

Puate XCVIII.

WoopEN ToBAcco-PIPES WITH ELABORATE DESIGNS INLAID WITH LEAD.

Fig. 1.—C.C. Transkeien type, from Johannesburg. Dark brown wood, with

fair polish. The bowl and its pedestal are carved to represent a railway engine

and inlaid with lead and copper, on the bowl purely for ornament and on the base

to emphasise the features of the engine. Wooden mouthpiece.

Fig. 2.—C.C. From Port Elizabeth. Dark polished wood. Bowl rests on a

carved hand. The stem is in two parts, one of which is like the wrist of the hand,

each of which is inserted into the end of a narrow brass tube, so that the tube

has the effect of being slipped over the stem. Wooden mouthpiece.

Fig. 2, a.—Front view of bowl and support.

300 Annals of the South African Museum.

Fig. 3—C.C. Transkei. Detail of ornamentation on dark wooden bowl of

woman’s pipe. The stars are inlaid with lead, and the brass band round the

mouth is a deep ring held on by exactness of fit.

Fig. 4.—C.C. Griquatown. Detail of lead inlay on light brown wooden bowl.

Fig. 5.—C.C. Transkei. Detail of lead inlay on dark brown wooden bowl.

Fig. 5, a.—End of stem of the same pipe, showing the alternate bands of lead

(white) and black horn (shaded) which are put over the wooden stem to make

a flanging mouth. Wooden mouthpiece.

Fig. 6.—C.C. Transkei. Detail of lead inlay on reddish wooden bowl that is

so highly polished as to look like mahogany.

Fig. 6, a.—Carved horn end to stem of same pipe.

Fig. 7.—C.C. Transkei. Brown polished wood. Large round bowl inlaid with

lead in naturalistic and conventional patterns. A piece of tin is nailed with brass

tacks at the base of the bowl over its junction with the stem, possibly to strengthen

a crack. The stem ends in a flanging mouth of ivory and has a wooden mouth-

piece.

Fig. 7, a.—Detail of lead inlay on bowl of same pipe.

Prats XCIX.

PrrEs FROM SoutH WeEstT ArRicA AND NORTHERN RHODESIA.

Fig. 1.—S.A.M., 2074. Batonga, N. Rhodesia. Black earthenware bowl,

widest just above its centre, where a deep horizontal groove separates the plain

polished upper part from the cross-hatched lower part. The bowl is on a short

stem, set at an angle on a small pedestal from the side of which a short tube,

all in one with the bowl and pedestal, projects and is inserted in a hole in the side

of a long reed stem. The pipe-head is secured to the stem by a binding of copper

wire round the top of the pedestal and below the hole, and by a single brass wire

which passes through two holes in the side of the bowl and round the stem. The

stem is a thick bamboo with ornamental bindings of zinc, copper, and brass wire.

At its base is inserted a pointed wooden peg bound in with copper wire, to stick.

in the ground so that the pipe may stand upright. Total length of stem with

point, 232 inches.

Fig. 2.—S.A.M., 3342. Baila, Kafue River, N. Rhodesia. Glazed black earthen-

ware pipe-head, of which the bowl rests on a ring, which is decorated with oval

depressions (? idea of head-pad for carrying pots), on the shoulders of an antelope,

the upturned body of which forms the stem of the pipe-head. The upper part

of the bowl is plain, and is separated by a deep horizontal groove from the cross-

hatched lower part. A long thin bamboo stem is inserted in the mouth of the

clay stem, and has a band of tin, secured only by fit, round the base. Length

of stem, 27 inches.

Fig. 3.—K.M., 2016. Ovambo, 8.W. Africa. Pipe-head of dark reddish-brown

hard wood. Bowl, pedestal, and short stem in one piece, with lines, zigzags, and

geometrical designs carved in relief. Should have a long stem of iron or wood

inserted in the mouth of the stem of the bowl.

