1>29 Dd Cee > * ort »>) 222 DD» >» wr vey yy Cie } Y Ud yuyu Pee oy -S eS D>. > »>d > 2 >> >>. > Se SS SPDe ay >) » D2. yD > > DD Md” © gs Bere 22-2. 2D mer 3D 2D DID DpH > DD. DS DD IDF D+ Dw . ag LDP » >: ss Ds D>» i DD 2I3ID> ID , DD Bee 2 s rs Do ean 2. 322. ye. DPR Dp. 2D? Dd 22> eae 5: eo a: i Soigort if : hel Nga ‘ ‘. h spe " ante fils NPM nt sh ee ANNALS OF THE SOUTH AFRICAN MUSEUM FOLOM SE VItT. erates! ANNALS OF THE SOUTH AFRICAN MUSEUM VOLUCM EV LF. PRINTED FOR THE TRUSTEES OF THE SOUTH AFRICAN MUSEUM By ApLtarp & Son & West NEwMAN, Lrp., Lonpon, L9id. TRUSTEES OF THE SOUTH AFRICAN MUSEUM. The Rt. Hon. Joon XAvrer MERRImMAN, P.C., M.L.A. Tuomas Morr, C.M.G., LL.D., M.A., F.R.S., F.R.S.E. Harry Bouvs, D.Sc., F.L.S. SCIENTIFIC STAFF OF THE SOUTH AFRICAN MUSEUM. Louis ALBERT Pértnavey, D.Sc., F.Z.S., F.E.S., Director. ArtHur WititAm Rogers, D.Se., F.G.8., Keeper of the Geological and Mineral Collections. Rosert Broom, D.Sc., M.D., Keeper of the Palaeontological Collections. Epwtn Percy Putruires, M.A., Assistant in charge of the Herbarium. Stpney Henry Haventon, B.A., Assistant in Geology and Palaeontology. KepreL Harcourt BarNnarp, M.A., Assistant in charge of Fish and Marine Invertebrate Collections. LIST OF CONTRIBUTORS. ' L. P&RINGUEY. PAGE The Stone Ages of South Africa as represented in the Collection of the South African Museum : : : , : : : ; 1 With chapters by : A. L. pu Tort. A chapter on the Sources of Rock for the Manufacture of Stone Imple- ments : : : : : : : : . : 5 ATK! F. C. SHRUBSALL. A chapter on Bushman Craniology . 3 3 ‘ ; 5 . 202 DATE OF ISSUE. July 5th, 1911. Note. This volume is to be regarded as complete, as no further parts will be issued. 4 ze “, eet is es =~ a i ee —s > re ; aes; ey”? f., o> a babe | r _ a Ria: 3} a | sys) cS 7) i a ew ot : in : a, Soi : : y ) a . ‘ Dae he ty ¢! ee J a ' A, : ' Lh F ) = : : 4 >| 1 i " s * ; saa — . 1 ' ) : v 1 - i f é & - ® v t = f - 7 3 a “+ Pils A > —- ehh, S "7 + | A 7 * : : a aie ay ay Nie 7 et eee ANNALS OF THE SOUTH AFRICAN MUSEUM. (Vou. VIIL.) 1.—The Stone Ages of South Africa as represented in the Collec- tion of the South African Musewm.—By L. Périneuey, D.Se., Director. INTRODUCTION. I must explain, a priori, that although for the purpose of illustra- tion I divide, or make an attempt at dividing, the South African Stone implements into several series or types, no definite proof has as yet been forthcoming, as will be seen hereafter, that the small, rude implements used here until a few years ago were not utilised con- temporaneously with the large, roughly or finely trimmed, tongue- or almond-shaped implements, or the smaller and perhaps still more perfectly worked tools which are met with so profusely in South Africa. As will be explained as this paper proceeds, the industrial traces left by the people that inhabited South Africa in times which may or may not synchronise with the stages recognised in Europe, con- sist of three distinct types, and also of a fourth, which is, however, difficult to define. The First Type.—This type is one considered to be comparable with the oldest division in Europe, which is based there on very good paleontological evidence, but not always indisputably backed, however, by geological proofs. Tools of this type are known to occur all over the world except in Australia. So much alike are 1 2 Annals of the South African Museum. they that it is difficult to believe they could have evolved inde- pendently, but it is probable that the African examples had their origin in fragments due to accidental fracturing of rocks. The proposition that their evolution took place in Africa, and perhaps South Africa, seems to me quite plausible. Were it possible to postulate for man (whether anthropomorphous ape or not) a very ancient origin—ancient in a geological sense—it is certainly to Africa that one would turn to find his original home, because for incalcu- lable time a large part of Africa has been uncovered by the sea. This is, however, a hazardous plea, because the existence of Ter- tiary man has hitherto not been asserted with any likelihood of probability. The type aforesaid consists of massive fragments of rock trimmed sometimes on one, sometimes on both sides, into cleaving, digging, or smiting artefacts, which, whether found in Europe, Asia, Central America, &c., bear such an extraordinary resemblance to each other that one is forced to the conclusion that the type could not have been invented in places so far apart in a spontaneous manner. Although it is quite possible that stone implements of a different character may have evolved from the growing intellectual power of man, it seems impossible that a uniform shape could have resulted from what must of necessity have been rude, uncouth methods, and from material differing in composition. ‘To intercourse or migration of races this result is probably due. On the other hand, if the natural texture of the material used allows under any kind of concussion the preliminary fracture that leads to the evolution of the finer artefacts, then, of course, we may dismiss from our mind these important factors—immigration, emigration, or intercourse. Professor W. J. Sollas has proposed lately the term ‘‘ bouwcher”’ for this type, which I should have otherwise dubbed “ paleeolith,” in honour of Boucher de Perthes, the French savant who was, if not the first discoverer, at least the first interpreter, of the occurrence of implements of that type which he found in the gravels of the river Somme. It is highly desirable that this term should be adopted, because the other appellations either mean nothing, or imply a purpose for which they were probably never intended. One point, however, cannot be disproved, and that is, the South African paloliths, other than the ubiquitous and probably surviving type of knife-scraper, correspond to the “ celt” of the English, the ‘coup de poing”’ or the “ hache 4 talon”’ of the French, the “ beil”’ of the German, the “‘hachas’’ of the Spaniard, &c., &e. Of the The Stone Ages of South Africa. 3 authors of these bouchers of palwolithic type known as Chellean, no other cultural trace remains, or is supposed to remain, in places outside South Africa, but this may be due to the fact that relics of their domestic utensils have disappeared, or at least have not been found in siti together with the ‘ bouchers.”’ Neither are the traces of the Mousterian stage of culture as clear as one could desire. In South Africa, however, the doubt is no longer permis- sible, as the evidence I adduce will show. The Second Type.—The second type of stone implements is in some respects more primitive ; occasionally it is of a superior finish, but still primitive. It has lasted until quite lately—a few years back, as a matter of fact; it has also probably replaced the former lithic industry. It may be termed South African Neolithic. It includes household utensils, mortars, querns, mullers, un- doubtedly polished by usage and not intentionally ; we have the “!kweé,” or perforated disk or orb; we have also pottery of a type unknown elsewhere, beads and ornaments of stone and clay, of shells and ostrich egg-shells, bone tools, &c., and also rock paintings and, possibly, rock gravings. It is, however, doubtful if the latter _ should not be ascribed to the Paleolithic. These two types are not often found together, yet they are occa- sionally met with in close proximity, owing probably to orographical conditions, such as the neighbourhood of streams or rivers that have persisted in their continuance ;* subsequent occupation of some points of vantage has led also to these artefacts, made at periods widely separated, coming together ultimately. The Third Type.—Uastly, we have recorded a few instances of implements the technique of which is that of the true Neolithic European period: small arrow-heads trimmed on either side and with a carefully worked peduncle, or ‘‘ tang,” for hafting, and a stone axe with ground edge, all made of local rocks. But before adducing my reasons for believing that this multiplicity of form is ascribable to a plurality of races, some extremely ancient, others less so, and others again well-nigh contemporaneous, I must warn the student that it should not be taken for granted that the evolution or transformation of an industry is always on the lines of progress. Evolution often stops, sometimes to start again on its onward career, sometimes—and oftener than not—to retro- grade, in the sense that the first line with which we connect it is gradually abandoned for a less complex but not necessarily * It is a well-accepted fact that primitive people always settled near running waters in the valleys. a Annals of the South African Museum. ineffective one, which, through supplanting the former, obliterates its traces. For comparison I have, purposely, somewhat neglected indications afforded by results obtained in England, Northern Europe, or Northern America, not that these indications are not valuable in themselves, but because the composition of the material used in the lithic industry of South Africa, and the resulting produce of the same, clearly assimilates it to that prevailing in Southern Europe, from the Pyrénées eastward and southwards. This paper is not an attempt to try and solve problems of great consequence for that section of the science of Anthropology dealing with the stone implements, the artefacts of man who had ceased to be anthropomorphous ape. It is a recapitulation, it can hardly be called a narrative, of information obtained in South Africa, classified wrongly or rightly according to the tenets obtaining now. It is the embodiment of some thirty years’ research, and if the explanations can be challenged or criticised the numerous illustra- tions will doubtlessly escape that fate. Renewed activity for the last ten years, in the search for these relics—an activity resulting from the discovery of important deposits—in which many and zealous collaborators have joined—has enabled us, at the South African Museum, to accumulate material from every part of South Africa, and of many from beyond. This material forms certainly the most complete collection of its kind. In addition, I was enabled by the courtesy of their owners to examine, photograph, and make casts of certain examples not repre- sented in the Museum Collection. I have been greatly aided by the members of the Staff of the Geological Survey of the Cape Colony in matters relating to the geological formations or sites of the implements found, many of them through their own exertions. I would fail in my duty if I did not make special mention of Mr. J. M. Bain, without whose intelligence, liberality, and absolutely gratuitous aid my attempt at discriminating in the intricate questions of the South African Lithice Ages would have been greatly impaired, and the results more incom- plete. Many are those who also proffered help, advice, and suggestions. To name them all would necessitate many lines of print; but the omission of their respective names will not, I know, be by them taken amiss, for, indeed, unselfishly they toiled. To Mr. A. R. Walker, of the South African Museum, I am much indebted for his assistance in photographing many of the numerous objects illustrating this paper. THE PALAOLITHIC. CHAPTER I. SITUATION AND COMPOSITION OF THE PALHOLITHIC AND NEOLITHIC SoutH AFRICAN IMPLEMENTS. In 1866 the late Sir Langham Dale discovered close to his residence on the Cape Flats, near Cape Town, stones showing plain marks of artificial working. These examples were submitted to experts in England, who pronounced them to be undoubtedly man’s handiwork. To the present generation it seems almost incredible that doubt about the workmanship of these implements could have ever been entertained, because among them were the best finished examples of a Solutrian type ever found, and of which two more only have been met with since. Willing searchers volunteered their services, and this discovery was followed by numerous ones in the Cape Colony, the Transkei, Griqualand West, where these artefacts were found embedded in mining claims “intermixed with precious stones in the diamond- diggings’; later on in Natal, the Transvaal, Southern and Northern Rhodesia, Swaziland, Bechuanaland, the Kalahari region, Mossamedes, &e., &e. In fact, these relics of primitive civilisation, be they digging- stones or hand-picks, cleaving-stones or axes, flakes having served as knives, saws, burins, piercers, scrapers, or perforated disks for weight-making, orbicular stones for hand-throwing, or perhaps slinging, smoothed pounders, mullers, querns, or mortars, stones grooved by sharpening bone skewers or bodkins, or by reducing to shape the bone shaft of arrows, whether of huge size or ridiculously small, they all abound in South Africa from west to east, from south to north. When they are of a type that might be assimilated perhaps to the 6 Annals of the South African Musewm. Aurignacian or Magdalenian, they are exposed on the surface, or occur in shell mounds or in rock-shelters. They are found, occasion- ally also with more ancient types, on the floor of huge sand-dunes by the sea-coast, when these are exposed and bared, to be no less periodically covered again, by the boisterous prevailing winds. They are common near the water-places called here “‘ fonteins,”’ 7.e., springs, and mostly always near to, or in, depressions where rain-water accumulates in the season: the ‘‘ vleis,”’ or “‘ pans” of South Africa. When of a more ancient type, Chellean-Mousterian, they are bedded in alluvial deposits, often very deeply. They occur in numbers on the talus of mountains and high hills. They are met with, but then mostly singly, in the exposed banks of rivers ; occasionally they are found on the surface, or where river terraces occur which, however, are not proved to be old. Often also they are found singly, where no trace of land erosion is perceptible or traceable, unless we go back to early pliocene—this showing plainly that their presence there is purely accidental. The material is always a rock of hard texture; no implement made from a flint nodule has as yet been found, because the material does not exist in South Africa. The hardest stone occurring locally or at some distance off has been selected for the large and small implements. It isTable Mountain sandstone (more or less quartzitic), Karroo quartzite, dolerite, lydianite, or shale indurated by the in- trusion of dolerite, surface quartzite of various textures, cherty sandstone, Dwyka chert, banded jasper, diabase, agate, and chalce- dony, white quartz either sugary or transparent, even granite. Implements made of green bottle and white plate glass have also been found. It happens not unfrequently that implements are met with in situations where the rock of which they are made is known to be absent. Barter may account for their presence there, but it is most likely that they were carried and left where found by owners of migratory or roaming habits or dispositions, clan-forming aborigines that have disappeared, leaving behind them, however, these artefacts as a testimony to their former existence. It soon becomes plain, even after a superficial examination, that the making of implements of forms so various cannot have been simultaneous. The technique is too dissimilar, the general facies also. Next to the scraper-knife flaked off a hard stone for a passing want and probably discarded immediately after, we find a laurel leaf- shaped lance-head worked by careful secondary trimming on either side and of nearly pure Solutrian type; a ‘coup de poing”’ of a finish The Stone Ages of South Africa. 7 equal to the best Acheulean. We have a cleaving-stone surpass- ing the best Mousterian ; a rude, irregular stone with cutting edges fixed with a gum-cement to a wooden handle in the manner obtain- ing among the Australian aborigines ; arrows, the cutting or piercing heads of which is obtained by minute chips set in a triangular piece of similar gum-cement, a few arrow-heads with tang, worked on both sides; and a ground axe of neolithic type are aiso recorded. The evolution in the manufacture of these tools took probably a very long time in South Africa, as elsewhere. I have already ex- pressed my belief,* based on purely antiquarian grounds, and according to the tenets of the classification generally accepted, that ‘we have in South Africa evidence of two periods: a paliwolithie and a recent one, which I hesitated to term neolithic; but that there is no evidence as to the time when the former was replaced by the latter, and moreover, that this point will remain for long conjec- tural.”’ In a word, a very ancient race had peopled Africa at the paleolithic stage. One or more races have supervened, possibly absorbed the former, and perhaps replaced it. Unfortunately, neither geology nor paleontology has been able to give us, so far, a clue to the possible age of the South African finds. The question is still more complicated owing to material of a palewolithic type of the highest finish, as well as of ruder kind, having been found in valleys where old river terraces cannot be traced, as well as where river terraces exist, or are said to exist. Then, alongside of these we have implements quite modern, as will be seen subsequently, and yet so primitive in appearance that one can excuse, yet not agree with, those antiquarians who, re- quiring a beginning to everything, have postulated that thorny subject an ‘‘eolithic” age preceding or accompanying the ‘ Strepyan.”’ Nor does the difference in composition of the material of which the implements are made help us to elucidate the point of antiquity. A hand-pick of dolerite will be weathered to a stage of unrecognition, while a quartzite one will, during the same time, merely acquire a patina, or polish ; a chert or banded jasper tool will remain almost as fresh as when made, while a diabasic one will become deeply pitted or smoothed under similar conditions. Eolian agencies have also to be taken into serious consideration, in a country where desiccation has been in progress, especially in the * “The Stone Age in South Africa,’’ in ‘* Science in South Africa,’ Cape Town, 1905, a publication prepared for the visit of the British Association for the Advance- ment of Science. The present paper is an enlargement of the necessarily highly condensed précis I then gave of our knowledge of the question. 8 Annals of the South African Museum. north-western part, from probably the beginning of the quaternary period. Great, almost unsurmountable, therefore are the difficulties front- ing the Antiquarian in South Africa: first, because geology and paleontology fail him in affording precise indications of an old period from which deductions other than speculative might be drawn ; secondly, because the Stone Age is not yet an age of the past, or if so, it ended yesterday; thirdly, because, with one exception, there is no evidence of a Polished or Ground Stone period having replaced the former and preceded a Bronze or Iron Age as in the Palearctic Region. The only resource left to him is to turn to the comparative study of the implements themselves, but he is soon led to conclude, on lithological grounds, that these South African implements do not fit in with the classification that answers to the requirements of, and is founded upon, the evidence obtained in Europe. The latter classification is based on stratigraphical and palzonto- logical evidence, and it depends also on certain industries * which unfortunately did not extend to South Africa. Classifications are made to be unmade when new discoveries occur. But it is not possible to make the known South African finds fit in with the classifications of Mortillet or of other authors. There is, moreover, a chain of evidence being slowly forged which points to a resemblance between implements from the Old and those of the New World. This similarity of form is so striking that it makes the Antiquarian pause when he considers the question of the possible identity of the races of mankind that manufactured these implements. Nor is he easy in his mind that this lithic industry is not the result of causes due to the growing intellectual power of man, affecting people in widely distant countries at the same or different times. He has then to eall the Anthropologist and the Ethnologist to his aid. In spite of the fact that a community of races is not implied by a like condition of culture, the Philologist may also be asked to add his quota, although his great error is, and has always been, to ‘‘ treat a communicable character as an inborn gift.” Thus reduced merely to a lithological comparison, the study of the South African implements might appear to prove barren of results. But it is not so. The Chellean type is the Chellean type of the Palearctic and other regions. This is indubitable. But the types * The Magdalenian, connected with the reindeer, and perhaps late on with the stag, is a case in point. The Stone Ages of South Africa. 9 that might correspond with the Aurignacian, Solutrian, and Magda- lenian cultures, especially the last, have an indescribable facies of their own which may be said to be South African. On the other hand, the “‘ pygmy’ implements, and others with the “ bord abattu”’ of the French, cannot be very readily distinguished from the English, French, and Indian implements of the same type, except, of course, by the material of which they are made; but they more closely approximate the Algerian and Morocco examples. The South African Aurignacian or Magdalenian type, may have been, and probably was, as old as that of the corresponding period of Europe ; but it has outlivedit. The‘ pygmy” culture lasted in the Cape Colony until the sixties of the last century, or thereabouts, and is lasting still in the Kalahari.* Truly we have not here a definite line of separation between the artefacts that are hacked stones or those that are polished stones, in so far as concerns weapons or tools that might have been used as weapons; but we have here an abundance of household utensils that might prove a counterpart to the age of the polished stone, but which have a facies eminently South African. This peculiar feature of what I prefer, rightly or wrongly, to term the South African Neolithic type is that, although for certain purposes stones were polished, yet it cannot be said that an attempt was made to make the weapons of the same period more serviceable or more effective by this polishing or grinding process. There is thus a big hiatus in the evolution of South African stone implements. * We find it connected also in some Cape caves with large implements of paleolithic type. CHAPTER II. DIVISIONS OF THE PAuLxonITHIC SERIES. Eouiras. PALXOLITHS OF LARGE SIZE OTHER THAN SCRAPER-KNIVES. ‘THE DIFFERENT Types oF SoutH AFRICAN PALHOLITHIC IMPLEMENTS. In order to compare our South African series and make them synchronise with those met with elsewhere, it may not be out of place—in fact, it is necessary—to give here, but on broad lines, he generally accepted divisions of the Pleistocene, or ancient Quaternary. Whether they are justified or not for the South African implements, the general terms will prove useful; but whether these divisions as now accepted will prove provisional, in view of the later discoveries of human remains, is a question of the future. These divisions, beginning from the lowest, are :— Strepyan Chellean Acheulean Mousterian Solutrian Magdalenian Azilian Another arrangement rejects the Strepyan and accepts the— Chellean Acheulean Mousterian Aurignacian Solutrian Magdalenian Azilian The Stone Ages of South Africa. Ad A third arrangement, and one which would seem more suited to the South African finds, rejects the Strepyan and unites the Chellean, Acheulean, and Mousterian in a sub-division called the Chelleo- Mousterian. The reasons for these divisions are as follows :— CHELLEAN.—During this period Europe was a warm country with a mild and damp climate. The fauna is characterised by the pre- sence of the Hippopotamus and Elephas antiquus. On the archeo- logical side were prevalent stone implements trimmed on both faces, and of an amygdaloidal (almond) shape. ACHEULEAN.—With the Chelléan is closely connected the Acheulean, which may be termed a phase of transition. The fauna consists of the animals of the Chellean period, among which, however, and in certain places, are found others belonging to the period following, 1.e., the Mousterian. Mousterian.—During the Mousterian, termed also Middle Paleolithic, the temperature is lower and the fauna is that of a cold, moist climate. The Mammoth, an animal of some 16 to 18 feet in height, takes the place of its still larger predecessor, Elephas antiquus; it is associated with the woolly rhinoceros, haying two horns over the nose, the larger of the two sometimes 3 feet in length. The Acheulean implement is still occasionally, but seldom, met with. The Mousterian implement is usually chipped on one face only; it is probably detached from the matrix, or nucleus, whether flint or quartzite, at one blow after or before the outer face has been shaped into the requisite manner; the reverse side shows well-nigh invariably the convex node called ‘‘bulb of percussion.”” The maximum size is about 6 inches long, the average 24; the edges are often very carefully retouched (secondary chipping). SoLUTRIAN- MAGDALENIAN.—The characteristic of this period is a dry cold succeeding to a moist, damp cold. During this period the severity of the climate, especially in winter, induces man to seek shelter to avoid partly its rigour. That shelter he finds in the caves or grottoes formed naturally in calcareous formations, or under hanging rocks. The fere nature multiply. Man’s mental powers are taxed to their utmost to resist his natural enemies, maybe by brute foree, but much more likely by craft, in order to obtain his food, and to secure garments as protection from the severity of the climate. This period, divided in two successive ones, is the age pre-eminently of the reindeer, and it leads progressively 12 Annals of the South African Museum. from the old Quaternary or Pleistocene to the actual Quaternary or Holocene. During the age of the hippopotamus or the elephant (H. antiquus), Chellean man’s only relics are stone implements—feeble weapons, after all, against the redoubtable foes he has to encounter. But from sheer necessity growing wants lead to the invention of more complicated implements or tools, and during the Reindeer period these forms were multiplying in adaptation for special purposes. Slender, even delicate, tools of ivory, bone, horn occur side by side with stone implements of different shapes, possibly pre-Solutrian. Flint lance-heads admirably worked on each side, arrow-heads with peduncles, or “ tangs,’’ exhibiting a wonderful progress in ‘“ hand- knapping,” are found together with simple, or barbed darts, arrow-heads, some simple, some notched, and others triangularly incised at the base; bodkins, &c., made of ivory or bone are common. It is at that time that the sense of what we call “art” appears to awake in the mind of this cave-dweller, or Troglodyte, and he gives expression to it in the shape of sculpture or painting, petroglyphs and glyptics, some of which are admirably preserved to this day, and treated in a manner that throws our so-called Bushman paintings completely into the shade. Yet this man’s predecessor, the late Mousterian, is dolichocephalic with an index of 75, and a breadth height of 62:5, that is to say with measurements corresponding with a lower stage than those of any existing race. The height is moderate, 1™ 60; he has large orbits, “the superciliary ridge forms a kind of vizor above them”’ ; the jaw is powerful; there is a total absence of chin, &c.’’ * However primitive this middle pleistocene man may seem if his physical characters only are taken into consideration, yet he buries his dead with care. Of the skeleton found in the classic grotto of ‘‘ Le Moustier,”’ it is said: ‘The posture is that of repose with the face turned to the right, the right arm is under the head which is surrounded by flint flakes. Beside the skeleton were found, in addition to the flint implements of the Mousterian type, some of the Acheulean, among them a splendidly worked ‘“ hand- wedge.”’ + A human jaw has been also lately discovered to which an older * Haddon, A. C., ‘‘ Paleolithic Man,” Nature, 1909, July 29th. + Boule, M., ‘‘L’homme fossile de la Chapelle aux Saints,’”’ L’Anthropologie, 1908, p. 519. The Stone Ages of South Africa. 13 still origin is attributed, but no implements seem to have been found with it.* For our purpose the lithological or other divisions following those already enunciated beyond, if not including the Azilian, do not bear any distinct connection with the South African; and even the Solutrian-Magdalenian stage does not seem to exist here in the succession claimed for it in the Palearctic region. But we have traces of culture in the shape of petroglyghs or glyptics that remind one of those of the Solutrian, if not Aurignacian, period. We have also an autochthonous race, or the remains of it, the ‘‘San,’’ Strand Looper, Barwa or Bushman, the cephalic index of which, inferior to that of any other living man, is perhaps only slightly superior to that of the man of the “Chapelle aux Saints.” + Members of that race were undoubtedly capable of producing glyptics; of this there is no doubt. It is not so certain, however, that the numerous petroglyphs (rock engravings), fairly numerous in South Africa, were executed by them.{ But of one thing we are sure, and that is that these rock engravings have been executed with stone tools.§ EOLItTHS. It seems logical that there should have been a beginning in the manufacture of these stone implements that show a considerable skill in the making. Primitive man made use of round, flat, or pointed stones, either rolled by water, detached, or sharpened by natural or accidental agencies; that afterwards he attempted to improve on them, or made them more suited to a requisite purpose, appears to us now very natural. This hypothesis is, moreover, borne out by the progress made in the manufacture of implements from the palzolithic type to the polished Stone Age forms. When, therefore, flints somewhat or greatly amorphous in general appearance, and which were afterwards dubbed “ eoliths,’’ were firstly discovered on the Kent Plateau, in England, by Harrison, it was claimed by him and others on the ground that evolution should * Alsberg, M., ‘‘ Recently discovered Fossil Human Remains,” Globus, vol. xev., 1909. + The cranial capacity of Bushmen is 1-285 c.c., according to Shrubsall, ‘Notes on some Bushmen Crania, &c.,’’ Ann. 8. Afric. Mus., v., 1907, p. 235. Previous observations had given 1:330 ¢.c. (male) and 1-255 (female). { Péringuey, ‘‘On Rock Engravings of Animals and the Human Figure, the Work of South African Aborigines, &c.,” loc. cit., vol. xvi., 1906. § Péringuey, ‘‘ On Rock Engravings of Animals and the Human Figure found in South Africa,”’ Trans. 8. Africa, Phil. Soe., vol. xviii., 1909. 14 Annals of the South African Musewm. have had a beginning, that these flints were the first attempt of man at trimming and working stone for his requirements. This discovery, followed by similar ones in Europe, led many Antiquarians to adopt the theory. It may not be out of place to give here an explanation of the making of tools by blows or percussion. If the blow is direct and delivered with great force the detached part is not bounded by a plane surface. Close to the part struck there is produced a conchoidal fracture, which gives rise to a ‘‘bulb of percussion.”” That is to say, one of the faces of the detached part presents at its thicker extremity a convex swelling corresponding to a concave cavity in the matrix or nucleus. This I may add, in passing, invariably occurs not only with flint nodules but also with all kinds of rocks used for that purpose in South Africa. This conchoidal fracture seems, however, to be altogether absent in the flint “eoliths.’ M. M. Boule has also shown pretty con- clusively that mechanical agents easily and naturally transform flint nodules into ‘eoliths.””* The discovery of mechanically made * pseudo-eoliths ’’ has undoubtedly modified the views of many of those who, endowing primitive races with characters which could have resulted only from evolutionary progress, were inclined to find in these so-called implements with “trimmed edges,” ‘“ double-,’’ “ crescent-shaped,” ‘ hollow-end,”’ ‘‘ horse-shoe’’ scrapers the most ancient attempts of man at the manufacture of tools. I may add that with the human remains of the Neanderthal-Spy— Le Moustier, et La Chapelle aux Saints race—remains of a man of an inferior type, more closely connected with the anthropoid apes than with any other ancient human group, such as the ‘“Cro- Magnon’’—were found stone implements, not eolithic, but of the Acheulean and Mousterian types. I would not have entered into this thorny question were it not that it is claimed that ‘“ eoliths”’ are found, among other places, near Pretoria, Transvaal, and also that they have been figured as such. | It is therefore not out of place to give here the history of that discovery. The late Mr. G. Leith made in the neighbourhood of Pretoria a collection of stone implements, mostly from the ironstone gravels through which the Aapies River flows. Mr. Nichol Brown, a co-worker of Harrison of the Kent Plateau eolith fame, had occasion to inspect that collection in 1897. He showed me some of these * M. Boule, ‘‘ L’origine des éolithes,’’ L’Anthropologie, 1905, p. 253. + G. Leith, “‘On the Caves, Shell-mounds and Stone Implements of South Africa,”’ Journ. of Anthropolog. Instit., i., 1899, pl. 18. The Stone Ages of South Africa. 15 notched scrapers, asking whether I considered them to be the handi- work of man. I replied in the affirmative, adding, however, that Nature’s agency had a great deal to do with their present state. Two years afterwards Mr. Leith published his paper, illustrated by one plate bearing the legend, ‘‘ Eoliths from Pretoria.”’ It is not possible to attribute to the implements figured by Leith a more ancient origin than to the Chellean, Acheulean, or Mousterian types occurring so abundantly all over South Africa; and this for the following reasons :— 1, These Pretoria ironstone river gravels are not very ancient. 2. The notched scrapers are the rarest. 3. These scrapers are not as smoothly polished as represented in the plate, this smoothness being due to the process-block reproduction. 4. Only those examples that would prove likely to support the ‘‘eolithic”’ theory were selected for reproduction. 5. Well-finished palwoliths were afterwards found in sitw, showing therefore contemporaneity with those so-called Eoliths. I have received from Mr. Leith himself, and also from other con- tributors, a considerable number of these implements; they show either no abrasion of the edges in many cases, and again consider- able abrasion in others; some are notched, others are not ; some are more or less polished ; some again show no sign of having been smoothed either by eolian or water action. But when abraded or partly polished the action is clearly due to the flow of an intermittent river. If one compares these specimens with some of the crescent-shaped Eoliths of Harrison, and Reutot, one is certainly struck with the similarity of the deep arcuate emargination of the thinner part of the Pretorian pseudo-eoliths. But the explanation is a simple one. Drawn into the vortex of turbulent waters while in flood, the denser part of the chipped stone offers more resistance to the moving agent, and the thinner lighter part is therefore thrown more forcibly into contact with the abrading obstacles, and suffers in consequence. This is especially illustrated in the knife-scrapers, or chips of the silicious rock found near the Victoria Falls. In no case is the thicker part of the chip dented in the horseshoe fashion claimed for the Kent eoliths * (cf. Cuts 3, 6, 7, 8 of Fig. 119, Pl. XV.). For the above reasons the “ eolithic ’ origin of the Pretoria imple- * Fig. 28 is a good instance of an abraded chip that on Mr. Leigh’s theory should be considered as an eolith. Yet it was found at the foot of Port Elizabeth Hill together with extremely water-worn quartzite implements (Figs. 25, 26, 27, Elev.) 16 Annals of the South African Musewm. ments must be dismissed. They are not the precursors of the highly finished or rougher South African types, and are neither the initial nor the secondary stage of the South African paleolithic forms. PALAOLITHS OF LARGE SIZE OTHER THAN SCRAPER-KNIVES. But if we reject, for the reasons already given, the evidence of the Pretoria implements as eolithic, we are faced with forms of a type and technique so truly palzolithic, and especialiy lower paleolithic, that doubt as to the identity of shape is no longer possible.* Their resemblance to the quartzite implements found in Kurope, Aigeria, Congo, India, is indegd extreme, and the process of manufacture appears to have been the same. The late E. T. Hamy was quite justified in stating of a quartzite cleaving, or hand-wedge boucher found at Koffyfontein, in Orangia, that, if not made aware of its source, a French or Spanish ethno- grapher would be justified in pronouncing it to have been found in one of the palolithic “ stations’ of the Haute-Garonne, in France,. or in the neighbourhood of Madrid. + So alike, indeed, to the bouchers of the neighbourhood of Toulouse, as figured and described by J. B. Moulet,} are the bouchers of the Stellenbosch-type, that the reproduction of Moulet’s own plates could have served for a great part of the illustrations given in this paper. It is not only to the quartzite implements of France or Spain that the verisimilitude of the South African bouchers is restricted. As far back as 1868 R. Bruce Foot called attention § to implements. met with in the laterite of Madras. Such artefacts have been found in the Narbadda Valley in Hindustan || and other localities. We have in the Museum specimens from Codapah, in the Madras. Presidency, resembling so much in material and workmanship the South African examples as to be verily indistinguishable. The same may be said of some of the Congo quartzite implements {/ ; of those * ITcannot refrain from quoting here a phrase in a letter I received from M. Cartailhac, the veteran of French Antiquarians: ‘‘ Mais leur aspect suffit pour les faire reconnaitre, aussi bien que si c’étaient des médailles de César ou de Victoria.”’ + Hamy, Bullet. Mus. d’Hist. Natur., 1899, No. 6. t ‘‘Etude sur les cailloux taillés par percussion du pays toulousain,” Archiv. Mus. Hist. Nat. Toulouse, pt. 2, 1880. § ‘On Quartzite Implements of Paleolithic Type from the East Coast of Southern India,” Intern. Congr. Anthropol., Norwich, 1868, p. 249. \| Sollas, ‘‘ Paleolithic Races,’’ Science Progress, pt. 2, 1909. “| Cf. X. Stainier, ‘*‘ L’Age de pierre au Congo,’? Ann. Mus. d. Congo, i., pt. 1, 1899. V. Jacques, ‘‘Instruments de pierre du Congo,’ Mém. Soc. Anthrop. Bruxelles, xix., 1901. The Stone Ages of South Africa. i, from Algeria,* &e., &¢., which form part of the Museum collec- tion. I shall never forget my pleasurable astonishment in discovering quite accidentally in a clay pit in Stellenbosch one of the most perfect implements of its kind, but of relatively small size (Pl. L., Fig. 4). This find led to the discovery of deposits or ‘‘ stations’ all over the Stellenbosch and neighbouring districts in the Cape Colony— discoveries which still continue. THE DIFFERENT TYPES OF SOUTH AFRICAN PAL2OLITHIC IMPLEMENTS. In spite of their general resemblance to each other, a glance at the illustrations completing this paper suffices to show that these bouchers exhibit a discrepancy in form and workmanship, in tech- nique and composition, which, although in many cases slight to the unaccustomed eye, may be taken to indicate that they have not all been manufactured by the same race, or at the same time, or in the same localities. Probably through long practice, I find no difficulty now in dividing the South African bouchers into several types, owing to their appearance or facies or to the material of which they are made. I thus venture an attempt at the classification of the types as I understand them. I must explain, however, that this classification is based on those implements which by their shape, size, or con- figuration correspond with the cleaving, digging, or smiting tools of the Chellean-Mousterian periods, as generally accepted. The evi- dence afforded by scrapers, knives, missiles, burins or bores, pottery or other implements of domestic use cannot be taken into account here. It will be fully treated in other chapters. I may preface the description of the stones by explaining that in the best finished implements which are not obviously cleavers, hand- wedges, or axes, the tongue or amygdaloidal shape prevails; the ovoid or discoidal is extremely rare, no matter to what type the implement may belong (cf. Fig. 24, Pl. III.). I may also state that the specimens figured have not been selected on account of their singularity or finish, and although no two pieces are ever alike, yet the figures are strictly representative of various forms occurring in South Africa. * M. Boule, “ Ktude paléontolog. et Archéolog. sur la station paléolithique du lac Karar (Algérie),” L’ Anthropologie, xi., 1900. 2 18 Annals of the South African Museum. STELLENBOSCH TypE.—It is in the valleys of the Eerste, Berg, Breede, Oliphants, and sundry rivers and their affluents in the Cape Colony that some of the best finished implements of that type have been discovered hitherto. Made of quartzite varying in closeness of texture (Table Mountain sandstone), but the grain of which, however crystalline the sand- stone may be, precludes the possibility of their ever acquiring a fine polish, they are of a type so numerously illustrated in all the South African districts of the Cape Colony and also beyond (Cape, Stellen- bosch, Paarl, Worcester, Tulbagh, Ceres, Clanwilliam, Malmesbury, Piquetberg, Caledon, Mossel Bay, Knysna, Port Elizabeth) that they may well be ranked under the name,* “ Stellenbosch type.” They are all broadly flaked either on one face or both; the edges are sharp, but very sinuous, and often continued round the butt, but they often also retain there the rounded original pebble shape; they show no distinct secondary trimming except faint traces at the point ; none are rectilinear in profile. Figs. 1, 2, 3, 4, 20, 32 represent the best finished tongue-shaped, Stellenbosch-type implements ; they are chipped on both faces, but the trimming of one side does not correspond with that of the other. Very variable indeed in shape and size are the artificially worked stones of that type. The crudeness and imperfection of some of them contrast singularly with that of Figs. 1, 2, 3, &c.; they are no longer tongue-shaped or amygdaloidal (see Figs. 12, 13, 14, 19). The butt is shaped into a rough point, as in Figs. 12 or 15; both ends may be reduced almost equally to a sharp point (Fig. 14), or one of the points into a broad wedge (Fig. 13); the apices are thus variable in form. But whatever the shape be, crude or finished, the median, always sharp, ridge of the two faces as shown in the absolute profile of the bouchers does never correspond. Next to, and found together with, the highly finished implements comparable, except in size, with the best Acheulean (Figs. 1, 2,-3, 20, 32), is Hamy’s “hache a talon’”’ (Figs. 5, 6, 7,8), in which a part of the water-worn, rounded quartzite boulder has been retained more for convenience than through accident. This retention of the contour of the boulder, or of part of it, is of very frequent occurrence in the Stellenbosch—more so than in any of the other types I have seen (Figs. 7, 8, 9, 10, 29) ; it is extremely pronounced in Figs. 25 and 26. * I have seen an implement of that type alleged to have been found in Natal. It is in the G. Leith Collection, now in the Pretoria Museum. I know of similar ones from East London, as well as from Swaziland. The Stone Ages of South Africa. 19 Many of the bouchers of the Stellenbosch type retain also on one side the primitive cleavage with very little paring; but this is not restricted to that type only. Orance River Type.—If we were to consider only the palzolithic industry of the southern districts of Cape Colony its homogeneous form would certainly stand as absolutely typical, but proceeding northward we find that in the so-called ‘“‘ Karroo”’ parts of Cape Colony the palzoliths are made of dolerite, oftener of shale indurated by the intrusion of dolerite, or of a hard chert band occurring at the top of the Dwyka shales. There also the ‘“‘knapper”’ has, where possible, made use of large pebbles or boulders rounded either by water or other natural causes. These implements are rude, and often extremely primitive. In the best examples the flaking, either through imperfect know- ledge of the craft, possibly also owing to the texture or composition of the material used, is irregular: one of the faces is often hollowed or very concave; the surface of the rock from which it was detached is often, and the rounded part of the boulder occasionally, retained. Further discoveries may reveal implements of a finer finish; hitherto only three “stations,” two in the Ceres and the other in the Cradock District of Cape Colony, have been found. The work- manship, although akin still to that of the Stellenbosch, merges, however, into the type that one meets with in the valleys of the Orange, Hartz, Vaal, Limpopo Rivers and their affluents, as well as in the Eastern Provinces of the Cape Colony, and which I propose to call the Orange River type. When we reach the Griqualand West District or its immediate vicinity, especially Prieska and Kenhardt, the difference between the above-mentioned Karroo forms and those occurring there is in some cases extreme. These palzoliths are mostly made of banded jasper, brown or yellow, and occasionally white with bluish veins— an extremely hard material which, owing to the banding, splits or flakes into small facets seemingly more readily than quartzite. The implements thus produced rival the best Acheulean flints in finish (Figs. 36, 37, 38, 39, 40, 41, 42, 44, 48, 49, &.). They do not usually attain the great size and heavy weight of many of the Stellenbosch- type examples, yet the Museum possesses one 255 mm. long, 106 mm. broad, but only 40 mm. in thickness, found near Griqua Town. At Griqua Town also was found the boucher (Fig. 35), which is made, however, of semi-translucent chert. In some examples a part of the natural banding is retained (Figs. 37a, 39a); others exhibit the trans- verse fracture of the point (Figs. 36, 40); on one face of Fig. 41 20 Annals of the South African Museum. there remains a part of the conglomerate in which it was embedded, and which contains chips of jasper, the material of which it is made. But in the Prieska, Kenhardt, Hay Districts (Cape Colony), imple- ments of quartzite or diabasic rock are found of a shape and finish rivalling those of Figs. 36, 37, 39, &. Fig. 36 is made of banded jasper ; 37 of crystalline quartzite ; 43 of diabase. The workmanship is of so superior a type, apart from the material used, that it might perhaps be ranged as a sub-type of the Orange, 7.e., the Griqua. The composition of this Orange River type is various; it is made of more or less coarse or sub-crystalline quartzite, of diabasic rock often very deeply pitted and occasionally vesicular, banded jasper, dolerite or aphanite, sometimes, but very seldom, chert (Districts of Bedford, Alice, East London, Carnarvon, Kenhardt, Prieska, Warrenton, Pniel, Vryburg, Modder River, Smithfield, Transkei, Pretoria, Witwatersrand, Potchefstroom, Vereeniging, Swaziland); jasper (banded ironstone series), quartzite, felsite, opaque vein quartz, diabase, granite, chalcedony in Southern Rhodesia (Hartley and Charter Districts, Zambesi River) ; chalce- dony also or other silicious rocks in N’Gamiland, and Bechuanaland Vryburg, Morokwen, Mafeking, &c. Near Pniel and Warrenton, on the Vaal River, the paleoliths, made either of dolerite or diabasic rock, are worn so smooth that some might, on a superficial examination, appear to have been artificially polished (Figs. 53, 55, 56, 57, 59, 60); the nearly oblong example (Fig. 43) from Calvinia in the Cape Colony is quite smooth ; but in igs. 53 and 61, the latter from Prieska, a faint contour of the chipping is retained, showing thus that the smoothness is not due to an intentional polishing of these artefacts. Some of these Vaal and Harts’ River Valley bouchers, as well as some from the Transkei, are often of a very large size and quite equal in that respect to the Stellenbosch. Generally they are of a better finish, but it must be remembered that in all likelihood only the best implements are picked by the casual collector. On the whole, and in spite of the differences mentioned, the facies or general appearance of these two, or perhaps three types is astonishingly alike, as the following examples will prove. Fig. 29 is that of a rough, massive implement made of quartzite, and was found in the Tulbagh District of the Cape Colony. A part of the original rounded surface of the boulder from which it was made has been retained. It is a very unwieldy tool, and very heavy. Fig. 30 comes from the Charter District of Southern Rhodesia. It is almost a replica of Fig. 29, and is made of impure jasper ; a part The Stone Ages of South Africa. 21 of the original surface of the matrix or nucleus has also been retained, as in the Tulbagh example, and although, owing to the composition of the rock of which it is made, it has undoubtedly required more knowledge of trimming, its resemblance to the Tulbagh implement, as well as its dimensions, are strikingly alike. If we turn to the best finished tongue-shaped or amygdaloidal implements the similarity in examples originating from localities far apart is equally suggestive. Fig. 31 is made of granite and is the first African implement of that rock found hitherto.* It comes also from Charter, in Southern Rhodesia. The right edge was broken by a waggon going over it. It is very finely grained and shows a great deal of weathering. The felspar is kaolinised. On comparison with Fig. 32, which is the reverse of Fig. 3, one of the best finished quartzite implements of the Stellenbosch type and picked up near that town, one finds it to be of the same shape, of the same type of manufacture and to have the same finish. The cleaver, Fig. 33, also from Charter, and Fig. 34 from Stellenbosch, exhibit the same resemblance, although differing in composition. These Southern Rhodesia implements were discovered by Mr. W. H. Kenny, a prospector, who found them isolated, some on the surface, others “sticking out” of the banks of a “ spruit.”’ + In that collection were examples very similar indeed to the banded jasper paleoliths of Griqua Town, but made of impure, not banded jasper. Some are small chalcedony bouchers chipped on one face only or nearly flat, and resembling greatly, in fact identical with, the Zambesi silicious implements ; but most interesting were three made of white quartz from vein, and respectively 133, 95, and 80 mm. long by 80, 68, and 66 mm. broad. The larger is opaque, the outer two crystalline; the smallest of the three is broken at one end. They are somewhat coarse, and the edges resulting from the ‘“knapping ” are very blunt, but they have probably taken as much time as, and demanded even greater skill in shaping than, the chalcedony paleolith from the Zambesi River described and figured by Mr. Henry Balfour. } * Granite implements have also been found in the neighbourhood of Toulouse, but, I believe, there only. + [here take the opportunity of thanking Mr. Kenny for allowing me not only to examine his very interesting collection, which I understand is now in England, and make notes thereon, but also to have some of the examples photographed for the purpose of this paper. He has just lately sent me a series of the same for our Collection. + «Note upon an Implement of Paleolithic Type from the Victoria Falls, Zambesi,” Journ. Anthropol. Inst., vol. xxxvi., 1906 (170-171). bo bo Annals of the South African Musewm. This palolith is 137 mm. wide and 66 mm. thick. It weighs 26 oz., and is of rude manufacture, due certainly to the nature of the rock. A part of the original shape of the nucleus has been retained at the butt, and it is trimmed on both sides.