Fig. 4.—S.A.M., 5676. Mampukushu, Okavango River, $.W. Africa. Simple

pipe-head of reddish unglazed clay. May be smoked with or without additional

reed stem.

Natwe Pipes and Smoking in South Africa. 301

Fig. 5.—S.A.M., 3341. Mampukushu, Andara, 8.W. Africa. Said to have

belonged to the chief Libebe, about 1860. A stem, holder for bowl, and “‘figure-

head,” all cut in one piece of light reddish wood, decorated similarly on each side

with round depressions as shown. A jointless band of skin has been put on wet

round the mouth of the stem. A thin piece of iron has been bent round to a funnel

shape with edges slightly overlapping, and put right down into the wooden bowl

so that only about half its length shows above the wood. It is packed in with

wood, and a ring of thick spirally-twisted native-made iron wire at the top of the

wooden bowl helps to hold it in place.

Fig. 6.—S.A.M., 3332. Herero, S.W. Africa. Dark grey earthenware pipe-

bowl, all in one with pedestal and short stem. The mouth of the bowl has a

flanging lip. The pedestal and stem are strengthened with hide; a jointless piece

drawn over the stem while wet, and another piece sewn and bound round the

pedestal and base of stem. The bowl is decorated with rows of punched dots

in vertical bands. A wooden or reed stem should be inserted in the mouth of the

earthenware stem.

Fig. 7.—8.A.M., 3332. Herero, S.W. Africa. Black earthenware pipe-bowl

with pedestal and short stem. The mouth of the bowl has a flanging lip. The

bowl is decorated with cross-hatching and vertical bands of punched dots. The

mouth of the stem is bound with sinew. A wooden or bamboo stem should be

inserted in the mouth of the earthenware stem.

ACKNOWLEDGMENTS.

I should like to thank Mr. R. C. Camp, of Plumstead, Cape; Dr.

Beukes of the Transvaal Museum; Mr. Duggan-Cronin, of Kimberley;

Mr. J. F. Schofield, of Durban; Miss M. Wilman of the Alexander

McGregor Memorial Museum, Kimberley; Dr. Fischer of the Linden

Museum, Stuttgart; and Dr. Germann of the Volkerkunde Museum,

Leipzig, for permission to study and publish some of the specimens

in their collections.

REFERENCES.

1. Dos Santos, in Theal, Records of South-Eastern Africa, vol. vii, p. 210.

2. LepzELTER, Eingeborenen Kulturen in Siidwest-und Siid-Afrika, p. 93.

3. ScHappRA, The Khoisan Peoples of South Africa, p. 101.

4, Matncarp, “The Lost Tribes of the Cape,” S.A. Journ. Sci., 1931, p. 492 f.

5. TEN Ruyne, in Early Cape Hottentots, ed. I. Schapera, p. 153.

6. VEDDER, Das Alte Siidwest-Afrika, p. 51.

7. VeppER, Die Bergdama, Bd. 1, p. 76. LeszELTER, op. cit., p. 131.

8. Scutnz, Deutsch-Siidwest-Afrika, pp. 302 f.

9, LEBZELTER, op. cit., p. 217.

10. LEBZELTER, op. cit., p. 93. SCHAPERA, op. cit., p. 101.

11. BuzEx, The Naron, p. 18.

12. Tan RuyYNE, op. cit., pp. 123, 131, 135.

302 Annals of the South African Museum.

13. SCHINZ, op. cit., p. 96. DapprEr, in Early Cape Hottentots, ed. I. Schapera,

p. 35.