* I have seen another boucher made of the same material but smaller, and unlike any other that I have seen. It is in the shape of a transverse wedge, blunt at the top and tapering thence to a fairly sharp edge, intended probably for cutting. It is the property of Mr. F. White, of Bulawayo. Large bouchers made of chaleedony seem to be, however, very rare, but some of moderate sizes have been found; they pertain, though, more to the scraper-knife type (Pl. XX., Figs. 121, 122). Implements made of similar material have been lately found in German South-West Africa (Pl. XIX., Fig. 147), but they can in nowise be compared with bouchers. They are plainly ostrich egg- shell borers, and if they are made of the same material as that of some of the large bouchers of the Zambesi Valley, it is due to the fact that silicious rock, and perhaps no other exposed one of hard texture, was available. Together with these chalcedony implements, quartzite bouchers have been found in the so-called river gravels of the Zambesi,t but these implements must, at least provisionally, be ranged like those from Charter (Pl. IV., Figs. 30, 31, 34), among the Orange River type. * Mr. Balfour informs me that he found several examples, together with non- silicious ones. + Discussing before the Geological Society the occurrence of stone implements in the neighbourhood of Victoria Falls, Mr. T. Codrington states that he found four implements of a paleolithic type ; ‘‘ the three of brown quartzite are very like implements from India, labelled in the South African Museum as greatly resembling South African implements in material and workmanship’ (Quart. Journ. Geolog, Soc., Aug., 1909, vol. Ixv.). CHAPTER III. THE MANNER IN WHICH THE BOUCHERS WERE MANUFACTURED, AND THE TOOLS USED FOR MANUFACTURING THEM. THE MANNER IN WHICH THE BOUCHERS WERE MANUFACTURED. All the South African bouchers have been shaped by percussion, Pressure applied with bone or wood could not produce the desired object with the material of which they are made. The nucleus is of two sorts: water-worn or. naturally rounded boulders or large pebbles, and fragments artificially detached from rocks. Hither might have been heated and flakes split off by the application of cold water. The sudden contraction would not, however, produce the concave fracture of the matrix which invariably corresponds with the convex side of the detached part (Fig. 69, Pl. X.). This convexity which corresponds with the ‘bulb of percussion ’”’ of smaller imple- ments is very seldom faint; in Fig. 70, Pl. X., however, it is not very marked. I have, by good luck, been able to make observations on the site of a most interesting ‘“ station,” or deposit, extending for several miles (Simondium, Cape Colony), and have collected material illustrating most clearly, and also most abundantly, all the phases of the manufacturing process. The primitive man who made the Chellean-type bouchers had, in all likelihood, found by experience that implements are more easily obtained by the fracture of pebbles or boulders than from pieces detached from the outcrop of rock. This opinion is certainly borne out by the very numerous implements of both the Stellen- bosch and Orange River types that retain part of their original contour.” So numerous indeed are the bouchers of this type in the Draken- * In the Toulouse and Madrid deposits the same thing occurs. 24 Annals of the South African Museum. stein Valley of the Cape Colony and elsewhere, as to almost justify the belief that the makers, satisfied that the stones could be of service at that stage, gave them merely a preliminary paring until they had either the time, or had acquired the skill necessary for transforming them into that tongue- or amygdaloidal-shaped boucher which is equal to the best Acheulean, the rarity of which among roughly shaped objects seems to point to the difficulty inherent to, or the skill requisite for, their finish. A careful inspection of the numerous artificially worked, originally rounded or water-worn boulders of the Simondium “ station,” indi- cates that the ultimate shape of the boucher depended mostly on the manner of the fracture. Thus, in Fig. 68, Pl. X., a fragment has been detached from a river boulder in the manner shown. The face of the detached part is nearly even, the fracture of the left side is irregular; in the side view (Fig. 69), the convex part of the fracture corresponds with, or is akin to, the concave part of Fig. 72. On the right side of Fig. 69 the marks of blows delivered in detaching the flake are very noticeable. This fact does away with the rock-heating and cold-water-throwing hypothesis. So much, then, for the preliminary fracture. Let us now try to realise what the ultimate shape of this implement would have been, had the ‘‘knapper”’ been allowed to finish it. The pointed part is already obtained; paring the butt would probably be the next step, because a comparison with Figs. 74 and 73 shows that the reduction of the butt-end has been effected in the manner suggested before the greatest part of the original surface of Fig. 73 was chipped. The boucher would be greatly reduced in size by this operation, but its utility as a hand-pick would have been increased rather than impaired by the smooth, even, original surface being retained on one side instead of being facetted as on the other. In the Simondium “ station ” these “ bouchers”’ pared on one side only were, as I have already remarked, very numerous, but together with them I found there also some of our best finished examples (Fig. 5, Pl. I.). Fig. 20, Pl. II., comes also from the same spot. In the Eerste River Valley I met with the same experience. Great importance attaches to the fragment Fig. 68 (face view), 69 (side view), Pl. X., because, unlike what is alleged for European bouchers, and other implements of the Mousterian type, the upper face has not been fashioned first and then detached. This conclusion is borne out by most implements of that kind which I examined. The Stone Ages of South Africa. 25 It is true that in Fig. 70, Pl. X., the preparatory stage bears no trace of the blows that detached it from the matrix. The fracture is almost vertical, and the strongly convex outer side is artificially pared ; but this is probably due to the difficulty experienced in detaching the same from the nucleus. Owing to its shape, it would prove almost an impossibility to turn this first stage (Fig. 70) into either a tongue-shaped or amygdaloidal boucher. With it I found, however, several examples having the shape of Fig. 71—a fact which shows that only in this form the artefact shown in Fig. 70 could have ultimately resulted, unless extreme skill had been applied to turn it into an amygdaloid boucher of very much smaller size; but as Fig. 71 has been used in its present shape, our conclusion is justified. In Fig. 63 we have a split boulder, which leads to the evolution of either the axe or cleaving tool (Fig. 66) or the spade-like pick (Fig. 67). Fig. 64 is that of a water-worn boulder, bearing on the left and right sides traces of ‘‘ knapping,”’ the beginning of its trans- formation into either the broad-end boucher (Fig. 65) or the pointed one (Fig. 62); Fig. 63 is a good illustration of the preliminary trimming that will result in the ultimate production of Fig. 62. The examples I give here are all taken from one locality, 7.e., Simondium, but I could have quoted as effectually from several other places. In the chapter dealing with the purpose for which bouchers were made I allude to the great size and correspondingly heavy weight of some of the South African implements. Fig. 20a is that of the largest and heaviest I have as yet met with, and the point plainly bears marks of usage. Many were the speculations which its shape and its partly unfinished appearance and heaviness suggested ; but if the workmanship of the fine implement shown in Fig. 20, Pl. III., is carefully compared with that of Fig. 20a, it becomes clear that the latter would have ultimately been turned into Fig. 20. All the facets, although reduced in size in proportion to that of the finished implement, are traceable: the concave depression on the left side of Fig. 20 is the reduced one of Fig. 20a; the median ridge of the apex is discernible in both; the flat part on the right side of Fig. 20a is still indicated in Fig. 20, &c.; and the same thing happens on the reverse of both. The trimming of Fig. 20a into the shape of Fig. 20 would, however, cause a considerable diminution in size and weight before the esthetic sense of the maker would be satisfied with the ultimate finish of the implement into the characteristic shape of Figs. 1, 2, 3, 4, 16, 17, 26 Annals of the South African Musewm. &e. Only large boulders would be selected for the first stage of an implement like that of Fig. 20a, and such large fragmented ones are not uncommon. One that I remember especially, because it was of such a great weight that I could not carry it with me, was oblong, 450 mm. in length, and clean cleft longitudinally. In the same manner as the sharp-pointed bouchers so would the cleaving implements be produced, only that they would in all likeli- hood be fashioned more easily. Both faces need not be knapped to procure a serviceable weapon in the shape of an effective cleaver. From the detaching blow there has resulted a smooth, slanting shape on one side (Figs. 47, 50, 53, 57), while the other face bad to be similarly treated to obtain a corresponding slanting side in order to make the cleaver effective. Sometimes the natural face is retained (Fig. 51). It would seem at first sight that a cleaver could be more easily produced than an amygdaloidal hand pick, and it probably was. Yet Figs. 48, 49, 56, &., denote a great deal of skill in the making. They are worked on both sides, and so are Figs. 33 and 34— the first from Stellenbosch, the second from Rhodesia. Perhaps they were intended both for cleaving and digging, but as cleavers they would certainly prove less serviceable than Figs. 47, 53, and 57, which have undoubtedly been produced with less skill. The ex- planation is that the cleaving propensity is due, as in the case of the digging-tools, to the preliminary trimming on which the shape depended. It could not be corrected without much trouble or dex- terity of hand. It must be alsoremembered that these cleaving tools ave fewin proportion to the number of digging bouchers. One would expect that, this mode of manufacture being the easier, the propor- tion would be reversed, and also that the style or type would be older were it not that the Mousterian (and most of the axes are Mousterian in type) has in the palearctic regions been preceded by the Acheulean and the latter by the Chellean. It does not follow, however, that such has been the case in South Africa, but the comparative scarcity of these cleaving-, in contradistinetion to the digging-tools, is worth noticing. THE TOOLS USED FOR MANUFACTURING THE BOUCHERS. When the large palxoliths are made from a rounded boulder or large pebble, water-worn or otherwise, it is probable that a boulder of the same size, or perhaps heavier, was hurled against the one which it was intended to split in order to obtain by concussion the preliminary frag- The Stone Ages of South Africa. 27 ments from the nuclei. The impact would, in many cases, result in the partial cleavage of both boulders, and of this there are manifest and numerous proofs in the Stellenbosch and Drakenstein deposits, where rounded boulders seem to have been exclusively used. The skill evinced in some of the rougher examples, as seen in Figs. 7, 8, 12, 15, 21 is, however, of such poor description that I am probably justified in postulating that paring tools other than any fragment or splinter of rock were not required for the purpose. But together with these roughly fabricated artefacts there are found, in situations pointing clearly to the same origin and to the same epoch, bouchers highly finished, although broadly facetted, and fully worked into a point (Figs. 1 to 4, &c., Pl. I.). For fashioning a split pebble into an amygdaloidal- or tongue-shaped implement, flaking and paring tools were necessary, and were evidently used; but so far, in the deposits of the two localities already mentioned (and I consider, for reasons to be given hereafter, that they are among the most ancient in South Africa), no evidence of these instruments has as yet been obtained. Nothing has been found to my knowledge resembling even the tools that might have served this purpose (cf. Pl. XI., Figs. 86-90), and which are known from other localities where bouchers have been found, as well as from places where they have not been met with. All bouchers, however, have not been evolved from rounded boulders ; some are made from fragments detached from large rock masses. To obtain the material required, correspondingly large hammers, which I may term detaching hammers, in the shape of Fig. 75 of Pl. XI., were used. Artefacts of this kind are now recorded from the neighbourhood of Cape Town, but I am by no means certain that the epoch when the detaching hammers were in vogue coincides with that of the Stellenbosch and Drakenstein deposits. There are occasionally found nuclei of the same shape as the detaching hammers, but with edges so numerous and sharp that it is plain that they are the residual nuclei or cores from which chips were detached, and not hammers; were it otherwise the edges would be abraded, which is not the case. Quite lately there were discovered at Fishhook, Cape Colony, on the slope of a hill usually covered by a huge sand-dune, which had been temporarily removed by abnormally strong winds, quite a number of these extremely large detaching hammers plainly bearing marks of the use to which they had been put. They were associated with river-boulder paleoliths of a most ancient type, as well as with others that had been fashioned from non-rolled pebbles, but of equally ancient appearance. On a part of the wholly exposed floor those 28 Annals of the South African Museum. detaching hammers had been deposited by gravitation, so as to form here and there irregular circles. The one shown in Fig. 75 of Pl. XI. is only of moderate size; some of those met with were more than twice as large. Were it not that the edges are so abraded, it might be possible to imagine that these stones had been cores as already stated, or perhaps missiles, in spite, or may be on account, of their weight, but together with these artefacts we found the fragment represented in Pl. XI., Fig. 84. Made of the same crystalline quartzite as the detaching hammers, a quartzite not found 7 szti, it shows plainly that it was detached from a rock surface by a very large tool of the style of those mentioned as detaching hammer. The reverse is flat and the fracture clean. The obverse is already pared on the upper side by the repeated blows of the detaching tools, somewhat in the manner claimed for the Mousterian of Europe. On the Cape Flats, in the neigibourhood of Cape Town, there used to be exposed here and there outcrops of a close-grained cherty sandstone, or surface quartzite, and the grains of which, even when not closely set, are cemented together by silicious matter. Some of these outcrops were covered with large bosses and depressions due to the removal of large or moderate-sized fragments by percussion. Fragments of the same rock considerably too large for paring or trimming the somewhat delicate, small-sized tools found there in abundance were not uncommon in the neighbourhood of these out- crops, and they may also be considered as detaching hammers, but none of these seen by me was equal in size to those found at Fishhook. That the detaching hammers, possibly hurling stones as well, made of this surface quartzite were also used as cores seems to be proven by the find, also at Fishhook, of similar examples, together with large flaking tools, of the same material, one of which is figured in Plate XI., Fig. 90. But as will be seen later on, the implements made of such material are comparatively recent. ( 29 ) CEeNe iii Ne WHAT WERE THE BOUCHERS USED FOR ? Let us endeavour now to realise the purpose which these artificially worked stones, which we are now terming ‘ bouchers,” were in- tended to serve. Primitive man did not use his uncouth tools for exclusively one purpose. Had he done so, he would not have been primitive ; but it is generally assumed that the bouchers of large, or moderately large size, were held in the hand, and nof hafted. This admission holds good for the South African ones, and for reasons which have not, to my knowledge, been adduced for the European, Algerian, or others, and which I give here. 1. In all the examples of these artificially worked stones—be they of the more or less perfect tongue- or amygdaloidal-shape, as well as in many, if not most, of the cleaving tools of the Stellenbosch type— whenever they are worked on both sides, the sharp, median, uneven ridge of one face never corresponds with that of the other. The stability of a boucher inserted in a hollowed- or cleft-handle of horn or wood would be greatly impaired thereby in spite of thong attach- ments, or the use of gum-cement in the manner obtaining among the Australian aborigines—a manner which did also obtain in South Africa, but for smaller tools only, as will be seen hereafter. Yet the thing is not impossible with bouchers of the Mousterian type, one face of which is the result of the cleavage, while the other is pared with the resulting median longitudinal ridge, because in these the flat face would add firmness to the hafting. This applies especially to the cleaving implements, but whereas the axes, Figs. 33, 34, 54, 55, 56, could not have been hafted easily, the same cannot be said of Figs. 47, 48, 49, 50, 51, 52, 53,57, which represent cleavers or hand-wedges belonging to both the Stellenbosch and Orange River types. 30 Annals of the South African Museum. One thing, however, is noticeable in these wedge-shaped tools, namely, that with extremely few exceptions, resulting from the mode of cleavage, the most bulging part, as shown in the profile of the figures here given, is the centre. Had these been hafted, a forcible impact would undoubtedly tend to dislodge them from the socket. Another reason for assuming that the bouchers worked round the butt in a tongue-, or amygdaloidal-shape, were not hafted is that the edges are too sinuous to produce a clean cut (see Figs. 1, 2, 3, 4, 11, 29). 2. The bouchers in which the original rounded part of the boulder has been retained (Figs. 5 and 22), might have been used as hafted club-heads, the sharp end being inserted in a handle. But had this been done, this rounded part would show traces of abrasion. So far, however, only one such boucher from among the many which I examined has shown traces of wear at the rounded part, as if it had been used for pounding or smashing a hard object. Moreover, in many of the examples of this type, the sharply pointed apex shows signs of retrimming, and this leads me to conclude that the sharp end and not the butt was used. — That the same occurred also in the case of the tongue-, or amygdaloidal-shaped bouchers is proved by Figs. 16, 17, 18, 19, in which the transversely broken part would need fresh trimming to become serviceable again. Had they been hafted, that is to say had the sharp point been inserted first into a socket, the fracture would not have taken place where it did from the force of a blow, but more towards the middle, as happens for the Solutrian type lance-, or javelin-beads, of the Cape Flats (Figs. 110, 112, No. 1, 9, 3, 5, Pl. XIV.). The bouchers truncated in this manner thus give the impression that they broke at their weakest point while being used as picking or digging tools. Figs. 5 and 6, Pl. I., represent hand-picks in which the rounded part of the nucleus has been retained as best suited for hold- ing in the hand. It is wonderful how, even in an inexperienced hand, adaptation to this primitive instrument is demonstrable. Fig. 9 is that of an implement of the same type, but with a much broader end, specially suited for spade-work. It is too large, how- ever, and too heavy (it weighs over 5 lbs.), not to have necessitated the use of both hands in using it. The same may be said of Fig. 20, the point of which shows very distinctly that in spite of its weight (9 lbs. 3 oz.) and size (33 cm. long by nearly 16 broad), it has been used as a digging instrument. Fig. 10 is especially instructive ; no The Stone Ages of South Africa. 31 longer amygdaloidal- or tongue-shaped, it is plainly curved. A point was what the maker wanted; this being obtained, he left untrimmed the remainder of the nucleus from which he fashioned it. This boucher weighs 3? lbs. In the same locality where I found it, I discovered several examples having the same peculiar bend. In Fig. 7 the apex is, in proportion to its size, as broad as in Fig. 9, and it is not easy to contend that this somewhat uncouth boucher could have been put to any other purpose than the one to which the indented point bears corroborative testimony, 7.e., digging. More- over, the points of Figs. 5, 7, 9, 10 show marks of retrimming. Fig. 22 is the representation of a quadrangular implement somewhat unique in shape. Had its purpose been that of a hammer, pounder, or club-head, the butt or edges would show signs of wear, which they do not in the least, but the point has been retrimmed ; moreover, in this particular case, the four sharp edges preclude the possibility of its being held in the hand as a club, except possibly in a case of emergency, but as a digging tool it would be most serviceable. In most of the examples pointed at each end, one of the ends shows more service than the other (Figs. 13, 15). Oblong or ovoid implements, of which it is difficult to say which end is the butt or point, are extremely rare among the Stellen- bosch type. Fig. 24 is an example. It is too large to have been a scraper, its edges are too sinuous to permit of it having been a cleaving tool. Moreover it is, like Fig. 18, made of a surface quartzite differing from the material used for the Stellenbosch type, and is certainly less ancient. But in the deposits of the Orange River type, especially the Griqua sub-type, these ovoid implements are occasionally found. As already mentioned, the striking characteristic of the well- executed South African paleoliths is their huge size. We have in the collection, apart from these already quoted, many specimens weighing from 24 lbs. to 6 lbs., and the average of a boucher 20 cm. long is a little less than 3 lbs. This great weight is also a strong argument against the likelihood of their having been hafted. They would have proved very unwieldy even to men of powerful physique. But it does not follow that we have not in South Africa bouchers of moderate and also small dimensions, and comparable in that respect with those obtained elsewhere. Figs. 4, 18, 19, are good examples of some, although they are rare, but I know of others which are smaller still, and of as perfect a type as the best Acheu- 32 Annals of the South African Museum. lean (Figs. 18, 76, 77, 79). Some were found in association with larger implements, others by themselves. It is not unreasonable to suppose that they were made either by,. or for the children, and possibly also for the women. The singular implement, Fig. 23, Pl. III., is of a type that has been discovered hitherto in the Berg River Valley (Cape Colony) only. Three are known from Wellington, and I lately discovered two more at Simondium, where there is what I consider to be the oldest station yet found, lying together with some of the best finished and largest paleoliths it has been my good fortune to discover, as well as with still more numerous unfinished ones, I think that the explanation corroborated by the figures here given makes it quite clear that the makers of these paleoliths fashioned at the same time two kinds of bouchers: a pick more or less sharply pointed at one end, and a cleaver or hand-wedge more or less broad at the cutting edge, both edge or point of which, in nearly all cases, Show conspicuous marks of wear. The hand-picks do not, however, imply agricultural pursuits in the sense of crop-growing. Man of the Middle Pleistocene was small, about five feet in height. It is hardly probabte that he trusted either singly or collectively to his strength alone to attack or repel the fere nature that threatened his existence, or to capture those that were necessary for his suste- nance. Probably armed with a heavy club, he entrapped the game required for his food and clothing, and this he did by snares or pits. If the original maker of these South African bouchers, which are almost identical with the European, is, as I really believe, the ancestor or descendant of the negroid race that left traces of its industry and culture in Southern and perhaps also in Central Europe, then the use of these implements, whether made of quartzite or of flint, is now explained. Hither he imported into Europe the methods to which he has long been accustomed in Africa for trapping and securing the produce of the chase; or, if he is not the native of this country, he brought to Africa, during his long peregrinations, to the further progress of which the Ultima Thule of the South put an insuperable obstacle, methods borrowed from those people whom he has encountered and from whom he has borrowed. We know that until lately drives, leading to pits and trenches garnished with pointed stakes at the bottom, were used in South Africa for securing game. In the Humansdorp District of the Cape The Stone Ages of South Africa. 33 Colony there is a narrow gorge in which game-pits and stakes in very good state of preservation are still to be found.* This is clearly a case of survival of methods. Primitive man, or the early South African aboriginal, lived on game, followed game, and entrapped and snared it in the same manner as the present aboriginals did until a few years ago. But unacquainted with the use or making of iron, as he undoubtedly was, how could the pits be dug but with the-stone picks or spades ; how were the stakes cut and sharpened for impaling the game at the bottom of the trenches, or for palisading the enclosures for the drives, but with the stone picks or stone axes mentioned? And as for the hypothesis that his weapon was a club, the survival of type seems to me to be also borne out by the discovery of such an implement in a rock shelter that had been partially filled with bat’s dung (cf. Pl. XIX., Fig. 152). Made of an extremely hard and heavy wood (Olea sp.) it is plainly fashioned with stone tools (scraper-knives). It is shaped as a phallus, the handle has been trimmed so as to make prehension by a small hand more effective. With it was found a stone bead (PI. XVI., Cut 3 of Fig. 186); also a small cube of iron pyrites. At the entrance of the shelter there are still traces of bush paintings, and it is not out of place to remind the reader that in he gorge at Humansdorp, where the game traps are preserved, there occurs a rock engraving, painted with red ochre.+ Let us assume that primitive man originated in Africa. When he invades Europe in the Chellean times, the climate is attractive ; he brings with him his primitive weapons, the weapons of the chase, defence or offence. Are the fere natwre which he has to encounter such formidable and unknown beasts as to daunt his courage? Cer- tainly not. Hyena spelea he knows well, it is the present H. crocuta, found only in South and Central Africa; Hyena brunnea, occurring now from Senegal to South Africa, he also knows well. The tooth- sabred-tiger or the cavern lion could have for him no more terror than his old acquaintance Felis leo or Felis pardus, the present lion and leopard which, besides, he meets again in that country new to him. Is he frightened by Hlephas antiquus ? No, it is his old acquaint- ance, now called H. africanus. Hippopotamus major is his old friend H. amphibius ; Rhinoceros mercki he cannot distinguish from * Péringuey, ‘‘ Rock Engravings of Animals, &c.,’’ Trans. §.A. Phil. Soc., xviii., 1909, p. 417. + But these facts are instanced here merely as cases of survival. It does not in the least follow—in fact, it is to me, at least, certain that neither the club nor the bead here mentioned haye any connection whatever with the Chellean-Mousterian boucher industry. 3 34 Annals of the South African Museum. f. sumus, R. bicornis, or R. keitloa. No such niceties in identifica- tion for him. He either defends himself against them, or uses his growing cunning in mastering them, especially the formidable cavern bear Ursus speleus, which he has not met before. He finds no longer the numerous antelopes of his acquaintance it is true, but Bos bison has the same attraction for him who has slain Bubalus baini or B. anti- quus. It is quite possible that he has not known these denizens of an intensely cold climate, the woolly rhinoceros, the mammoth, the reindeer. He would follow the animals which he knew, beasts driven back by cold to receding warmer climes—to climes where, as in South Africa, the total absence of traces of pleistocene ice-age clearly proves that there did not exist at the time the increasing rigour of the elements that has come to prevail in the country whence he retreats, either following the migration of the game on which he subsists or migrating to where it is found still. And if he is not of African origin, if he is of the Neanderthal- Chapelle race, but, unlike the latter, has not been able to ac- commodate himself to the new climatic conditions, then in his retreat southward, and especially to the African continent, ne probably accompanies or comes across there most of his old ac- quaintances ; if not all, many of them, @.e., the hippopotamus, the elephant, the hyenas. He finds himself among antelopes which he did not know, but horses which he knew. The hyenas follow him, for is he not providing crumbs for them? He continues the application of methods which he has perfected elsewhere. In his emigration southwards, where he no longer finds the flint nodule so easily worked into implements, he resorts to any stone hard enough to ensure its object ; hence the use of quartzite, hence also the discrepancy in technique, more apparent than real, since ulti- mately the “‘ knapping’”’ becomes as perfect as that of the best flint. But the primary use for which the new material implements are made is the same. They are intended to be used as picks or spades for digging trenches, cleavers to cut stakes or palisades, and, as will be seen in the next chapter, the manufacture of other tools for domestic use here accompanies or follows that of the bouchers, if it has not preceded it. ( 35 ) CHAPTER V. WERE ToOLS OTHER THAN BoucHERS MANUFACTURED BY THE MAKERS OF PALZOLITHS ? It remains a moot question whether or not the Chellean boucher, which so many of ours resemble, was the only manufactured im- plement of that period, leaving out of account the flakes or chips that resulted from the paring of the same—the by-product, so to say, of this lithic industry; a by-product which is very rarely associated both in Europe, and in many cases here also, with the finished or unfinished tool. It is, however, safe to conclude that in the case of the Chellean era the bi-facial boucher was the predominating, but not the exclusive tool. On the other hand, the Acheulean type is often found together with the Mousterian, and with the latter’s concomitants of ‘ points,” “‘scrapers,”’ ‘‘ graving tools,” &c. In the chapter dealing with the process of manufacture of the South African paleoliths I have gone into details which seem to show that on the shape of the first fracture of the pebble the ultimate object, cleaver or pick, probably depended. But although the forcible impact of a boulder, or of alarge pebble against another, was the pre- liminary step for fashioning the tool, smaller stones more appropriate to the purpose, such as minor fabricating tools, would be required to give, when desired, to a boucher of the type of Figs. 1 to 4, PL. I, these finishing touches, such as facet-flaking or edge-paring, which so clearly imply that a sense of esthetics prevailed over that of mere utility with some at least of the primitive makers. FLAKERS, PARING TOOLS. The two first-mentioned tools are seldom found in South Africa with the completed objects. Finishing tools they may be called where the Chellean- or Acheulean-type bouchers are concerned, fabricators where the comparatively thin, lanciform, lamellate scraper 36 Annals of the South African Musewm. knives are produced, Figs. 86 to 92, Pl. XI., give a general aspect of their shape. Figs. 87 and 88 are tools proved to have been used for facetting and edge-trimming by the situation in which they were discovered. Made of ferruginous banded jasper, Fig. 89 was found together with highly finished bouchers and flakes of the banded jasper of the Griqua sub-type, and so are others which, made of different rock, have the same general appearance. The Tywmi Deposit—Fig. 87 deserves more than a passing mention. It was found with three of four similar ones embedded deeply in the banks of the river Tyumi, in the Victoria East District, Cape Colony.* With it was found a number of implements made of the same material (banded jasper), among which a long, oblong pick of the style of Fig. 43, but narrower, as well as pieces of the most finished Acheulean type I have as yet met with here (Figs. 76, 77, 85). There were also knife-scrapers (Cut 4 of Fig. 131), nearly as well finished as the implements not uncommonly found in the neighbourhood of Cape Town. Others were of the type and size of those of Fig. 105 (from Nooitgedacht), one, especially, being almost the replica of Cut 7. There were also found lydianite scrapers, one of which (Fig. 124) shows the same workmanship as the bottle- glass example (Fig. 129, Pl. XVI.). No other conclusion can be come to than that all these implements, bouchers, flakers, scrapers, rude or well finished, uniting as they do most of the forms met with in South Africa, except the typically crude, large bouchers of the Chellean type, are coeval, and that they have been manufactured by the same people. They are also evidence of an advance in culture, especially in so far as the digging tools are concerned ; but it must not be forgotten that equally finished ones made of a less easily worked material are also known from deposits that are undoubtedly older (cf. Figs. 1, 2, 4, &c.). At any rate, it is proved in this case that the makers of the bouchers did at the same time manufacture implements superior in technique to the mere scraper-knife. But no flaking tool having some resemblance to LETS ESS WE SI) Pl. XII., was discovered in this Tyumi River deposit. These laminate, more or less lanceolate tools, which probably served also, at a push, as knives or scrapers, in spite of their irregu- larly dented edges, are met with lying on the surface or under a depth of soil; they are found solitary, but oftener than not they * This deposit was discovered by Mr. A. Johns, who presented several examples to the Museum. The district is perhaps better known as ‘‘ Alice.” The Stone Ages of South Africa. 37 occur together with other and smaller products of a lithic industry of a doubtful date or period. As far as my experience goes, they have seldom been found in a situation which incontrovertibly proves them to have been associated with bouchers of the old-fashioned Chellean type. They all bear marks of use. Some are so weathered that the faces are partly disintegrated. Fig. 102, Pl. XII., is from East London, Cape Colony; Fig. 103 from Douglas, also in the Cape Colony— localities far remote. They are each made from an igneous rock that would, from its texture, become more pitted after an equally prolonged exposure than, for instance, the lydianite tools (Figs. 88, 91, 92, 97), but the abrading process in these two examples, although probably due to dissimilar actions, has resulted in the same worn appearance. Fig. 102 is from the banks of the Buffalo River, Hast London (Cape Colony), and is one of the few for which a connection with a river terrace may perhaps be claimed. The under surface has become quite granulate ; the upper is considerably pitted. Fig. 103, found under a shallow layer of soil, if not actually on the open, is even more worn or pitted than Fig. 102. We possess from the cave-shelters of Knysna (Rob Berg) a lanci- form implement of quartzite, exhibiting traces of weathering equal to that of these two examples. Worked on one side only, and always without any traces of secondary trimming, most of the lamellate tools were originally flakes produced in the shaping of a large tool, as the convex under- face shows; and it is most probably by means of these flakes turned into flakers that the spokeshave-like scraper-knives seen in Figs. 88, 89, 100 were detached from the nuclei. This curious bend or curve plainly indicates that the flakes were obtained by percussion from an already convex or naturally rounded boulder. But, whereas the very great age of flakers similar to Figs. 102 and 103 cannot for a moment be doubted, while that of Figs. 91, 92, 93, 97, owing to their well-preserved appearance, remains doubtful, the same cannot be said of Figs. 90 or 91, Pl. XI. Made of a surface quartzite that seems to be restricted to the western part of the Cape Colony, from Cape Town to Namaqualand, they are con- nected with implements of small or moderate size, some of which are the nearest approach to a Solutrian type as yet discovered in South Africa (Figs. 110 to 111, Pl. XIV.), and with others which are of very recent date, comparatively, as will be explained. Yet, the primitiveness of these fabricators is about equal to that 38 Annals of the South African Musewn. of similar tools for which a degree or character of great antiquity may justly be claimed. One thing, however, we may consider as established. It was by means of flaking or paring tools of the kind here figured, whether very ancient or comparatively modern, that the cutting, scraping, even sawing tools, with edges either worn or as sharp as if they were made yesterday, were obtained or produced. With extremely rare exceptions, and these possibly accidental, they show at the base, on the reverse side, a conchoidal fracture so bulging, or they are so incurved as to preclude even in the most lanciform examples the possibility of their having been firmly hafted. We must regard most of them also as having been detached from the nuclei merely for obtaining a cutting tool, a tool that could be discarded at will—one that often served a momentary want, and on which much care would not be bestowed.* Their extreme abundance seems to justify this hypothesis. The Noottgedacht Deposit.—Although standing alone, the Tyumi River deposit shows so clearly the connection of small or mode- rately sized scraper-knives, with and without secondary trimming, with bouchers of large and moderate sizes, but of an advanced type, that their contemporaneity cannot be challenged. We owe, however, to Miss Wilman, of the Kimberley Museum, a discovery of a similar nature as that of the Tyumi, only that the scraper-knives are very primitive and the bouchers still more so. This lady investigated in the dry diamond diggings of Griqualand West, at a place called Nooitgedacht, on the Vaal River, a deposit containing a great number of bouchers of both the digging or cleaving forms. Some are extremely large, others are of more than moderate size, but all are so worn and polished by water (no other agency could have produced such a smooth surface) that they have returned in many cases to their original state of pebbles. They very much resemble the implements in Pl. IX.,, Figs. 57 to 61, but are still more polished, or in a more deformed condition. Intermingled in great quantity, and worn as smooth as the * Mr. Redmond Orpen told me some years ago how, when in company with his uncle, C. Sirr Orpen, a buck was shot in Griqualand West, but when they came to disembowel it, so as to make it lighter to carry to the waggon, they found that no one had a knife. Undaunted by the untoward circumstance, the native who accompanied them, and whose race is doubtful, knocked together two pieces of stone, and produced from them an implement sharp enough to cut the animal open. This anecdote is also found in Stow’s ‘‘ Native Races of South Africa,” but I had it betore the publication of that book. The Stone Ages of South Africa. 39 bouchers themselves, were found scraper-knives, from which I have selected for illustration seven examples (Fig. 105, Pl. XII). It will be seen from these figures that the outline corresponds with that of almost all implements of this nature which one meets with in South Africa. Many of these flakes or scraper-knives are too small to have been the original splinters detached in trimming the bouchers, judging from the evidence afforded by the Stellenbosch type, and they must thus have been shaped intentionally, probably from a small nucleus ; they were, therefore, not mere accessories. The great age of this Nooitgedacht deposit is not, however, proved, but in what I consider to be the oldest, 7.e., Cape, Draken- stein, Stellenbosch, these scraper-knives are indeed very rare, yet I know of some that seem to have been trimmed intentionally into shape; such aoneis Fig. 93, Pl. XII. This lanciform implement could not, as in the majority of cases, have been hafted on account of the bulging butt-end. The intention of paring this flake into a more serviceable tool than the mere accessory flakes (Figs. 94 and 95), found in the same locality, is apparent. At Simondium I found a scraper (Fig. 96), the shape of which does not appear to be accidental; the bulb of concussion is very strong. I have also seen from there two smaller ones shaped alike but not connected with the large bouchers, and belonging to the more recent type. Near the mouth of the river Nahoon, not far from East London, Mr. J. Wood has found and sent me some very large, crude bouchers made of different Karroo rocks. They were found exposed, to- gether with a very large number of chips of the usual flat, flake- form. But the absence or rarity of these lanciform blades, flaking or paring hammers, or of scraper-knives among the worked or partly worked bouchers, wherever the deposits are found on the talus of mountains or on the mamelons left in the erosion of valleys, does not in the least imply that bouchers alone were manu- factured in the original situation in which they were made. Their absence is explained by the removal of the flakes, owing to their lighter weight, occasioned by the same denuding agents to which, however, the very much greater size and bulk of the bouchers, cores, or detaching hammers offered a greater resist- ance—a physical resistance that made their downward progress very slow in comparison with that of the flakes. But they would ultimately reach either a flat terrace which greatly impeded their downward course, as at Bosman’s Crossing, or a pot-hole, as seems 40 Annals of the South African Museum. to have been the case at Nooitgedacht; and thus all these imple- ments, large or small, would be found lying pell-mell. IMPLEMENTS OTHER THAN FLAKES OR SCRAPER-KNIVES. In South Africa, however, we have now proofs that the lithic industry was not confined solely to the production of digging or cleaving tools and their accessories, the flakes that ultimately became scraper-knives. In these deposits or “statons’’ one meets occasionally with flaked stones facetted in the manner of Figs. 78 and 79, Pl. XI. These might at first sight be taken for nuclei from which short, narrow flakes had been detached by percussion in the manner of Fig. 139, Pl. XVITI., in which case the makers knew already how to utilise the ‘‘ pigmy ” tools as burins, or ostrich egg-shell beads parers, perhaps also for heading arrows. But the character of the Table Mountain sandstone, or other quartzite rock of which these pseudo nuclei consist, precludes the possibility of these splinters having been used for that purpose ; the texture is not sufficiently fine-grained or compact for obtaining a very sharp edge, nor would it be resistant enough. These stones might, on the other hand, be taken for small hammers used for retouching or retrimming, but there is no sign of secondary trimming of the kind that could be produced by them on the implements with which these nuclei-like artefacts are associated, and, moreover, the edges of the fractures are always extremely sharp, which would not be possible had these come into forcible contact with another stone of the same material as them- selves. Most likely these stones were missiles for hand-throwing, and possibly, also, sling stones. They vary in size, but none is as large as the detaching hammers which they resemble except for their irregularly polygonai shape. The evolution of the hand-throwing stone into the sling-stone is easily conceivable; both were intended as missiles, primarily for defence, and probably at a later stage for attack. In addition to these polygonal or core-like implements, plainly artefacts, there have been found, especially in the south-western deposits of the Cape, but also in Vereeniging, Prieska, and the ‘‘dry diggings” of the Vaal River, rounded, seemingly water- worn pebbles, which one hesitates to say are artefacts, whereas in others the doubt is not possible. These latter have one or more smooth, flat, or slightly convex facets, the abraded planes of which, when the stone is multi-facetted, never correspond with those of another (Figs. 80 to 83, Pl. XI.). These facets are, The Stone Ages of South Africa. 41 moreover, so small that they preclude the possibility of the stones having been used as pounders or mullers after the manner of those of the recent period, such as Figs. 168 and 170, Pl. XXIII. No mortar or quern has been as yet found in these deposits. There is also no justification for seeing in them the precursors of the rounded or flat stones perforated in the centre and called “! kwé” (Figs. 152 to 160, Pls. XX.-XXI.). Doubtless these rounded pebbles (Figs. 80, 81, 82, 83, Pl. XI.) owe their shape wholly or in part to physical agents, such as the displacement and friction of boulders caused by the torrential river-flows that, in the majority of cases, form in South Africa the short rivers or streams which, well-nigh dry most of the time, become impetuous torrents at others. One such pebble, tri-facetted, I found in the temporarily dry bed of the Eerste River (Cape Colony) among large water-worn boulders which in time of flood are hurled against each other with such force as to produce a deep rumbling noise heard from afar, in the same manner as in many of the Pyrenean “ gaves.” This action would easily suffice to grind facets into a small, already rounded pebble falling between revolving masses of considerable size. But it is, indeed, seldom in such a situation that these rounded stones are found. They are met with in the Drakenstein, Paarl, Stellenbosch, and neighbouring valleys, and invariably with bouchers of large size, on the hills or talus slopes, often at great heights. It is probable that if no more have been recorded outside the region of Table Mountain Sandstone, and of the Vaal River deposits, it is because they were considered to be natural products, not artefacts. While some of them are fairly well preserved and patinated (Figs. 81, 82, 83, Pl. XI.), others are partially disintegrated, and this, curiously enough, round the facets. ; Fig. 80 is peculiarly instructive in that respect. At the two extremities are two facets not quite concordant, and of the size of a shilling and a sixpenny-piece- respectively; both retain the glaze or patina of the original surface, while the rest of the periphery is disintegrated to the depth of from 2 to 3 mm. This example I found at Simondium at a height of 300 feet above the level of the river draining the valley, which is some 3 miles away as the crow flies. It was lying among huge bouchers, all showing more disintegration than any I have as yet met with.* This partial * At the time of my visit these implements had just been dug out in making a plantation. The constituents of the soil in which they were embedded, entirely void of lime and containing very little potash, could not have acted as potent 42 Annals of the South African Musewm. disintegration is not restricted to that one rounded stone, for we have in the collection others showing more or less evidence of a great age. In some of the best preserved the shape of the facets would seem to imply that these stones have been used as small grinders or mullers (Figs. 82, 83), only that the facets are slightly concave, or deeper in the centre than at the edges, and differ in this respect from the neolithic mullers figured in Pl. XXIII. There is no reason for disbelieving that the primitive people who made the paleoliths availed themselves of these objects primarily fashioned partially, if not completely, by natural agencies with a view either to improve on them (cf. Fig. 23, Pl. II1.), or of making use of them as they were. The missile hypothesis seems to be strengthened by a recent find in one of the numerous Stellenbosch- valley deposits of an almost orbicular quartzite stone of moderate size, with a very small depression on one side either artificial or more pro- bably left from the original contour, and of another larger spherical one made from milky-vein quartz bearing unmistakable traces of artificial working. I discovered similar ones several years ago, but I rejected them at the time as being doubtful artefacts.” These rounded or polygonal stones are not restricted, as I have already stated, to the Cape, Stellenbosch, and Drakenstein deposits. In those of the Vaal River, rounded stones of a similar facies but not facetted are found. They are often difficult to distinguish from naturally rounded ones. A small specimen in our Collection, from Waldeck’s Plant on the Vaal River diggings, bears, however, marks of usage in the shape of a small artificial dent, or rather depression. At Vereeniging (Transvaal) these implements are of the same type as those from Stellenbosch or Drakenstein, and are made of quartzite or dolerite. In some, the facets correspond more or less on each side; the edges of the depressions or facets are usually sharp, while in other examples they are smoothed and partly obliterated. + But to whatever use these rounded stones may have been put, factors in the disintegration of sandstone implements all plainly made from rounded water-worn boulders, selected because their texture was denser or more compact. This lack of potentiality in the chemical agents of the soil leads to the conclusion that this partial disintegration is due to an extremely ancient age. * T found no less than five such rounded stones in a clay-pit of great depth, and only half a mile from the spot where the two examples here mentioned were dug out. + Lam ereatly indebted to Mr. T. N. Leslie, of Vereeniging, who discovered that deposit, for loaning me the typical specimens for identification, comparison, and reproduction. Mr. J. P. Johnson subsequently described and figured some of these implements (Trans. S. Afric. Phil. Soc., xvi., pt. 2, 1905). The Stone Ages of South Africa, 43 and that of brayer is quite likely, one important point is now estab- lished, namely, that in parts of South Africa, very far distant from each other, balls of sandstone, quartz, or dolerite have been found together with bouchers of undoubtedly palzolithic shape. The importance of their occurrence here is enhanced by the fact that in the “Grotte de l’Ours,” in France, balls of sandstone, of caleareous stone, and of flint have been found together with Mousterian domestic utensils (mobilier), which include amygdaloidal Acheulean bouchers, Mousterian points, burins or graving tools, borers, &c.* Similar discoveries were made in the ‘‘ Grotte de l’Hyéne,”’ on a distinct Mousterian level, as well as in the “ Quina”’ deposit of the Reindeer period, z.c., Magdalenian, &c.+ These discoveries, viewed in connection with ours, are of extreme importance. Better than the bouchers, so many of which are of the Mousterian facies,t better than the flakers, scraper-knives—them- selves types of great survival—these nuclei-like, and especially these partly facetted or rounded stones, justify us in connecting the older lithic industry of South Africa with a Mousterian culture—a culture which is not restricted to the manufacture of large cleaving- or digging-stone tools. FLAKES AND SCRAPER-KNIVES. Had we to deal only with the more or less almond-shaped, or with partially tongue-shaped tools, or even the rougher implements that served an identical purpose, doubt would not be permissible, and the South African implements of that type might, with safety, be considered as belonging to the Chelleo-Mousterian, and even to correspond in age. But the presence of flakes, mostly, but not necessarily always, arge, greatly complicates matters, because their fabrication and use are continued until a time which is here practically that of yesterday. This may be due to their useful primitiveness. That these scraper-knives were originally evolved from spalls resulting from the preliminary trimming of the bouchers no one * Cf., Chauvet, ‘‘ Stations Quaternaires”’ ;—Parat, ‘‘ La Grotte de l’Ours,” &e. + The quartzite bouchers of the Pyrenean region, which so greatly resemble those occurring here, possibly on account of the difficulty they offer to trimming, are connected by Déchelette with Acheulean and Mousterian periods. t The Mentone caves as well as the Taulbagh deposit have demonstrated the fact that in the early quaternary period (pleistocene), there were to be found primi- tive implements resembling more the Mousterian than the typical Chellean or Acheulean “coup de poing.”’ 