14. LEBZELTER, op. cit., p. 131. Irnwx, Die Gerero, p. 151.

15. IRL, op. cit., p. 114.

16. SCHINZ, op. cit., pp. 302 f.

17. TéngES, Ovamboland, p. 85. LEBZELTER, op. cit., p. 200.

18. Garon, Tropical South Africa (Minerva Library Ed.), p. 126.

19. TONJES, op. cit., p. 85.

20. ScHuLtTze, Siid-west-Afrika, pp. 253 f.

21. BurcHeEty, Travels in the Interior of S. Africa, vol. ii, pp. 322, 579.

22. Morrat, Missionary Labours, ed. 1846, p. 146.

23. CHapMAN, Travels in the Interior of 8. Africa, vol. i, p. 141.

24. ELLENBERGER, History of the Basuto, p. 71.

25. Stayt, The Bavenda, p. 50.

26. Hemans, ‘Customs of the Bashankwe,”’ Proc. Rhod. Sci. Soc., vol. xvi, pt. 2,

195. Ue

27. Dornan, Pygmies and Bushmen of the Kalahari, p. 122.

28. Hunter, Reaction to Contact, p. 87.

29. Buttock, The Mashona, p. 245.

30. Fritscu, Die Eingeborenen Siid-Afrikas, pp. 68 f.

31. Stayt, op. cit., p. 50.

32. BLEEK, “Bushmen of Angola,” Bantu Studies, vol. iii, pt. 2, p. 119.

33. CHAPMAN, op. cit., vol. ii, p. 286. MuLLER and SNELLEMAN, Industrie des

Caffres, pl. xi, figs. 4 and 5.

34. FRITSCH, op. cit., p. 69.

35. ScuuxttTzE, Aus Namaland und Kalahari, pp. 626 f.

36. THomas, Eleven Years in Central S. Africa, pp. 179 f.

37. DoRNAN, op. cit., p. 123.

38. Le VaILuant, Travels in the Interior of Africa (1796 ed.), vol. ii, p. 71.

39. BLEEK, op. cit., vol. iii, pt. 2, p. 119.

40. ScHuutz and Hammar, The New Africa, p. 202 f.

41. VepprErR, Die Bergdama, vol. i, p. 76. LEBZELTER, op. cit., pp. 131 f.

42. SCHAPERA, op. cit., p. 241 (quoting Kolb).

43. Krics, The Social System of the Zulus, p. 50.

44. Drury, Ann. 8. Afr. Museum, vol. xxiv, pt. ii, p. 99.

45. ENGELBRECHT, The Korana, p. 109.

46. Stow, The Native Races of 8. Africa, p. 52.

47. ScHINZ, op. cit., p. 52.

48. CHAPMAN, op. cit., vol. i, p. 311 (quoting Green).

49. CHAPMAN, ibid.

50. FLemrine, Southern Africa, pp. 202 f.

51. ScHINZ, op. cit., p. 93.

Plate LXXXYV.

. XXIV.

Vol

.S. Afr. Mus

wae aN

isso

Gal

Neill & Co., Lid.

M. Shaw.

Ann. 8. Afr. Mus., Vol. XXIV. Plate LXXXVI.

M, Shaw, Neill & Co., Ltd.

Ann. S. Afr. Mus., Vol. XXIV. Plate LXX XVII.

M. Shaw. Neill & Co., Lid.

Ann. 8. Afr. Mus., Vol. XXIV. Plate LX XXVIII.

M. Shaw. Netll & Co., Lid.

Ann. 8. Afr. Mus., Vol. XXIV. Plate LX X XIX.

sail A

M. Shaw. i Neill & Co., Lid.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XC.

(0 eres eel net Sas:

M. Shaw. Neill & Co., Lid.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XCI.

> === Sa 5

4 a at a /

M. Shaw. Neill & Co., Ltd.

Ann. §. Afr. Mus., Vol. XXIV. Plate XCII.

M. Shaw. Neill & Co., Lia.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XCITI.

M. Shaw. Neill & Co., Ltd.

Plate XCLY.

. XXIV.

Vol

Mus.,

Ann. S. Afr

Neill & Co., Lid.

M. Shaw.

Ann. 8. Afr. Mus., Vol. XXIV. Plate XCV.