44 Annals of the South African Musewm. can deny who looks at the representations of some of them given in Pl. XII. Figs. 97 to 100 are particularly instructive. There can be no doubt, also, that from their appearance and situation many of these tools, flakers, scrapers, or knives, are to be considered as old as some of the bouchers. A fragment of scraper absolutely similar to, and of the same composition as, the Cuts in Fig. 104, was found embedded in a raised beach situated some 3 miles, as the crow flies, from the present sea-coast. It is identical with the large flakes found together with large bouchers at, or close to, Cape St. Blaize—a locality not far removed from the raised beach alluded to. A point of importance is that none of these knives or scrapers that from their position might be considered ancient, moderately ancient, or very ancient, exhibits traces of paring on the reverse side, and in that respect these flakes are doubtless Mousterian in shape. But the question naturally arises: why did these flakes not follow here, and also elsewhere, the same evolution in technique as the bi-facial bouchers with which many are in South Africa undoubtedly con- temporaneous, and produced by the same artisans ? Being coeval with the boucher, it seems natural that the paring of either face of the flake or spall into a lance- or javelin-head, or a hand-throwing spear, should have of necessity followed, as it has elsewhere, but such has not been the case either in the South African paleeolithic type or in the neolithic. It is therefore almost certain that the South African maker of paleolithic stone implements, as well as the aboriginal who fabri- cated the neolithic, were ignorant, with a few exceptions to be mentioned hereafter, of the method of stone lance-head manufacture, the use of which weapon would in itself denote great bravery. But the neolithic maker, unlike the Magdalenian hunter with whose culture his has so many analogous points, is well acquainted with the bow, and arrow tipped with very small chips, the minuteness of which implies of necessity the use of poison. Craft versus brute strength. The failure of the evolution of the flake into a lance-head pared on each side in the Solutrian sty ¢ may, however, be explained by the process of trimming the pebble. This trimming of a boulder in most cases already naturally rounded, into a more or less amygda- loidal shape, is the natural result of artificial concussion or impact, as I have endeavoured to show in Chapter III., and the boucher is therefore older in date than the flake. But in many instances it is impossible to assert that a scraper-knife The Stone Ages of South Africa, 45 is more ancient than another, or belongs to a different type; and as indications of its age, we can only be guided by the site or position in which it is met with—indications which are far from being conclusive evidence. I have endeavoured to figure all the types in our Collection. Illustrations of implements of that kind give a better idea than descriptions. All the figures of Pl. XII. I consider to belong to the paleo- lithic type. Many of them were found where bouchers occur. I have explained their shape in a former chapter. In Pl. XIII. all the Cuts of Fig. 104 may be said to be also of a very old type, but with the exception of Cuts 7 and 12 they were not found in connec- tion with bouchers, although these paleoliths occur in the same district. Cuts 1,2, 5, and 8 are from cave-shelters on the littoral. Cut 6, also from Knysna, has been hacked into a sawing tool. It bears at the base a dent that looks like a notch—a rare occurrence which, however, must be looked upon as accidental. Cuts 7 and 8 have traces of a similar one, and so has Fig. 101 of Pl. XII., in which the dent is not well shown. The Cuts of Fig. 105, Pl. XIII., represent scrapers found together with the bouchers in the Nooitgedacht deposit of the Vaal River. As much worn and polished as the large paleoliths not only of that particular deposit, but also of others found on the surface (Pl. IX., Figs. 57 to 61), their shape is indistinguishable from that of Figs. 106, 107, Pl. XIII., found in the Karroo and the Cape Flats respectively, and which may be said to be of yesterday’s date. Cut 3 of Fig. 105 and Fig. 108 are, however, somewhat out of the common, because the base seems to have been trimmed into a wedge-like shape, as if they were intended for tipping a lance, but on close observation it will be noticed that in Cut 3 the thinned reduced part is too short to allow of hafting, but Cut 1 of Fig. 108 could have been hafted. We find also among relies of that old type a few, very few, tools with a peduncle. This ‘‘tang’’ was intentionally produced, doubt- less, in Cut 2 of Fig. 108, which, like Cut 3 of Fig. 105 and Fig. 108, is from the Kimberley District, Cape Colony. Doubt is permissible for Cuts 1 and 2 of Fig. 109, both from the Vereeniging deposit. But from a neolithic situation we have Figs. 125 from Queen’s Town, Cape Colony, in which the ‘ tang” is undoubtedly inten- tionally produced, and Mr. Cottell found at Cradock two examples with a somewhat similar peduncle, of which he sent me a sketch. These were subsequently figured by Mr. J. P. Johnson. 46 Annals of the South African Museum. Figs. 119, 121, and 122 represent the chalcedony implements of the neighbourhood of Victoria Falls. The technique is the same as in many of the Cape Flats type as represented in Figs. 118 and 120. Cuts of Figs. 135, which are stone implements found in the Matoppo caves, where paintings occur, are the very counterpart in size and shape of quartzite scrapers discovered at the foot of the Paarl Rock in the Cape Colony, which show traces of enormously long exposure. From the examples given it is seen that the shape or style of paring the scraper-knives or fabricators cannot give a clue to their respective age or to their relation to each other. Nor is the size of the flakes an indication. Some of the Victoria Falls chalcedony implements are certainly as small as many of our Cape-Flats and inland districts neoliths, with the exception of some of the ‘‘ pygmies,”’ and it may be said that flakes of sandstone or banded jasper are, as a rule, larger than those of silicious or partly silicious rocks. Negative, however, as the results of comparison between flake- knives or scrapers, undoubtedly very ancient, and equally un- doubtedly recent, are, we have fortunately evidence of another kind which throws light, of a sort, on the progress or regression of the lithic industry in South Africa. We have examples of a solutrian type in the shape of lance- or javelin-heads, chipped on each side, and laurel- or willow-leaf shaped (Figs. 110, 112); of aurignacian-solutrian-type scrapers with plain, secondary trimming of edges, either at one end (Figs. 123, 128) or even all round (Cuts 1 and 2 of Fig. 131). We have also tardenosian pygmies (Figs. 140, 141, 143), polished tools, and arrow-heads with tangs (Figs. 116, 117), as well as a bone- and a stone-culture commingled. But this evidence, for reasons which will be duly explained, I include in the South African neolithic. This digression into the neolithic period is here necessary for the proper comprehension of the finds in the various deposits which, rightly or wrongly, I assimilate to the chelleo-mousterian of Europe. (27%) CHAPTER VI. 22 ‘An ACCOUNT OF THE “ STATIONS. In order that my conclusions regarding the South African Palxo- lithic Age should be tested eventually, Iam giving here an account of the main deposits. This account will, of course, contain repetitions of some of the points dealt with in the previous chapters, but it is to be regarded as the documentary evidence on which my explanation of the Paleolithic and Neolithic Ages in South Africa is based. Tue DEPOSITS OF PAARL, SIMONDIUM, STELLENBOSCH, XC. The first find in the neighbourhood of Cape Town of very large- sized quartzite paleoliths was made on the Koeberg Road by the Honourable W. F. Lyttelton in 1880. The implement found is of the rare ovoid-discoidal shape, and is represented in Pl. IIT., Fig, 24. The second was met with in the streets of Paarl, Cape Colony, which are coated with what is here called “ gravel,” and thought at the time to be laterite, but now considered as a coarse ironstone. On visiting the quarry whence the gravel was extracted, I found several paleoliths of very large size. In the centre of the ironstone deposit was a huge block of the same material, too compact to be easily broken, and from one of the median faces projected a large palxolithic boucher, which I could not detach without fracture. Some 400 or 500 yards away on the higher slope of the hill, and about 114 or 2 miles from the Breede River, a piece of ground had been ‘‘delved’’ for establishing a new vineyard. Alongside were two heaps of paleoliths and nuclei, thrown aside by the workmen. From one of these I selected the implement, Fig. 22 of Pl. III., and others. Not a single scraper or small spall was to be had. These paleoliths and nuclei were found at a depth of 24 or 3 feet—the depth of the ‘ delving.” I could not continue my investigations at a higher level, as no new ground was put under cultivation that could reveal the extent of the lithic deposit. 48 Annals of the South African Museum. The importance attaching to this deposit is that quartzite does not occur in the locality, only granite. The water-worn quartzite boulders have therefore been originally brought from the river to their high, commanding position—a position doubtless more ele- vated than the spot where they were found. That they had gravi- tated there is not a mere assumption, because on the top of the mountain, at the foot of these large granitic bosses, after one of which the Paarl* is named, there were ultimately found scraper- flakes of quartzite as worn as any I ever saw, and this in a place which not only surmounts the “ Factory Site’? where I found the paleoliths, but where also quartzite is not a local rock. The Simondium ‘“ Station’? presents the same feature as that of the Lower Paarl, and the sketch here given will make the recital more easily understood. The ‘Station’ is situated on a some- what abrupt talus of Simonsberg on the Drakenstein valley side. On the very steep slope of ‘‘ Pontac Hill,” near the dwelling- house, on the Pomona Estate, an extremely large number of arti- ficially worked stones had been exposed by cultivation on what is a saddle of moderate width sloping on one side towards ‘‘ Donker- hoek,” and ‘‘ Pomona”’ on the other. A narrow, sinuous ‘“nullah”’ had been eroded to a depth of 10 or 12 feet. At a depth of some 9 feet I espied on one of the sides the projecting butt-end of a boucher which proved to be one of the finest I ever obtained. On the surface, exposed by digging or ploughing, and almost in a line with the first embedded boucher, I picked one of the heaviest ‘‘hache a talon”’ I have yet found. The two had plainly gravitated from a higher altitude, as others had done, which lying on the surface were, either spall-like, partly manufactured, or nuclei— and what cores! some more than 18 inches in length—littering the ground, I traced a number of these paloliths for nearly a mile along the ascending talus up to 550 feet above sea-level (see cross- dotted scheme in sketch). At that height, also, the implement had been exposed by delving the ground for a plantation, and there is no reason to disbelieve that, were cultivation to be carried higher on this mountain talus, palsoliths would be revealed. It is at the higher altitude that two implements like Fig. 23 of Pl. III. were found, and some of the bouchers were almost in a state of disintegra- tion. A search at a lower level led to the discovery of other paleo- liths deeply imbedded in stone-gravel in cuttings on each side of the Simondium Railway Station, 370 feet above sea-level, and about * Corruption of the French word perle, given it by the French Hugueno settlers. The Stone Ages of South Africa. 49 2 miles from the Berg River, which runs there at an approximate altitude of 300 feet. I doubt not that a renewed search will lead to the discovery of these paloliths close to or in the proximity of the riverbanks. This Simondium deposit as now traced extends Fic. 1.—Simondium Factory Site. for nearly 2 miles on a steep slope. Hitherto I have found there no undoubted scraper-knives ; only large unused flakes, large spalls, and large nuclei. It is possible that flake-knives were once there and utilised as such, but one thing can be definitely said, and that is, that they are no longer found with the finished, partly 4 50 Annals of the South African Musewm. pared or unfinished palzoliths exposed on the surface or found at shallow depth. Another point which this Simondium deposit has in common with that of the Paarl, which is nearly on the same level, is the selection by the makers of such a high position for their workshop or abode. Was the river that drains the valley nearer than it is now? The great thickness of the iron-gravel at the Simondium Railway Station, in which some of the implements were found, has, like the large block of iron-stone at the Paarl, taken a very long time to accumulate. In that deposit, while forming, the stone implements, artefacts already of great antiquity, have gravitated, following the course of hill denudation and valley formation. And this occurrence is repeated wherever the Stellenbosch-type quartzite boucher is found. Look for it in iron-gravel or below it, when not very thick, and you will find it there as often as in alluvium or silt, perhaps because it was arrested in its downward course by the formation of conglomerate. Stellenbosch Station. At Bosman’s Crossing, near Stellenbosch, the occurrence of these palzoliths is somewhat different. At the foot of a steep hill called Papagaiberg runs a small rivulet, a tributary of the Eerste River, which it joins close by. The spur of that hill abuts on that rivulet, and is intersected on one side by a cart-road and a railway cutting on the other. The space thus left has been used for a good many years as a brick-field, from which a thickness of some 20 feet or more of material has been removed. I found there in the vertical wall, from which the clay was detached by pickaxe, two superposed layers of fractured, water-worn boulders, spalls, nuclei, finished bouchers, such as Figs. 2 and 4 of Pl. I., and a few triangular scrapers that, ike Fig. 95 of Pl. XII., had probably not been utilised as such. They had been deposited on the granite formation, and I traced these layers on the other side of the railway cutting, also resting on the mass of granite which terminates abruptly on the bank of the Eerste River. I traced them also on the other side of the cart-road, flush with the floor of an excavation 30 feet deep, and corresponding, with only a slight incline, with the granite surface on which the deposit rested. The great accumulation of this brick-clay is in itself a proof of great antiquity. But in addition to this Bosman’s Crossing deposit I have since proved that palzeo- liths are found in the whole of the Stellenbosch (Herste River) The Stone Ages of South Africa. 51 valley high above the river or nearly on a level with it, and seldom accompanied there by chips which are not more in the shape of spalls than of flake-knives. The same conditions obtain in the other river-valleys of South Africa. In most deposits the partially pared paleoliths are more numerous than the better shaped tongue-implements, but this, I venture to say, is ascribable less to the degree of the maker’s skill than to the greater resistance offered to the continuous action of rain-wash which, gradually, brings the artificially worked stones from the original position in which they were left or manufactured to a lower level. The heavy, round, butt-end of a partly pared paleolith offers much more resistance to the denuding agent than the flat- tened, almost navicular, tongue-shaped implement which rests on a very much reduced surface (see profile of Figs. 1, 2, and 8). There is, therefore, no reason militating against the acceptance for our South African paleoliths of as great antiquity as for the deposits of the Pyrénées, of Spain or of India, where, as here, the smaller artefacts in the shape of scraper-knives are equally rare. BEUKESFONTEIN DEPOsIT. Very primitive-looking are implements of a large type found in the Ceres District of Cape Colony. They could not have attained a much more water-worn aspect had they been rolled for centuries in the shallow yet impetuous waters of a spruit—a treatment which they evidently never experienced, this water-worn aspect being caused by weathering alone. In this locality * the outcrop of the bluish-grey chert band of the Dwyka forms a little ridge, at the base of which is a gentle slope of soft shales. This slope, for a distance of fully 200 yards from the outcrop of the chert band, is strewn with blocks of chert and worked fragments of the same. There is very little soil, as the shales crumble away into small fragments. It is therefore impossible to determine what length of time the implements have lain there. No small scraper-knives or flakes were found. But from the outcrop of the Dwyka chert band Mr. A. R. Walker, one of the Museum Assistants, has brought back from Matjesfon- tein (Cape Colony) a few implements, among which is a boucher of moderate dimension, whose original facets have disappeared, and which is as equally smoothed and pitted as any one example from Beukesfontein. For an implement to be so closely pitted one must assume a very long period of exposure to the elements. This * Described in Ann. Rep. of Geol. Com. for 1903, p. 25. 52 Annals of the South African Museum. boucher, a digging tool of no considerable technique, is slightly curved in the line of the long axis, but it is as well, or very nearly as well, smoothed and pitted on one face as on the other. It is doubtful if rain-wash alone caused it to roll regularly in such a way and for such a period that each face was alternately and equally subjected to disintegrating influences, but a stream, now no longer traceable, could have done it ; and it must also be stated that there is a dry river not far from the spot to which leads the slope where it was found. According to Mr. Walker, from its position on the lower slope of the hill below the chert band, it might have been carried there from a higher level. At some distance from that boucher, but above it, Mr. Walker collected one of the usual Karroo type, a moderately small, lanceo- late, triangular scraper, patinated also, but not as deeply as the boucher itself. The patina was distinctly deeper on the obverse than on the reverse side with its usual conchoidal trace of fracture, showing that it had lain undisturbed. At a higher altitude than either the boucher or the scraper were found bluish-grey nuclei with sharp edges, the facetted contour of which corresponds to the shape of the removed flake of the usual sub-lanceolate form, and also a large trim- ming flake, the jagged edges of which testify to its use, and plainly detached from either a very large pebble or rounded boucher. Its curvature is even greater than that of Fig. 98 in Pl. XII., which in itself is not more incurved than some of the ‘‘ éclats’’ we possess from ‘“‘ Le Moustier’”’ Station of France. The importance attaching to the find of this boucher is that the apical part of one face has been used again slantingly as a muller, or rubbing implement. The abraded surface shows the original bluish-grey texture of the chert pebble—a colouring similar to that of the non-patinated tools found at a higher elevation. From this it would appear as if the makers of the surviving, some- what degraded type, such as the usual scraper, either no longer knew the manufacture of, or had never been acquainted with, the primitive use of the boucher. Prompted either by curiosity or a sense of adaptation, they put this tool to a purpose not originally intended for it. THE Crapock Drposits (CAPE CoLony). I have received from Mr. W. H. Cottell, formerly on the staff of the Cape Colony Public Works Department at Cradock, a represen- The Stone Ages of South Africa. 53 tative collection of stone implements discovered by him in the neighbourhood of that town. I give here his answers to some questions I put him—questions which are well indicated by his answers. I must, however, preface his response by explaining that the “large tools” are paleoliths varying from 6 in. to 8 in. in length, made of a hard shale indurated by the intrusion of dolerite. Some pieces are tongue-shaped, sharply pointed at apex, while others are more irregular; the flaking is very crude in most examples, and the cleavers plainly of a coarse style of manufacture. Although chipping has been carried round the butt in some examples, Fie. 2.—View looking towards Fish River and first Kranz, 3 miles north of Cradock. in a few others the contour of the original pebble is retained, much as in the Stellenbosch type; but as a rule these bouchers are less bulging at the centre.* They have a greyish patina that contrasts greatly with the black colour of the shale, but they have not been smoothed artificially either by water action or by very long exposure. A few pieces are not patinated at all. The ‘‘ Bushman tools ’’ are somewhat of the Aurignacian-Solutrian- type scrapers, bevelled at one end by secondary chipping, as figured * It seems, however, that I have not seen the ‘‘ best finished implements,” only photographs of some of them, but the latter are quite instructive enough. 54 Annals of the South African Museum. in Pl. XVI., Figs. 123 and 128; also nuclei, irregular flakes, scraper- knives, &c., in great quantity, but most of them also of a very crude type. Mr. Cottell’s answers are as follows :— ‘‘ Before replying to your questions as to sites, hill, or other for- mations, in which the paloliths and the seemingly recent imple- ments are found, I was anxious to visit a few places to make sure of the relation of the ‘ paleolithic’ type to the ‘ Bushman.’ I found a deposit of a considerable extent of the large tools some 5 miles south of Cradock, and about a mile and a half east of the river. The deposit is on the slope of a flat-topped hill, the west side of which terminates in a kranz facing the river (text-fig.2). The whole of the Fie. 3. slope is covered with chipped boulders of a hard, black shale. The same material caps the top of the hill. These tools are very crudely chipped and badly modelled. Some are of gigantic dimensions. The majority appear very modern, almost as if quite recently made. There were a few old ones of much better design and finish. I am sending a few of each. “There were no ‘Bushman’ tools with these large ones, nor remains of game or shells. ‘“On the kranz side of the hill, however, there were ‘Bushman’ flakes, mostly of the knife-scraper lanceolate form, and shells (? Unio sp.). ‘“‘T revisited the ‘ Baths.’ The large types there are on the fringe The Stone Ages of South Africa. 58 of the ‘Bushman’ deposit, and although there are scrapers near, none are mixed with the large tools. ‘These large implements are not deeply buried in the alluvium ; the majority are on the surface, and when buried are only covered by a few inches of sand, with the exception, of course, of those washed into ‘ spruits.’ ‘The ‘Bushman’ deposit is almost wholly confined to the river- banks. Scattered over the veldt, however, are occasionally flakes, lance-like in form, usually broken, and without secondary chipping. Fie. 4.—View showing top part of alluvium and implements left behind after soil has been washed away. Three miles north of Cradock. With the deposit are quantities of shells, pieces of pottery, and remains of game. Here and there are little piles of broken flags of sandstone, with their faces ground smooth from wear, and water- worn pebbles with one or more ground facets, evidently obtained by being used as mullers or grinders. ‘One cannot estimate the age of the tools from the weathering ; some found at the surface have the exposed side bleached and the unexposed side almost, if not quite, fresh ;* those near the thermal * This statement is not entirely borne out by some of the examples received ; in several both faces have a greyish patina, but a patina shared also by many of the smaller ‘‘ Bushman ’”’ implements. 56 Annals of the South African Musewm. springs seem to have weathered much more rapidly than else- where.”’ It is difficult to speak with any degree of authority of a deposit which one has not personally inspected, but from the tenor of the letter here given, which is in answer to direct questions sent by me, as well as from the examination of the implements, I am led to conclude, until further investigations disprove my conclusions, that the Cradock deposits imply a contemporaneity between the large bouchers and the small or only comparatively large scraper-knives quite equal to that of the Nooitgedacht deposit. In one case there were no ‘‘bushman”’ implements found together with these palzo- liths; while in another the latter were found on the fringe of the ‘‘bushman’”’ deposit, which is almost wholly confined to the river-bank. We have thus probably a repetition of an instance of denudation by rain-wash, the same as in the Stellenbosch, Simondium, and other sites, only that at Cradock the smaller implements have not been carried away, or only partly so, into the streams or rivers. THE ‘‘ NOOITGEDACHT’’ AND BARKLY WEsT DEPosIts. At a place called ‘‘ Nooitgedacht,” close to the Vaal River, in what is called here the “dry diggings”’ for diamonds, there have been found scraper-knives, and small and large bouchers made mostly of diabasic rocks, all mixed together. This find is of very great interest, as showing more conclusively, perhaps, than even the Tyumi River deposit, that knives and scrapers of the Mousterian type were made or used by the same people who manufactured the large bouchers of the chellean- mousterian form. Miss Wilman* who sent me some of these implements, so abraded that identification was difficult, and whom I had asked to ascertain if there were old or recent river-terraces near the deposit, wrote as follows :— “The Nooitgedacht implements, big and small, all occur together in a bed of gravel that is being worked on the water’s edge of the river (Vaal). In fact, the diggers have been flooded out at times. “The diggers have removed the overlying sand at A and are clearing out the gravel at B for sifting and sorting. All the imple- ments sent come from £. This, then, is the lowest terrace, and * A former assistant of this Museum, and now in charge of the McGregor’s Memorial Museum at Kimberley. The Stone Ages of South Africa. 57 it rises to a couple of hundred feet and has quite large trees growing upon it. In the sand are no implements, but lying all over the surface are some of the same type, exactly as those found in B. Large implements are not common on the surface, and some are Fic. 5.—4dA—Fine sand. B—Gravel, diamond bearing, consisting almost entirely of rolled agates, &c., wherein are the large and small implements. quite unlike the water-worn rocks that they accompany, and I am convinced that while some are undoubted implements, others were implements in the making.” The pieces received from Miss Wilman are almost smooth, with the angles and facets quite obliterated. Many are tongue-shaped, and others more or less regularly almond-shaped. The examples represented in Pl. IX., Figs. 57 to 60, which were found at Barkly West, not far from Nooitgedacht, give a good idea of their appear- ance, though some are even more amorphous. In common with many collectors who pick specimens to oblige friends, Mr. A. du Toit, of the Cape Geological Survey, who pre- sented these last-mentioned examples, left many the conditions of which through abrasion made him doubtful as to their being artefacts. On the whole, these bouchers show traces of superior workman- ship, most of them having been ‘“ knapped”’ on either side. Some are of the unusually large ‘‘shard’”’ form trimmed only on one face, and it is to these that Miss Wilman alluded as being in the making.* Several of the smaller, about 100 mm. in length, are well finished, others not; one still retains its wedge shape. The cleavers are easily recognisable, and round the butt of one there is a very irregular depression that, if not accidental, might make hafting by ligatures possible. The cleavers of Pl. VIII., Figs. 52 to 56, met not far from Nooitgedacht, are of the same type. As for the scraper-knives, their resemblance to similar tools found so commonly all over South Africa is indeed extreme. I have selected a few types for illustration (Pl. XIII., Fig. 105). I am, unfortunately, debarred from giving in this modest paper illustrations of tools of the same technique found in mid-Europe or England, but I should like to call the attention of antiquarians to the extreme likeness of No. 7 of Fig. 105, to one of the famous * One sent me is 260 mm. long, and weighs 4 lbs. 4 oz. 58 Annals of the South African Musewm. “« Micoque ”’ lanciform knife-scrapers, attributed to the Chellean- Acheulean ; of that of No. 4 of the same Fig. to the ‘‘ Chatelperron ”’ scraper-knives, &e. The importance of the Nooitgedacht deposit is that the scraper- knives and bouchers are found together, all bearing clear proofs that they have been subjected at one time to the same process of abrasion, and we are therefore justified in arriving at the conclusion that the two types of implements, the boucher and scraper, were artefacts of the same makers. But when we try to assign a geological date not only to these relics, but also to similar ones from the deposits of the Vaal River, we are confronted with the same difficulties as in the case of the other South African deposits. In this Nooitgedacht gravel-bed numerous potsherds were found associated with the implements. One piece, now in the Collection, is of moderately close texture; the greatest thickness is 8 mm.; the edges of the fracture are not abraded in the least, and the outside glaze is wonderfully well preserved. There is nothing to differen- tiate it from the pottery made by ‘“Topnaar’’ Hottentots (vulgo “Strand Loopers’’). This pottery must have been deposited accidentally in ‘the gravel- bed before the latter was covered by the layer of sand, but it points also to the implements having been deposited there in their already abraded condition, because the evidence of the unabraded potsherds preclude the possibility of the implements having been water-worn in their distinctive manner in that particular cul-de-sac, or ‘ pot-hole.”” They must, therefore, have been brought down with alluvium, rom higher levels, from river-terraces. We shall see what evidence is obtainable on that point; but one must not forget that specimens found in the gravels of the Vaal River must have been left there in the same worn condition as when deposited; that is to say, worn by gravels in the process of formation. A close examination of some of the Nooitgedacht pieces reveals traces still fairly visible of the pitting that long-prolonged eolian agencies impart to diabasic implements. This is very important. It implies that the paleoliths were already of great antiquity before being subjected to the fluviatile attrition. Messrs. J. P. Johnson and R. B. Young, in a paper read before the Geological Society of South Africa,* state that ‘all along the * « The Relation of the Ancient Deposits of the Vaal River to the Paleolithic Period of South Africa ’’ (IX., 1906), p. 53. The Stone Ages of South Africa. 59 Vaal River are two well-defined terraces of gravel, the longer one usually being covered with a considerable thickness of loam.” Mr. J. P. Johnson in a subsequent publication * qualifies this statement, ‘All along the Vaal River there are well-defined terraces. There are usually two: the upper and older one consisting of a thick bed of gravel; the lower and newer one being, as a rule, a stratum of gravelly detritus lying at the base of a varying thickness of alluvial loam.” | The two writers above quoted, when speaking of the Barkly deposit and mentioning the extraordinary abundance in it of the typical paleolithic implements, “all but a very few equally rolled being practically reduced to pebbles,’ state, inter alia: ‘At Barkly, between the bridge and the village, the upper terrace is well exposed in the old diggings. It lies at the foot of a ridge of hills, hence the talus element is predominant, though the river gravel is in evidence throughout.” They found, together with the rolled bouchers, ‘(a few sharp implements which led them to conclude that that deposit consists of two distinct series, the one probably older, the other perhaps con- temporaneous with, perhaps newer than the deposits.’ Unfortu- nately, these ‘“sharp’’ implements are not figured, and it is thus impossible to traverse this conclusion, but we possess a good many pieces from Barkly and its near neighbourhood. I have also seen a good many besides those we own,} and I would certainly not differ- entiate respecting the contemporaneity of a ‘‘ sharp’ implement, like Fig. 53, of Pl. VIIL., a sharply wedged tool, Fig. 56, or an obtuse boucher like Figs. 57 or 58 of Pl. IX. Mr. A. L. du Toit, of the Cape Colony Geological Survey, who has surveyed this very district, was kind enough, at my request, to investigate this deposit, and this is what he wrote to me, accom- panying his remarks with the following diagram given on next page. “There are several old river-terraces in the neighbourhood of Barkly, one immediately above the present river-bed, and a second from 30 to 40 feet above it as recorded by Messrs. Johnson and Young; this higher terrace is indicated at A on the accompanying Fig. At B there is a gap in the hills through which a loop of * The Stone Implements of South Africa, 1908. + In the Bloemfontein Museum are three specimens found 40 feet deep in the old course of the Vaal River. Two are of the usual worn boucher-type, 100- 115 mm. in length; the third is a small circular chip 22 mm. at its greatest length. 60 Annals of the South African Musewm. the Vaal River very probably flowed in former times, C then being an island. ‘Traces of still older gravels are found on the tops of the ridges C and D about 100 feet higher up; but to the north this upper terrace is continuous and well developed, but does not appear to carry any large implements. The gravels at Riverton (west bank) and Warren- ton seem similarly to be without implements. BARALY 4 g “ G “ 3 tz Z Ae D aa - hs (oe é Orr, 3 ‘ = i! 1 a vn L Ni Cp et wy A iB CHT 0 got, 2 at 10 y= BOE NE a MY ty, z i= (les rend VM A 3 SB ee . wll WM, t= ie ee: >: [ef Bop a (fe = > A “Uy ENE cu, SS fi NAG ta US 2 EAS SUN Wiss (UN) SSNS «This middle terrace at Barkly, which therefore corresponds to the ‘upper’ of the two authors quoted, carries many rolled implements, and at the point marked A they have been brought up in thousands in the course of sinking shafts to reach the diamond-bearing gravel. Among the inclusions of diabase, some only slightly rolled others very well rounded, are numerous large, well-shaped implements considerably water-worn, their abundance at this one spot being most remarkable. “Tt seems most likely that the implements were manufactured on the slopes of the little hill C, and that they gravitated down into the river at 4. At the time when this terrace was being formed there would probably be an eddy at A where the two branches of the river reunited, and this might account for their extremely rolled condition and possibly for their concentration at this point.” THE GRIQUALAND WEST JASPER IMPLEMENTS, AND THE TyuMI RIVER DEPOSIT. The Griqualand brown, yellow, or whitish jasper implements have a certain aspect of their own, perhaps due to the material The Stone Ages of South Africa. 61 of which they are made; but many bouchers are exquisitely finished, The diggers, or cleavers, are most effective implements, cf. Pl. V., and Figs. 48 and 49 of Pl. VII., but they are very seldom found on the surface, and are met together with by-products, chips, and flakes, when opening or cleaning wells or springs that have become obstructed, or have disappeared altogether owing to the deposition of carbonate of lime. Some of the implements in the Collection have been encountered at a depth varying from 18 to 20 feet. At the springs that supply Griqua Town with water these banded jasper implements were found in great quantity. These finds, made while cleaning water-holes, are extremely common, and have given rise to the belief there that when- ever Bushmen had, for fear of attack or for other motives, to abandon the springs round which they lived, they would throw all their domestic appliances into the water-holes and obliterate the latter. it was at one time thought that the jasper implements were restricted to Griqualand West and the Prieska Districts of the Cape Colony, until a very interesting discovery was made in the banks and on a drift of the Tyumi River, near Lovedale, in the Alice District of Cape Colony. I have seen twenty-eight pieces discovered by Mr. A. Johns, some of which he has presented to us. Among them are many beautifully worked Acheulean-type bouchers, some of which are reproduced on PI. [X., Figs. 77, 78, and 86, and lanceolate flakes pared on one side only. There are also a few flaking hammers (Fig. 88, Pl. XI.). These implements are made of brown and yellow jasper ; parts of the faces that have been exposed exhibit the beautiful glaze peculiar to this sub-type, while others are quite lustreless. With them are three scrapers of black lydianite, and a large and somewhat thin boucher of a rough type that might have served as a broad-edged cleaver. This boucher was made of a local rock. The finish of the bouchers, and of some of the lanciform knives or scrapers which retain, however, the strong conchoidal bulging of the reverse side, point to a great development in the lithic industry This find also adds another proof that bouchers and scraper-knives were produced simultaneously and by the same makers. An interesting point in the Tyumi River discovery is that the jasper rock is not as yet known to occur in this locality. The first suggestion, that the material already manufactured has been transported a distance of 300 miles from Prieska (Cape Colony) by the nomadic makers, had to be abandoned when chips and small fragments were found in some pieces of conglomerate adhering to 62 Annals of the South African Museum. the bouchers, showing conclusively that the tools had been made in siti. The origin of the jasper has to be sought, therefore, in the Dwyka conglomerate which crops out about 20 miles south of that spot,t or else the tools were shaped from rough material brought from the northern Karroo. But the dates of finds, as well as the names of localities, inscribed on the pieces are not always the same, and it would appear that this Tyumi deposit was not restricted to the bank of the river from where most of the examples were obtained by Mr. Johns, or to the river spruit. Unfortunately I could not keep in touch with that gentleman, who discovered the deposit in 1906. But that the makers carried with them their implements during their migrations or wanderings is borne out by the find in an Kast London shell-mound of a scraper-knife made of yellow jasper. THE VEREENIGING DEPOSIT. Mr. J. P. Johnson has given an account of this deposit in the Trans. S. Afric. Phil. Soc., xvi., 1905, p. 107, which he describes thus :— “The Vaal in that part has cut a channel deep into the solid rock, and on top of the cliff thus formed, and extending, to my knowledge, some distance east and west of the town, is an old river-terrace consisting of gravel and small boulders embedded in and overlaid by loam. There is a small pit in it, east of the town, where flakes occur in great profusion, and nearly every pebble (which are all of quartzite) | has been chipped. They appear to be largely the result of unsuccessful attempts at manufacturing implements. No finished one has been found in this spot... . Mr. Leshe’s find is some distance west of the town, where long stretches of the terrace have been furrowed and spread out by the rain. There, for many hundreds of feet, unfinished implements occur in the greatest abundance, the flakes produced in their manufacture by the thousand, while here and there complete specimens are met with. The quartzite seems to have been of too coarse a grain, as a rule, for suitable working, as nearly all the failures and very few fmished implements are in this material, the majority of the good specimens being of greenstone (diabase). One or two unfinished examples of chert were found. * Bouchers of the same jasper rock have lately been found in the Bedford District not very far from the Tyumi River deposit. + A. L. du Toit. { Compare this with the Nooitgedacht find. The Stone Ages of South Africa. 63 “It is quite clear that these implements must have been made very close to where they are now found. Very probably the gravel is the sweepings of an adjacent land-surface where the implements were manufactured on a large scale. Many of them are as sharp and fresh as on the days they were made, while obliteration of the sharpness of the facets in others is more often due to weathering than wear.” Mr. Leslie had sent me these implements, I believe, before Mr. Johnson saw them, and I made copious notes of the same, the gist of which is: (1) That diabasic tools are as much worn and smoothed by water agency as those figured on Pl. IX. and of those mentioned in the Barkly West deposit; (2) that the quartzite implements are of a different style of manufacture. They cannot compare in work- manship with the Stellenbosch implements; they are more of the cutting axe (Mousterian) type, one of the faces having been cleaved at one blow, while the other has often only three broad facets and a few (?) secondary chippings along the edge. Fig. 57 of Pl. VIII. is peculiarly instructive ; its shape has been obtained by very few blows, yet it is an effective cleaving tool. (3) These quartzite tools show also signs of wear, but nothing like the diabasic ones. The scraper-knives of ;diabase have the edges much rounded and abraded, and several of them are almost polished, as if by water ; but not so the quartzite implements. Although it must not be forgotten that if the matrix, such as quartzite, is more difficult to work than flint or chert, the difficulty the manufacturer of the boucher encountered in working diabasic rock was probably greater. We must therefore attach no importance to the assumption that because these two forms are found together they are not only con- temporaneous, but produced by the same maker. In fact, I believe the contrary was the case. Mr. Johnson, as here quoted, mentions a small pit (in an old river-terrace) where flakes occur in great pro- fusion, and nearly every pebble (which are all of quartzite) has been chipped. These diabase, or quartzite, relics were not found alone, however. With them I received from Mr. Leslie the agate, quartz, jasper, cherty, small, somewhat amorphous chips, together with small scrapers of the usual modern type. Mr. Leslie’s letter in answer to my questions is very clear on the point :— ‘“ Now to your questions. (1) Have you found on the banks of the river the agate or chert-like small implements mixed with the large ones ?—Yes, but more often the large ones are mixed with the 64 Annals of the South African Museum. small ones in the sense that when a patch of small implements is found, large flakes or coarse chippings are found in smaller numbers comparatively. (2) Do the diabase implements occur in conjunction with the quartzitic ?—Yes; I can find no dividing line between any. When I say that really all the large implements sent you, with a great many more, were found on the river-banks within a distance of 200 yards along the river, and about 50 yards from it, it will be seen how very difficult it was to distinguish between their relative positions ; but with regard to the small agate ones, they, or the greater number of them, were found about a mile away associated with large flakes. ... ‘On the receipt of your letter I went to a spot about 500 yards from where the implements were found. On a space of 10 feet square I saw flakes, chips, &c., which I am sending you; they show the variety well. I am of opinion that the sign of great age of many of the large implements or flakes is due to weathering.” * The Vereeniging find was made in 1905. Mr. Ivor Guest kindly sent me a sketch-map of the locality. When he visited Bosman’s Crossing at Stellenbosch he found that the talus conditions were the same in both places. I pointed out the Bosman’s site and all its features to Mr. Leslie in 1908, and he informed me that the situation was identical with that of Vereeniging, which he described as “a talus of 4 or 5 feet in depth and extending some miles along the banks of the Vaal River and about a mile wide ; implements both large and small may be found in any part of it.” With the evidence before us it is safe to conclude that the Vereeniging deposit is another instance of accumulation in talus, and this by degrees. Whether the first makers used quartzite, and those that followed diabase for their manufacture is not likely to be conclusively proved ; | I believe that they were not the same race, or had not the same lithic culture or skill. As to the makers of the small, uncouth agate or chert implements, we can, I think, consider them to have been people, if not of another race, at least of one that had lost this state of culture, and this possibly, if not probably, because the in- vention of another and more effective weapon, or the acquired knowledge of iron, caused them either to discard the ancient types or to restrict themselves to one particular line. These agate, &c., implements are similar to those dealt with further on in the * In the face of the Nooitgedacht and Barkly West finds, this conclusion is erroneous. + The dolerite implements are of a far more advanced type than the quartzite. The Stone Ages of South Africa. 65 account of the domestic implements of the up-country caves or rock-shelters. THE FisHHOOK DeEposIrT. Fishhook, or “‘ Vishoek,”’ is a crescent-shaped sandy beach, through which meanders and usually loses itself in the sands a rivulet, or rather small stream. It is situated between Kalk Bay, a well-known seaside resort, and Simon’s Town, the seat of the Admiralty, and some 16 miles distant from Cape Town. On one side runs the road to the Kommetje, on the other that of Nordhoek, and the valley runs from False Bay on one side of the Cape Peninsula to Chapman’s Bay on the other. On one side of the valley is a high hill of Table Mountain sand- stone, forming a very long spur, but this hill, like others in the neighbourhood, is covered by unceasingly advancing and retreating dunes of white sand, in places several hundred feet high, rolling imperceptibly upwards over the rugged sides of the mountain crags under the propulsion of south-east winds of great violence, even capping the crest of the hills in the manner shown in text- figure 7. If the wind veers to the east, or more especially to the north-west, this covering is blown as huge walls of sand in an oppo- site direction, thus uncovering part or the whole of the original surface, but only for a short period, especially where the distance from either sea increases. It was on such an occasion that there were discovered at the foot of a sand-dune, advancing perceptibly to cover again the exposed rocks, bouchers of unmistakably Chellean, Acheulean, and Mousterian types, together with broad flakes of paleolithic form, detaching hammers (Fig. 75), &c. Intermingled with these were also flakes and scrapers in the style and of the material occurring so numerously on the Cape Flats (Pl. XV., Figs. 118, 120), flakers of the same quartzite (Pl. XI, Figs. 86, 90), and, lastly, sherds of Strand-Looper pottery. All this ‘ outillage’’ was resting on a sloping layer of ironstone gravel, in the process perhaps of formation, and producing in places quite a smooth floor, sufficient, however, to stop effectually its gravi- tation to a greater depth. Many of the implements were wedged in anfractuosities of the uncovered rocks (A in text-figure 7). The layer of ironstone was plainly continued under the readvancing sand- hill, at the very foot of which we found one of the most sharply ending bouchers I have as yet seen, and partially coated with ironstone. 5 66 Annals of the South African Museum. On ascending the hill we kept finding paleoliths, made mostly of a more quartzitic sandstone than the usual Stellenbosch type—a sandstone that, we found later, occurs only some 4 or 5 miles from the spot. Well-nigh at the top (B in text-figure) were paleoliths, some jammed into the interstices of the rocks, which are perforated and vermiculated by sand action in a very singular manner. Here and KG. if. there also we found the small Cape Flat type of flakes and pot- sherds. More numerous than the large implements, there occurred at that altitude, some 400 feet from the first find, large detaching hammers of the type of Fig. 75. Their weight had offered probably more resistance to their removal to a lower level. Proceeding along the line of the hill, we did not find any boucher, but at a lower level we met instead with many large hammers or nuclei similar to those found at the higher level. Fig. 75 of Pl. XL, picked up there, represents a moderately-sized one. Arrested no longer by the layer of ironstone, which is not to be found there, they have by the pulsating effect of the removal of the sand reached the floor of the dune, and have in several instances The Stone Ages of South Africa. 67 been deposited so as to form an irregular circle. Fig. 86, Pl. XL, was found on the slope of that floor. Fairly numerous, too, were the small Cape Flats flakes, and we picked up several mullers of small dimensions with the usual ground end of the type of Fig. 168 in Pl XXII: Some 2 miles off, and near a lagoon, but in line with this floor, quite a number of mortars, some of large dimensions, have since been discovered close to each other, some being very much sand- worn. ; On examination it is found that in a few of the bouchers occurring in that locality the original reverse smooth face (Mousterian type) has been retained ; others show usually an attempt at paring, while on each side, and round the butt also, many show signs of having been manufactured from a rounded pebble by retaining a part of the contour. The majority were plainly detached from large nuclei. The cleaving tools were very serviceable, and of the type of Fig. 33, PIT. In length, the digging tools vary from 26 cm. to10cm. I have seen an extremely well-shaped example of the size and shape of Fig. 4, Pl. I. We found a few throwing-stones, and the usual accompanying waste of broad flakes that could be utilised as scrapers. All these implements were very finely pitted by sand action. But, taken together, the bouchers, nuclei, and large flakes are undoubtedly of the Stellenbosch, Chellean-Mousterian type, the only difference being that they are made from a grained, more quartzitic sandstone. Yet there is no evidence that the narrow lamellate surface- quartzite flakes, the mullers with one rounded face, the mortars with the hollow depression, and the rough potsherds are connected with the Stellenbosch, Chellean-Mousterian type, although some reservation may be made in regard to the scrapers or flakes. How can the finding of types so very different lying together in close juxtaposition be reconciled ? The valley of Fishhook is extremely well suited to the require- ments of primitive man, whose main life support was fishing or hunting. An abundance of never-failing water—so attractive an object to man or beast in Africa—an extremely mild climate, ‘‘vleis”’ wherein to fish or obtain wildfowl, a superfluous abundance of sea- shell (Haliotis, Patella, Mytilus) to be found on either side of the isthmus, such were the advantages that primitive man found at Fishhook, and of which, judging from its relics, he made the most of. 68 Annals of the South African Musewm. But the sand-dunes, under the impulsion of impetuous and season- ably recurring winds, have swayed to and fro for years untold, moving forwards and backwards, thus disintegrating the sandstone of the hills encountered in their onward or retrograde march—a dis- integration greatly helped by saline matters. These dunes must have been, and still are, a most powerful erosive factor, as is but too plain when the rocks which they cover are laid bare. Now, it must be remembered that bouchers and nuclei are found at the top of the partly uncovered slope, some are so wedged in the deeply and broadly pitted sandstone that a hammer had to be used to remove them. As it is impossible that these very heavy imple- ments could have been rolled upwards from a lower level by the sand that forms these billowy dunes, we cannot but conclude that they were deposited in the situation in which we found them from a once more elevated level, a higher level eroded by the action of the shifting sands, and an erosion doubtless of great antiquity.