M,. Shaw. ~ Neill & Co., Lid.

; f yay!

WAC RE vyiei tre :

‘ Aras

Ann. 8. Afr. Mus., Vol. XXIV. Plate XCVI.

Ut Lea ge hay

pacer cee: Hat

Mea IM MAR

oo eh] %\

awe ll

Heaton | aoe.

Leip

M. Shaw. Neill & Co., Ltd.

Ann. S. Afr. Mus., Vol. XXIV. Plate XCVII.

M. Shaw. Neill & Co., Ltd.

Ann. S. Afr. Mus., Vol. XXIV. Plate XCVIII.

M, Shaw. Neill & Co., Lid.

Ann. §. Afr. Mus., Vol. XXIV. Plate XCIX.

Ng i

M. Shaw, Neill & Co., Lid.

Me eee

PARTS OF ‘THE ANNALS PREVIOUSLY ISSUED—

Vol. 1.—Part 1, out of print; Part 2, out of print; Part 3, out ee Pp

Vol. I1.—Part 1, out of print; Part 2, 5/6; Part 3, out of print; Part

“Part 5, 1/6; Part 6, 3/-; Part 7, 1/6; Part 8, 3/-; Part 9, Fa

Part 10, 7/-; Part i 3/-; Index, “Title, etc., 1/6. Reise iss. cd

Vol. I1I.—Part 1, 2/6; Part 2, 1/6; Part 3, 5/6; Part 4, 3/-; Part 5, abe

5/6; Part 6, 7/-; Part 7, 1/6; Part 8, 3/-; Part 9, 1/6; ncaa Byes

‘Vitle, etc., 1/6. Ris, Tae hn

Vol. LV (containing "Palaeontological papers published in conjunction with the Sie)

Geological Survey ).—

Part 1, ll/-; Part 2, 7/-; Part 3, 4/6; Part 4, 4/6; Part 5, 2/6;

Part 6, 4/6; Part 7, 14/-; Part 8 (with Index, Title, etc.), 8/-.

Vol. V.—Part 1, 4/6; Part 2, 8/6; Part 3, 2/6; Part 4, 1/6; Part 5, 2/-;

Part 6, 5/—; Part 7, 3/-; Part 8, 4/6; Part 9, 4/6; Index, Title,

etc., 1/6.

Vol. VI.—Part 1, 13/6; Part 2, 4/6; Part 3, 3/6; Part 4, 30/-; Index, Title, _

etc., 1/6. ‘

Vol. VII (containing Palaeontological papers published in conjunction with the

Geological Survey).—

Part 1, 3/-; Part 2, out of print; Part 3, 5/-; Part 4, 8/-; Part 5,

5/6; Part 6, 1/6; Index, Title, etc., 1/6.

Vol. VIIL. ._—Complete, out of print. Index, Title, etc., 1/6.

Vol. [X.—Part 1, 4/6; Part 2, 5/6; Part 3, 10/-; Part 4, 6/6; Part 5, 3/6;

Part 6, 11/—; Part 7, 9/-; Index, Title, etc., 1/6.

Vol. X.—Part 1, 3/-; Part 2, 2/6; Part 3, 2/-; Part 4, 3/-; Part 5, 20/-;.

Part 6, 3/-; Part 7, 10/-; Part 8, 2/6; Part 9, 5/-; Part 10, 2/6;

Part 11, 20/-; Part 12, 7/—. Complete. Title, etc., contained in

Part 11.

Vol. XI.—Part 1, 3/6; Part 2, 2/-; Part 3, 13/6; Part 4, 1/6; Part 5, 17/-;

Part 6, 11/-; Part 7, including Index, Title, etc., Appendix and

Plate IL, 3/-.