* But the recent quartzite scrapers are found in the same situation ; and why should they not have come also from the eroded higher level ? In dealing with the Tsumi River deposit, I have pointed out that scrapers of exactly the same type as those occurring on the Cape Flats, and associated there with beautiful Solutrian laurel- leaf-like implements worked on both sides, were found in such juxta- position with most finished bouchers ot an Acheulean type that it is impossible to doubt the contemporaneity of the two. The same thing occurs also on the Cape Flats. Small bouchers of Acheulean type, made of the same recent quartzite as the Solutrian implements, are occasionally met with, but, I must add, lymg by themselves. Fig. 18, Pl. 93, represents one of these bouchers. It would not, therefore, be safe to deny the possibility of these surface quartzite flakes having been coexistent with the Chellean- Mousterian type found at Fishhook, except for the following reason :— Further investigation has revealed, close by, a midden containing the usual domestic appliances of flakes and chips, querns and pot- sherds, and among them a javelin-head of the advanced type, but * Since this account was written bouchers of the same paleolithic type as those found at Fishhook have been discovered at the very top of the mountains flanking the valley. At Glencairn they show the same effect of sand-blast, and were dis- covered along the only possible opening that could lead to the ‘‘ Kommetjee,” on the other side of that part of the peninsula. We have thus a repetition of the occupation of heights by the early, or primitive makers, comparable to the Paarl, Simondium, and other sites. The Stone Ages of South Africa. 69 somewhat roughly worked on both sides. This midden is of the Strand-Looper type, and no Chellean-Mousterian boucher has as yet been found the connection of which with the midden is indubitable. The evidence afforded by the presence of this midden may and probably does account for the presence of these high-type flakes, as well as for that of the mortars and pounders, such middens never being without an abundance of them and some containing objects quite modern. The flakes were probably made and the debris of the industry deposited at the time when the hill, which commands even now a eood view of the whole valley, was partially uncovered, after a long period of erosion. Being much lighter, and some even of incon- siderable weight, they, as well as the potsherds, found their way with a rapidity appropriate to their size and weight to the level which the Chellean-Mousterian bouchers had reached before them. It must be remembered that no boucher was found on the furthest point of the uncovered floor of the dune, and also that about one mile further, and close to an existent lagoon, quite a number of mortars but no bouchers were found. This seems to point to the fact that the people that made the mortars and the pottery were not necessarily the manufacturers of the Chellean-Mousterian paloliths. Tue ‘Kast Lonpon” DEPposIt. The late Mr. George McKay, of East London, has left some very interesting notes on the occurrence of stone implements at or near the mouth of the Buffalo River at East London, Cape Colony, and, were his premises correct, we could conclude that the relics found there must represent an enormous lapse of time. We are greatly indebted to Mr. John Wood, of East London, for a very representative collection of these local implements. The finds consist of scraper-knives, mullers, perforated stones, and potsherds, 7.e., in many instances the usual domestic appliances of shell-mounds along our coasts. I may state, however, that the scraper-knife flakes, long or of moderate size, found there, are, with very few exceptions, the most worn of the South African implements of that kind that I have seen. Fig. 102 of Pl. XII. is a case in point, and although the finds sent do not include bouchers these scraper-knives or flakers belong certainly to the South African paleolithic, but Mr. Wood adds :— “The biggest implement I found here was a wedge-shaped one, 70 Annals of the South African Museum. 9 or 10 inches long, and perhaps 14 or 2 inches at the broadest. It is in the King Williamstown Museum.* Of course you understand that, so far as finding on the surface goes, the implements are most abundant away from the shell mounds: they are in plenty on the broken ridges just a little back from the beach.” Among the reasons given by Mr. McKay for attributing a great antiquity to these finds are the following :— “Some years ago in opening a quarry a very large mound of shells was discovered on the left bank of the Quigney River at its junction with the Buffalo. It was covered with vegetable soil, with trees growing on it, just like the section remaining to-day. The Harbour Works engineers have removed upwards of 375,000 cubic feet of these shells to fill up the lagoon. Any one examining what remains of this mound will find it composed principally of limpet, mussel, oyster, haliotis, and other shells of edible species, with bones of fish and birds as well as of antelopes, hippopotamus, and other mammalia, layers of ashes, fragments of charcoal, and pieces of coarse pottery. No stone implement has been found, but stones showing the action of fire are common.” “. . . to afford material for forming the railway embankment. The ground (covered by a dense bush) consisted of from 4 to 5 feet of stiff clay soil followed by about a foot of travelled gravel, that is rolled gravel having a close resemblance to shot of different sizes. A little above the gravel line a large number of stone implements were found, chiefly rubbers; a few spear-heads (lanceolate knife- scrapers) | were found with stone fragments of coarse pottery, &c. Mr. Gately’s residence stands on the rounded top of an isolated knoll or small hill which is connected with the site of East London by a narrow neck. This neck is the dividing ridge of two water- courses that nearly surround the hill before they unite and find their way to the first creek. These two water-courses have been the cause of the isolation of the hill. On its top there is a layer of black mud from 2 to 3 feet thick, below that there is from 1 to 2 feet of decomposed rock before solid rock is reached. Out of this black mud Mr. Gately has dug many implements and bones; among them stone-flakes, spear-heads (knife-scrapers),} coarse pottery, teeth and bones of hippopotamus .. . There was a time when Mr. Gately’s house was not a hill, and the implements found there are as old as that time. . . . Fringing in detached patches the whole South- Eastern coast of Africa there exists a peculiar wind-stratified * I have seen the example. It is a strongly incurved flake or flaker and of undoubted paleolithic type. t+ The words in brackets are mine. The Stone Ages of South Africa. 71 calcareous sandstone. At Cove Rock and Bat’s Cave it abounds with fossils, especially at the latter place. The peculiarities of the stratification leave no doubt whatever that this is an eolian forma- tion. Out of it I have taken three well-formed stone spear-heads (knife-scrapers) at the level of low water, in a position where they must at one time have been covered by 180 feet of this rock-drifted sand which has become consolidated into hard rock, &e.”’ Mr. McKay’s contentions as to the extremely old origin of these knife-scrapers, mullers, potsherds, &c., might at first sight appear warranted were it not that the whole “outillage,”’ or domestic appliances pertain to the midden or shell-mound type, and as I have endeavoured to show in another chapter, it is not in a midden, always void of the Chellean or Acheulean as well as Mous- terian bouchers, that we can expect to find a possible clue to the antiquity of primitive man. But Mr. John Wood, of East London, has been kind enough to send the following notes regarding the estuary of the Buffalo River :— a .; this would have been a revelation to the late Mr. McKay, who thought that the stone implements found upon the banks of the river among the fragments of sub-aerially decomposed rock (which he seems to have regarded as river gravel) had been dropped by the water margin and lain there—under a subsequent covering of soil—throughout the period taken by the river to wear its way down to its present level 80 to 100 feet below. He did not know the old channel was still another 120 feet deeper down.* My notion is that implements dropped as they were being made on the ground, and have gradually worked their way, assisted by well- known agencies—through the 2 or 3 feet or so of soil there is, as a rule here, covering the bedrock of the country. I have turned up with the spade in my garden, which is close to the river, a number of flaked stones, and once a nice ‘rubbing stone.’ “ Another mistake I fancy Mr. McKay made was when comment- ing upon the occurrence of stone implements along the beach near Bat’s Cove. His idea was that they had been dropped among the pebbles at tidal mark and then overwhelmed in a mass of sand— probably a dune of over 100 feet high judging from surroundings— which was converted into eolian sandstone. But as the spot where those implements and pebbles are found is a very circumscribed one, I should say the explanation is that we are upon the site of an old cave there, led up to by a gully or cove into which the waves heaped * A fact which came to light when the engineers bored for foundations for the bridge placed across the Buffalo after MacKay’s day. 72 Annals of the South African Museum. a collection of pebbles, and that the roofs afterwards collapsed. At present there is a detached block of eolian sandstone lying upon 3 feet or so of rolled pebbles—also bound fast by the lime—and it is amongst these the implements and bone fragments are found. ‘The coarsely cemented sand shell masses along our beach here ave hardly entitled to be called rocks; they are the cores of old sand- dunes quickly formed and quickly disintegrated ; and there are other similar spots near at hand where the occurrence I have been sur- inising is likely to be re-enacted at no distant date. ‘‘T think, too, his deductions from the ‘ finds’ in Gately’s garden are unwarranted. A shell-midden exists there lying upon the shale, the latter comes to about a foot from the surface. Anything im- perishable left lying about would surely soon sink through that shallow soil. It would be good camping ground for the Strand Loopers because of its outlook, and there was a vlei near at hand upon the site of which, by the way, our Town Hall now stands.”’ Negative again in this case is evidence of great antiquity, and IT would not have mentioned the East London deposit were it not that the late G. McKay’s account of it is on record, and may prove misleading. I have before me a large photograph of sixty-three implements collected by him. Some are large flakes that may be palzolithic, but the presence of numerous ‘‘!’ Kwés’’* whole or broken, perforated entirely or partly, large, and very small, with an accompaniment of mullers of different sizes, and also of ‘‘pygmy”’ flakes, denote but too well the domestic appliances of the shell-mound dwellers. This photograph was evidently intended to be an illustration, as representative as possible, of these stone implements, the deposition of which showing proofs of untold ages in the antiquity of man was such a cherished idea of the investigator, who, like Stow and others, unfortunately did not know how to discriminate between palezolithic and neolithic types. In this photograph there is not one figure approximating a boucher. Large flakes are there, but there are no means of ascer- taining their size. Some look weathered, others not; but on the whole these flakes are typical of those sent us by Mr. J. Wood. I have already sounded a note of warning as to attributing a very ancient or comparatively modern age to our South African * A “VKwe” is a more or less spherical stone perforated in the centre, the perforation being begun at each end and meeting in the centre (see chapter on YKwes, &c.). The Stone Ages of South Africa. ie scraper-knife-flakes, owing to the continuous survival of the type due probably to the facility of shaping or evolving the implement from any loose stone. But some, such as Figs. 102 and 103, have such a worn appearance that even a long continuation of eolian agencies would be required to impart the same, and the absence of bouchers cannot in their case be claimed for their rejuvenation. That this absence may be accidental is proved by the fact that although no paleolith in the shape of a tongue-shaped or cleaver- like tool can be said to have hitherto been discovered in the East London beach deposit, nor in the “broken ridges just a little back of the beach” where very large worn flakes do, however, occur, yet Mr. J. Wood has quite lately discovered at no great distance from there, near the mouth of the Nahoon River, bouchers, mostly huge (judging from the examples he sent us) * lying together with flakes of the kind which I consider belongs to the South African Paleolithic. And thus the discrepancy between the styles of the stone tools in McKay’s photograph are explained. The !’Kwés, mullers, some of the flakes, and the ‘‘ pygmies ”’ are recent additions deposited at or near the place where imple- ments of palolithic type had gravitated, or been manufactured very long before the Strand Looper had left there his own mark of occupation. * These bouchers are made of voleanic rock or of Karroo sandstone. They are very roughly trimmed, and all those sent us are large. None is tongue- or almond- shaped; some are pointed at both ends; in others, made of sandstone, the butt retains the contour of the rolled pebble. The flakes are of the usual knife-scraper shape, but are plainly worn. 74 CHAPTER VII. PALH ONTOLOGICAL AND GEOLOGICAL EVIDENCE. PALMONTOLOGICAL. The Pleistocene or early Quaternary period of the Northern Hemisphere was characterised by the extension of ice, and the deposit of deep alluvium in valleys and plains. There existed during this period not only in the palolithic Region or the greater part of it, but also far south of it, a mam- malian fauna including, among large forms, several that have survived in Africa and also in India, where ‘‘ some of the deposits show how long was the period during which the encroachment of some of the great African land animals into Europe and India must have persisted.” * But, as far as South Africa is concerned, no traces of Pleistocene glacial conditions have as yet been observed, although carefully searched for by the two Geological Survey parties who are mapping this part of the African Continent. The problem is still more complicated because most of our large mammals have outlived the Pleistocene, showing thus a longer survival than in Europe, while others which we know existed there in the early Pleistocene are still with us, and to all intents and purposes unchanged, although their progenitors were contem- poraneous with Chelleo-Mousterian man. Hyawnas.—Hyena brunnea, our so-called ‘‘ Strand Wolf,” lived together with Hlephas antiquus, perhaps even with EH. meridionalis, * Suess, ‘‘ The Face of the Earth,” Engl. Edit., vol. iv., Oxford, 1909. Ibid. ‘If we compare Pilgrim’s description of the fauna of the alluvium of the Godavari, and of the caves of Kamul (on the Kistna), where the remains of Manis still occur, with Boule’s account of the stratified succession in the Grimaldi grottoes (Monaco), we discover that man was a witness of this extension, both in Europe and in India. In Europe it extends into the interglacial phase of the Chellean.”’ The Stone Ages of South Africa. 75 Rhinoceros mercki, Hippopotamus major, &c., in the same manner that it does now, or did till a few years ago, with Hlephas africanus (that survivor of the elephants of the same African type to which E. meridionalis and E. antiquus belonged), Rhinoceros simus, our white Rhinoceros which several Palzontologists consider to be the same as R. mercki, Hippopotamus najor, of which H. amphibwus is merely a variety, or survivor, &e. Hienea crocuta, the spotted hyzna, which we call here ‘“ Tiger Wolf,” is now generally identified with the ‘‘ Cave Hyena ”’ (Hyena spelea), the remains of which have been found throughout Europe, from Yorkshire in England to Gibraltar, and as far east as the Madras Presidency in India. While its congener H. brunnea was seemingly pertaining more to the warm fauna of the early Pleistocene, H. crocuta kept company with the long-haired, woolly Mammoth, and the equally thickly woolled Rhinoceros tichorrhinus. Yet those two descendants of a very ancient race are now re- stricted to a warm climate, where intense cold comparable with that of the second period of the Pleistocene never prevailed.* With such cases of survival before us, it is plain that when we find in caves an accumulation of bones, some of which bear the distinctive marks imprinted on them by the powerful jaws of Hyena, more corroborative evidence is required to assign great antiquity to the deposit. Mastopon.—Part of the tooth (molar) of a Mastodon has been found in close proximity to, if not actually in, the Barkly West Deposit, of which Mr. A. L. du Toit has given here an account, and is described by E. Fraas.! This fragment is the only find of its kind known hitherto, but there is no reason to believe that the find is not genuine, and it will be remembered that the paleoliths, bouchers or flakes, found in this deposit are extremely worn (see Pls. VIII. and IX., and Fig. 105 in El eX tit.) 6 I wrote to M. Boule, submitting an account of the find of the * Hyena brunnea seems to be confined now to West Africa; Hy@na crocuta roams all over Africa, south of the Sahara to the Cape of Good Hope. + See account of Hawston Cave, Chapter XVI. + « Pleistocine fauna aus den Diamantseifen,” Zeits. d. Deutsch. Geol. Ges., 1907. § In the Bloemfontein Museum are two bouchers of large size bearing the inseription, ‘‘ Found 40 feet deep in old course of Vaal River, Barkly West.” It is probable that these implements were collected by either Stow or C. Sirr Orpen at the very spot mentioned by Mr. du Toit. 76 Annals of the South African Museum, tooth and of the paloliths, and also pointing to the fact that if it were duly proven that the paloliths were associated with the mastodon molar tooth, a much greater antiquity would have to be claimed for these artefacts. But the learned Professor of Paleontology of the Paris Museum answered my communication to this effect: ‘This contention would be true, provided the genus Mastodon did not last longer in Africa than it did in Europe, where it does not go beyond the limit of the Pleistocene. It has, perhaps, survived with you during the Quaternary (Pleistocene) period, as it has done in North and South America.” * The finds in the sands of the Fayoum seem to justify this contention. Moreover, the tooth belongs to the Bunolophon group of Mastodon in which all the North African species described by Gaudry and Depéret are included. On the other hand, this molar tooth may have come from a terrace that was once much higher than the deposit on which it was found. ANTELOPES.—If we turn to the Antilopine, a group of Mammals numbering so many species in Africa, but which, with the excep- tion of the cold-loving Antilope saiga, are not represented in the European Quaternary, we find an extinet Bubalis, B. priscus, Broom. + BuBALIS, or ALCELAPHUS, as it is sometimes called, is, like several other antelopes, recorded from the Pliocene beds of India (Siwalik). This find here is represented by one example only, and consists of the post-orbital portion of a skull with the proximal part of the left horn-core. The interest attaching to the example is that it was discovered in the banks of the Modder River, and apparently not far from where Rickard discovered implements of palolithic type. | But the connection of the Paleoliths with the remains of that extinct Bubalis is merely conjectural. This fragmentary skull is quite black, in the manner of many of the ancient bones found in liurope in river-drift. Burrato AND Horsre.—For some time now a Buffalo, which judging from the size of the horn-cores must have been gigantic, has been known from South Africa. It was described by Seeley from our example in the Museum—a part of the frontal and the * L’Anthropologie, xxi., 1910, p. 248. + Ann. 8. Afric. Museum, vii.,.1909, p. 279, ¢. fig. { Cambridg. Com, Antiq. Soe., v., 1880-1881. The Stone Ages of South Africa. 77 complete horn-cores measuring 14 feet in the curve. The locality is probably the banks of the Caledon River, in the Orange Free State. Three more, but quite incomplete, examples have been discovered since, one at Bloembosch, Cape Colony, of which more anon. Nearly completing the series of finds of Quaternary deposits, is now recorded a large horse—not zebra. Broom has lately described it under the name of Equus capensis.* The description is mainly based on the left lower jaw, although some teeth, still larger, were available, but were overlooked. These remains were discovered in the following manner: The owner of a cattle-ranch, called Bloembosch, some 20 miles from Darling, in the Cape Colony, was much troubled by the advance of sand-dunes which were threatening to cover a perennial spring, and he set about planting a certain kind of grass used to stop their pro- gress. At the foot of that dune, and not far from the spring, he came across a Skull and horn-cores of such a size that his attention was attracted, and he brought part of it to his homestead, where it was seen by Mr. H. M. Oakley, who, being acquainted with the first example of Bubalus bai, recognised at once the relic as belonging to the same animal. He informed me of the find, bringing at the same time a number of bones belonging to the Rhinoceros, Eland, Bubalis, Bubalus, Equus, &ce. Both the Geologist and the Paleon- tologist of the Museum, Mr. du Toit and Dr. Broom, as well as Mr. Bain and myself, visited the place, and we came to the same con- clusion, a conclusion arrived at from different considerations, that these bones had not accumulated at that spot by the formation of a deposit, but had been brought piecemeal to or near the water by beast or man. No complete skeleton was found. The teeth of the horse measure 198 mm., and are thus larger than those of the Hquus sivalensis. One atlas of Bubalus measures 351 mm. in width and 123 mm. in length. On the exposed floor of the dune we found the usual accompani- ment of a Midden in proximity to the water: mullers and small scrapers, borers or drills and ostrich-egg-shell beads, two brass buttons, &c.; and it is difficult indeed to come to any conclusion other than the occupiers of the midden were responsible for these bones accumulating in that spot. Neither large bouchers nor large scraper-knives were found, and if these large animals were slain by the Strand Looper—and slain by man they undoubtedly were—it must have been done by craft and by means of the tiny, neolithic stone weapons, found 77 siti. * Ann. 8. Afric. Museum, vii., 1909, p. 281. 78 Annals of the South African Museum. The horse is manifestly not old geologically, and the same may be said respecting Bubalus, but at all events in the case of the latter, whereas no one would be surprised to find its remains in marshy spots or near banks of flowing rivers, one is indeed surprised to find it in places now almost waterless, but prebably made so by a gradual rising of the coast belt, during which period the aboriginal was not acquainted with the use of palzeoliths. On the other hand also, this aboriginal would, of necessity, come to water, and the remnants of his domestic industry may have been deposited there long after the animals had been laid low. One of the femora of Bubalus had been split open as if to obtain the marrow, but it is very difficult to decide if it be the work of man or hyzena. And on the whole this restricted paleontological evidence would be of little weight were it not that in Algeria there have been found, at the bottom of a small lake called Lake Karar, paleoliths resem- bling so much in shape and even material the South African ones as to be almost indistinguishable. With them also were found small pieces of the Cape Town Flats and Karroo type, the better trimmed of which are made of flint; but their contemporaneity, according to M. Boule, who has investigated this deposit, with the quartzite paleoliths is doubtful.* The numerous bones found with the Lake Karar implements are those of Elephant, Horse, Hippopotamus, Rhinoceros, Pig, Gnu, Bubalus, and Stag. The Hlephant is an extinct kind, but belonging to one more related to the Quaternary, or even the Pliocene Elephants of Europe; others, except the Stag, are still living in the south of the African continent ; the Rhinoceros is &. sumus ; the Gnu is probably Connochetes taurinus or C. gnu ; the Horse, Hquus burchelli, &c.; and the Buffalo, Bubalus antiquus, does not seem to differ much from Bubalus baint, our extinct species. But with implements of such ancient palzolithic type one would have expected to find remains of the animals that characterise the finds at Chelles, &&. Why are not the Elephants, H. antiquus or EH. meridionalis, Hyena spelea or brunnea, the Stenon horse, Machau- rodes or Ursus spe@leuws, represented in that deposit if its great antiquity is to be assumed? M. Boule admits that the Quaternary fauna of Algeria and that of Europe do greatly differ. The former consists of species that have migrated in the manner followed by forms now northern, and the remains of which are found in the Kuropean Quaternary deposits. * M. Boule, ‘ Station Paléolithique du lac Karar,” L’Anthropologie, xi., 1900. The Stone Ages of South Africa. 79 But there are still some elements of doubt as to the relative age of this Lake Karar deposit. Those who are persuaded that the Chellean-Mousterian boucher is of extreme antiquity may agree that there is nothing to prove that the Palzoliths and the animals were deposited together during the same epoch. To those who think that too great an antiquity is claimed for the Paloliths, and that their presence or existence should be reckoned by hundreds instead of thousands of years, the presence in this deposit of large animals, mostly all living at the present time, would seem to militate in favour of their view.