Vol. XII (containing Palaeontological papers published in conjunction with the

Geological Survey).—

Part 1, 15/6; Part 2, 3/6; Part 3, 4/6; Part 4, 3/-; Part 5, 7/-;

Part 6, 6/-; Part 7, 20/-; Part 8, 20/-; Index, Title, etc., 1/6..

Vol. XIITI.—Part 1, 6/-; Part 2, 2/6; Part 3, 3/-; Part 4, 8/6; Part 5, 1/6;

Part 6, 5/-; Part 7, 30/-; Part 8, 1/—; Index, Title, etc., 1/6. -

Vol. XIV.—Part 1, 8/6; Part 2, 8/-; Part 3, 6/-; Part 4, 17/6; Part 5, 5/-;

Part 6, 9/—; Index, ‘Title, etc., 1/6.

Vol. XV.—Part 1, 17/-; Part 2, 17/-; Part 3, 14/-; Part 4, 12/6; Part 5, 5/6;

Part a 3/6; Index, Title, etc., 1/6.

Vol. XVI.—Part 1, 30/6; Part 2, 4/-; Part 3 (with Title, Index, etc.), 25/-.

Vol. XVII.—Part 1, 12/-; Part 2, 9/6; Part 3, 3/-; Part 4, 15/-; Part 5, 15/-;

Part 6, 2/6; Index, Title, etc., 1/6. :

Vol. XVIII.—Part 1, 20/-; Part 2, 7/6; Part 3, 30/-; Part 4, 12/6; Index, Title,

ete., 1/6.

Vol. XIX.—Part 1, 22/-; Part 2, 17/6; Part 3, 11/-; Part 4 (with Index, Title,

etc.), 5/6.

Vol. XX.—Part 1, 8/6; Part 2, 12/6; Part 3, 4/-; Part 4, 10/-; Part 5, 4/-;

Part 6 (with Title, etc.), 4/6.

Vol. XXI.—Part 1, 25/-; Part 2 (with Title, etc.), 30/-.

Vol. XXII (containing Palaeontological papers published in conjunction with the

Geological Survey).—

Part 1, 20/-; Part 2, 10/-; Part 3 (with Title, etc.), 3/6.

Vol. XXIII.—Part 1, 12/6; Part 2, 8/-; Part 3 (with Index, Title, ete.), 9/-.

Vol. XXIV.—Part 1, 10/-; Part 2, 2/6; Part 3, 4/6; Part 4, 11/-; Part 5 (with

Index, Title, ete.), 16/-.

Vol. XXV.—Part 1, 12/6; Part 2, 10/-; Part 3 (with Index, Title, etc.), 10/-.

Vol. XX VI.—Complete, 25/-.

Vol. XX VII.—Complete, 25/-.

Vol. XXVIII (containing Palaeontological papers published in conjunction with the

Geological Survey).—

Part 1, 10/-; Part 2, 17/6; Part 3, 6/—; Part 4 (with Index, Title, a

etc.), 10/-. " 4

Vol. XXIX.—Part 1, 20/-; Part 2 (with Index, Title, etc.), 20/-.

Vol. XXX.—Part 1, 15/6; Part 2, 20/-; Part 3, 8/-; Part 4, 9/-; Part 5 (with

Index, Title, ete.), 9/-.

Vol. XX XT (containing Palacontological papers published in conjunction with the

Geological Survey).—

Part 1, 20/—; Part 2, 13/6; Part 3, 17/6.

Vol. XX XIT.—Part 1, 2/6; Part 2, 7/6; Part 3, 17/-; Part 4, 1/-.

Vol. XXXIV.—Complete, 40/-.

The Annals of the South African Museum will be issued at irregular intervals, .

as matter for publication is available. vies

Copies may be obtained from— ~

Messrs. WHELDON & WESLEY, L2p., oad

2, 3, and 4 ArTHurR Street, New Oxrorp Srreet, Lonpon, W. C. 2; ie

The LIBRARIAN, Soutn Arrican Museum, Care Town. hs Spa lio

N “i bi