* But, as Boule justly remarks, no implement from among those extracted from the Lake can be included in the Neolithic +; and no remains of the animals now living in the country have been found in the Lake. In North-Western Rhodesia there has been discovered at a place called ‘‘ Broken Hill,’ in caves of very singular formation, an accu- mulation of bones of different animals, in which the Antelopes pre- dominate, and among them was found a Rhinoceros, said to be an extinct species, on the strength of a femur or tibia. In the same deposit were also found small stone implements quite similar to these occurring on the surface, in the Kafue River valley, and greatly resembling the examples found in the Matoppo caves (Bly XWIL.; Bigs.133): Mr. Franklin White informs me that he has found in these caves the unmistakable coproliths of the Hyena, identified, moreover, as such at the British Museum. But there seems to have been no absolutely reliable systematic investigation of the layers of these caves’ deposits ; and, although it is claimed that this is the first case when stone implements have been, in South Africa, connected with bones of an extinct animal— which is really not a true assertion—the evidence of an extinct species of an hitherto unknown Rhinoceros, based on a leg-bone, is not to be very seriously considered, and still less accepted, by those zoologists or paleontologists who are aware that even the dentition of these animals is not a sure clue to their specific identity. On the other hand, Dr. Broom, our Museum Palzontologist, is of opinion that among the remains received by us from this Broken * H. O. Forbes, ‘On the Age of the Surface Flint Implements of Egypt and Somaliland,” Bull. Liverp. Museums, iii., 1901. + Some have, however, been found on the surface close to the Lake, but they belong so unmistakably to the polished stone period that they cannot in the least be associated with the palwolithic forms. 50 Annals of the South African Musewm. Hill cave-deposits, there is included part of the skull of a large unknown bird. GEOLOGICAL EVIDENCE. So far Geology has not been of much use here in helping to elucidate the problem of the antiquity of our Stone Implements. But Mr. A. W. Rogers has recorded a find which is of the utmost importance. It is embodied in the Tenth Annual Report of the Geological Commission of the Cape of Good Hope, p. 293. In 1905 Mr. George Robertson, of ‘Klein Brak” River, near Mossel Bay, sent to the South African Museum some shells that came from a quarry on his property. Mr. Rogers had occasion to visit the spot and of obtaining a representative collection of the fossils preserved there. “A low ridge or terrace, rising to a height of about 15 feet above the level of the Klein Brak River vlei, marks the position of the shelly beds, and small quarries have been opened in the ridge for the purpose of getting out limestone for building. The limestone is a loose-textured, rather incoherent rock, but it hardens rapidly on exposure and appears to stand the weather well. It contains a number of shells which can easily be removed from the rock.” The species obtained are enumerated. ‘All these species, with the exception of the Cerithiwm and Calliostoma, are known from the South African seas, though some of them (Panopea natalensrs and Triton australis) do not appear to have been recorded from the coasts of the Cape Colony. ‘In the limestone I found a piece of quartzite which appears to have been shaped by man. It has a form common to many rude implements of small size found on the surface in various parts of the Colony. Mr. Robertson told me he had found round stones flattened at one end, evidently by use as crushers or pounders, in the limestone. . . . The limestone must have been formed at a time when the land stood at least 15 feet lower than now, and when the shore of Mossel Bay passed some two miles inland of its present position.” The quartzite implement mentioned by Mr. Rogers is in our Collection. It 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. That this primitive implement was deposited there in times far remote admits of no doubt whatever. The pity is that it should prove to be an instrument that can give no clue, although from its mere appearance, and without knowing its history, I would have The Stone Ages of South Africa. 81 included it in the South African paleolithic. On Mr. Robertson’s evidence, pounders of the type of our dune-middens and caves have also been found in the limestone. These I have not seen, and the presence of a vlei may account for their occurrence there, even at a depth that is, unfortunately, not stated. But Mr. Rogers’ evidence is that of a professional man, and, whether or not the mullers and the flake are contemporaneous, we have at least one example of great antiquity clearly established in the situation in which the said flake was found. ( 82 ) THE NEOLITHIC. CHAPTER VIII. AURIGNACIAN, SOLUTRIAN, MAGDALENIAN, &c., TYPES. Points, SCRAPERS OR FLAKE-KNIVES. Before proceeding to deal with the question, Have there been in South Africa periods or sub-divisions of the post-Mousterian corresponding to, or coinciding with the Aurignacian, Solutrian, Magdalenian, or post-Magdalenian? it is as well that a definition of the Mousterian type be given. Déchelete’s is one of the most concise.* “The implement is a flake usually triangular, the two lateral sides of which are trimmed again with care upon one of the faces. The part opposite to the sharp end, a part usually called base or butt, is oftener than not without secondary trimming. Noticeable is the place where the blow of the striking tool has fallen, which blow, as a direct result, produces on the reverse the bulb of percussion, a swelling which is very seldom subsequently thinned or reduced. The reverse of the Mousterian powt + is smooth, and it differs on that account from the Chellean implement; it is always lighter, being seldom more than 10 cm. in length, and is longer than broad.” Gradually, however, and sometimes too suddenly, the Mousterian outillage disappears in HKurope. It is replaced by implements of very superior technique. No longer prevails the spall with bulging bulb of percussion, very little trimmed on one face, and the other left as formed under the detaching blow; even the scrapers them- selves show traces of secondary trimming on the sides; pieces * « Manuel d’Archéol. préhist.,” i., 1908. + But the word ‘‘ point”? must not be taken necessarily to mean a lance-head, arrow-head or other penetrating implement, it corresponds even to the ‘‘ boucher”’ of the period. The Stone Ages of South Africa. 83 chipped on both faces are common; arrow- and javelin-heads, thin symmetrically produced into a laurel- or willow-leaf shape appear ; some of the points have a truly characteristic tang continued on one side for hafting. Every piece denotes a great advance in the lithic industry. The large palewoliths, whether amygdaloidal or not, bifacial, or simple on one side, are now discarded. Ultimately this Solutrian technique will also be partly abandoned, to be partly replaced by the bone industry which characterises the Magdalenian, the epoch that follows the Solutrian. This sudden advance from the Mousterian to the Solutrian is not easily explained, but in order to fill the gap the Aurignacian division has been adopted. | It is said to be characterised by a less finished style than the ensuing Solutrian lithic industry. Many pieces of the cruder Mousterian type are still found with it; the bone, in the shape of points with a transverse basal notch, begins to be used, together with ‘‘double’”’ stone scrapers with secondary trimmings (7.e., scrapers of which both ends could be used); boring tools, burins are found, for here begins the dawn of sculpture in the shape of gravings, and also that of painting in the shape of petroglyphs. The conclusions drawn from the stratigraphical evidence that led to the acceptation of the Aurignacian finds of France as older than the Solutrian, and the latter as having preceded the Magdalenian are doubtless justified. But what is sometimes true on this side of the Alps, is often erroneous on the other, and it seems as if the Aurignacian proves now in Poland and elsewhere to be quite equal in finish to the Solutrian. As far as the South African lithic methods go, and leaving out the bouchers of the Stellenbosch and other types, it may be taken for granted that a Mousterian facies is found. It cannot be said that evolution led generally to the adoption of this great advance, i.e., trimming on each side, and secondary trimming of edges, yet we have instances of crude imitation in that direction. Moreover, we have also, and apparently simultaneous or coeval, bone; and lastly, Tardonian, or early Neolithic, in the shape of ‘‘pygmies.”’ In fact, these pygmies are connected here with the ostrich egg-shell bead industry quite as much as the Magdalenian is associated with the Reindeer epoch. Had flint been available as material instead of quartzite, in- durated shale, silicious quartzite, chert or chalcedony, it is probable that the workmanship would have greatly resembled that of Europe. But on that account, perhaps also for other reasons, the South 84 Annals of the South African Musewn. African lithic industry, which, with the exclusion of the paleoliths, is so plainly of a Mousterian type has nevertheless what may be termed an African facies.* I have alluded in Chapter VI. to scraper-knives found together with bouchers in such a position that doubt as to the two being coeval is well-nigh impossible, and on the evidence of the bouchers these scrapers must be considered as being also Chelleo-Mousterian, that is to say, representing this type of culture. But on the surface, where no indication as to the age or date of deposition is obtainable, we find, seldom isolated, and often with remnants of cores or small spalls showing that they have been fabricated on the spot, scrapers, the technique of which is, I think, well illustrated by the plates accompanying or rather supplementing this paper. It is not only on the surface that these remnants, or effective implements, are met with. They are found in kitchen-middens, in rock-shelters, and in caves. For several reasons, such as difference in the shape and material of which they are made, or in technique, they may be divided into sub-types, which do, however, occasionally commingle and which I term the ‘ Littoral” and ‘Inland Districts ” respectively. Inttoral Type.—In the Cape Peninsula there are many small depressions which contain fresh water at certain times; round these vleis one is sure to find implements, mostly small, like those of Figs. 118 and 120. Occasionally examples of the best Solutrian type met with in this country are also discovered. The lance-heads of Fig. 110 show on the obverse and reverse a secondary chipping so carefully done, an advance so great in the process of hafting, that it becomes doubtful if the technique is not one that is altogether foreign to that of the aborigines. Have we here, then, a corroboration of the famous, yet contro- versial, Egyptian Periplus? Have these splendid points which so very greatly resemble the best specimens from the Fayoum been made on the spot by mariners wrecked or temporarily stranded, or by adventurers seeking pastures new ? Rare indeed are these entire specimens, nor are the fragments numerous which show so well that when used as lance-head the local material broke easily at the edge of the hafted part, a fracture due perhaps to the thick hide of the Hippopotamus, Rhinoceros, and Elephant, the remains of which are not uncommon in the parts * Cf. Foureau, Mission Saharienne, Paris, 1905. Zeltner, ‘‘ Note sur le Préhis- torique Soudanais,’”’ L’ Anthropologie, 1907. The Stone Ages of South Africa. 85 where the implements are occasionally met with (cf. obverse and reverse Cuts 1, 2, 3, 5 of Figs. 110 and 112). So far, only one example of the highly finished javelin-heads, Cut 8 of Fig. 110, has been found beyond the Cape Peninsula, namely, in the Berg River valley, close to Simondium, where paleoliths abound. And together with these lance-heads trimmed on both sides we find examples of truly Mousterian type, that is to say, trimmed on one side only (Figs. 111 and 113 showing both sides). The material is the same, the technique appears to be similar, and yet only one face is pared. Occasionally great care is devoted to this part, as in Cuts 2 and 5, but the other face with its largely bulging bulb of percussion is as Mousterian in shape as any example we possess from the Moustier, or is to be seen in the St. Germain Museum, in Paris. Unlike the highly finished lance-heads which have hitherto been found singly only, these last-mentioned implements are found together with rude ones made of the same hard surface quartzite, a material forming at the present time. Most of them obviously are scrapers. The notched Cuts 1, 2, 3, 5 of Fig. 118 are interesting because they probably fulfilled the same purpose as those in Fig. 138, which are from the Karroo and elsewhere, namely, paring into shape the piece of wood used for making a bow. The serrated edge of Cuts 8, 9, 10 of Fig. 118 suggests that they were used as saws. While the notch of Cuts 1, 2, 3 are on the left of the obverse part of the scraper, that of Cut 4 of the same Fig. is on the right, as if the maker was left-handed. Tools of this rough type, made of the same material, are to be found all along the western coast from Simon’s Town to Port Nolloth, but seldom far inland. They abound near the middens of that part of the coastal belt, and they are found there together with the “ pygmies.” On the Cape Flats they are, as I have already said, very numerous, but they always overlie the iron-stone gravel, whereas under this gravel, or in the layers of the gravel, small bouchers, or large flakes are occasionally found. Rough indeed is the method of manufacture of the Cape Flats implements. As far as I know, none has been found that shows unmistakably traces of secondary trimming, with the exception of the highly finished examples that are figured in Pl. XIV. and the tiny nucleiform scrapers, Fig. 130. But with these others are found that plainly show decadence in the exercise of the skill, until the more primitive style has again been reverted to. Of these examples 86 Annals of the South African Museum. it is very doubtful if, when blunted by use, the opposite cutting or scraping edge was ever made to serve. Inland districts* Type—Ift we leave the littoral and examine the Mousterian type of the inland districts, we find analogous cases of crude and better finished tools; but if none has as yet been found to be completely trimmed on each side, that is to say with the complete removal of the bulb of percussion, we meet occasionally with well-finished ones. Cut 1 of Fig. 114 from Swazieland cor- responds to Cut 2 of the same Fig. from the Cape Flats; the long knives of Fig. 106 from Aliwal North are very elegant and very serviceable tools, and Mr. J. P. Johnson has figured a scraper from the Embabaan Valley, in Swazieland, partly chipped on the reverse side but with the bulb of percussion left, which is probably one of the finest pieces of the ‘‘ up-country ” type as yet found. But the material is now changing. It is no longer the silicious sandstone of the Cape Flats which is such an indication of the locality ; it is an indurated shale that cleaves readily, lydianite, cherty material, jasper, chalcedony, or white quartz, and now we meet with a secondary kind of trimming which can compare only with that of the Aurignacian turning into Solutrian. In the Stormberg, Herschell, the Free State, Griqualand West, a part of the Western Karroo, and even occasionally on the Western littoral of the Cape, we meet scrapers like those of Figs. 123 and 128, in nearly all of which this trimming is at one end.t The “ bord abattu’’ is not accidental, and in many examples, especially those of moderate size, it is the thicker part of the flake which is thus reduced into a bevelled shape. When moderately large, these scrapers answer admirably for re- moving the particles of flesh adhering to a flayed skin, but was this their intended purpose? They greatly vary in size, and examples of Figs. 130 from the Cape middens, minor tools among already minute ones, could hardly be said to have fulfilled this function. Cut 1 of Fig. 131 from Smithfield, Orange Free State, is one of the very few examples known to me that show secondary trimming on more than one edge. In Cuts 3 and 4, given here for purposes of com- parison, the chipping of edge is not intentional, but has been produced by use. Cut 2 of the same figure is even more evenly treated all round than Cut 1 and is from the same locality (Queen’s Town, Cape Colony) as Figs. 125, 126 and 127, but in spite of the secondary trimming treatment it cannot be compared with a * This type is not restricted to the inland districts, it is occasionally found along the coastal belt east of Algoa Bay. + Somewhat in the manner of some of the Magdalenian scrapers. The Stone Ages of South Africa. 87 Solutrian scraper. Fig. 125, both faces of which are represented, has undoubtedly been worked intentionally into a “tang,” but not so the serrated tool Fig. 126 nor Fig. 127, and it is impossible to decide whether we have here a case of progression or one of re- gression. Two pedunculated tools nearly similar to Fig. 125 were discovered at Cradock, and another, which was unfortunately lost, came from Queen’s Town. The bevelled scrapers are never large, especially if we compare them with the flake-knives, but Fig. 132 of Pl. XVI. is quite an exception. It is from Matatiele, in the Cape Colony, and the secondary trimming on each side of the edge is very conspicuous. It is difficult to understand the purpose for which this large, moderately thin implement was intended. It is 210 mm. long, 85 mm. wide, and 23 mm. thick. It may have been hafted laterally in the manner of a Basuto battle-axe. ° These bevelled tools are seldom found in the Western part of the Colony, in fact, I know of none of the same size, although the diminutive ones in Fig. 130 approximate them; but they abound in the Karroo, and are met with in the Orange Free State, Griqualand West, and in the Southern parts of the Transvaal, but I have not seen any from Southern Rhodesia, Natal, or the N’Gami Region, and, so far, it is not unsafe to say that the technique is that of the ‘“‘pygmies,’’ many of which have, in spite of their small size, been subjected to a secondary trimming, more difficult, perhaps, to execute even than the bevelling of the nucleiform minute scrapers of Fig. 130; cf. also the bevelled agate chip, Cut 1 of Fig. 135, Pi; XVI. Yet, in spite of the undoubted skill displayed in the production of these stone tools, the progress seems to have been checked; the manufacturer has never conceived the method of paring the tool on both faces as the Aurignacian and the Solutrian makers did, and thus on pure lithological grounds it may be asserted that the South African industry, which I term Neolithic, has, so far as the “ points ”’ are concerned not proceeded beyond the Mousterian type. But, just as in the case of the Chelleo-Mousterian type, bouchers and scraper-knives, the evidence as to the age of those scrapers which we find nearly always associated with minute flakes and diminutive piercers, borers, or drills, which I range among the “pygmy” or “‘ Tardenosian”’ type of Rutot, is unsatisfactory. It cannot be ascertained, because, whereas there is no proof that the paleeolith, either Chellean, Acheulean, or Mousterian was used by the Aborigines in historical times, the same cannot be said 88 Annals of the South African Museum. of many of the smaller implements that retain still a Mousterian facies, as the following examples show Figs. 124 and 129 are made of lydianite; the first example is greatly polished, in all likelihood by eolian agency. Fig. 129, of a shape and manufacture absolutely identical, is made of green bottle- glass. It is from Pearston, in the Cape Colony, and was found with other rough scrapers. We know that the town of Pearston has not been long founded. Some eight years ago Rev. W. A. Adams submitted to me several examples of bevelled scrapers greatly resembling Cut 12 of Fig. 128, which he collected at du Toit’s Pan, close to Kimberley, in Griqua- land West. They are made respectively of green bottle-glass and of thick plate-glass. They are as carefully bevelled as the indurated shale, agate, or quartz implements of the same style of manufac- ture. Now Kimberley and du Toit’s Pan date from 1870, and it is highly improbable that plate-glass was introduced there before 1880. When treating of the pygmy flakes or borers, or of some of the Cape shell-mounds or sand-dune middens, it will be shown that implements have been found lying together with various objects of European manufacture. That South Africa was partially in the Stone Age some four or five hundred years back, and most probably even later than that, must needs be accepted as an established fact. A characteristic trait of the culture of the Aurignacian-Solutrian man is his ability to reproduce by pigment on parietal (rock) surface, or even occasionally on water-worn stones, images of the scenes that appealed to him, or to grave on the walls, or the floor of the recesses he occupied, and more numerously perhaps on bones of mammals, figures, often realistic indeed, of the animals he chased. We have here also presentments which compare with the artistic skill of the Solutrian man. Our rock-paintings are inferior as a rule, but not always, to those found in Southern France or Spain, whilst our rock-gravings are, in many instances, far superior. But comparison of the bone- or stone-implements found in, or near, the famous Altamira Cave, with those of our rock-shelters or caves where paintings or gravings occur, reveals a great dis- similarity, greater, however, in the stone than in the bone tools. The gravings here are in the open. In a few cases we can asso- ciate with them palzoliths of the Chelleo-Mousterian type; also in one instance gravers of the Mousterian long, scraper-form. The Stone Ages of South Africa. 89 With our paintings,* on the other hand, are associated simple, small flakes, of a type and size which are met with nearly every- where, but no palzeolith. In the Matoppos, Southern Rhodesia, the paintings are getting fainter every year, and the place is too much exposed for these glyptics to be of great antiquity. The stone relics left by the aborigines, the authors of the paintings, are represented in Pi XVIL, Hig. 135. Fig. 134 represents the implements discovered at the foot of a rock-shelter, hardly, however, deserving this name, on which are found remarkable paintings, at Modderpoort, Orange Free State. Fig. 136 is the representation of forms similar to those of the three preceding Figs., found close to a cave-shelter full of paintings. A comparison with the shape of these tools and those of the Cape Flats, especially that of Cut 2 of Fig. 136, with the same number in Fig. 120, Pl. XV., is very instructive. The material used differs, but the technique and also the purpose is the same. The notched scraper, Cut 10 of Fig. 134, from Modderpoort, has served the same purpose as the more broadly or crescentic ones in Fig. 138 from the Cape Karroo, and the Cuts of Fig. 135 from a Free State rock- shelter where no paintings occur show that, with the exception of Cut No. 1, a bevelled edge specimen, the others can match in size and shape many that have been found in rock-shelters where paintings are displayed. The technique, poor as it is, is the same, and however old the survival may be, there is no reason to endow these relics with a great antiquity, because the paintings with which they are connected cannot be of any antiquity. Nevertheless, in spite of the evidence adduced, it would be unsafe to conclude that only these most primitive chips are co-related with artistic skill. Text-fig. 8 is one that goes far to disprove this theory. It is that of as fine a Mousterian-type scraper as any found in South Africa. It was discovered in a cave full of well-preserved paintings. With it was a small, round, dolerite pebble, of the same size as Cut 1 of Fig. 185, and which, from its appearance, might have been a muller for preparing paint or poison. On the floor were a number of broken reed-shafted arrows, some retaining still the cement that served to fix the feather; the sinews binding the notch are also well preserved. Some of these arrows, more diminutive than the others, * These paintings are usually found in, or close to, rock-shelters. None has, as yet, been found in deep caves, 90 Annals of the South African Museum, are tipped with long pieces of hardened wood, and they, together with a fragment of a bow, plainly pared by a stone scraper, but almost tiny, seem to indicate that they were used for practice by children. These broken reeds tell a tale: the storming, or destruction by surprise, of a lair of Bush people in recent time. And therefore this highly finished tool dates from yesterday. Its finish contrasts singularly with the rough scrapers in Figs. 133, 134, 135, 136, or 138. Yet the culture of the people is the same as that of which the other rock-shelters give evidence. Associated with the scrapers, notched, lanciform and saw-like, which, as I have already remarked, and as the illustrations plainly show, are nearly always of a moderate size, and often small, there is an admixture of much smaller ones which show no secondary trimming in many eases, Figs. 135, 151, 140, whereas those of Figs. 144 and 149 show this kind of trimming in spite of their minuteness, as do also the examples in Fig. 143. The facies is doubtless not unlike that of many burins or drills of the Magdalenian period. The Stone Ages of South Africa. aL It is generally admitted that stone implements of the type here figured were held in the hand. But that some were hafted, and in that manner made more serviceable, is proved by Fig. 150, of Pl, XIX. Two examples of that sort are now known. The specimen figured is in the Museum Collection. Both were found in the sepultures of the Outeniqua caves. Details of the find of our example will be found in Chapter XVI. The wooden handle is greatly decomposed. Had it not been care- fully protected from atmospheric action from the time of its dis- covery, it would have crumbled to dust; even the first exposure to the air caused it to break. It is so saturated with sodium chloride, that even in the few days it was removed from its glass case for reproduction by photography, an efflorescence of salt was produced. The ovoid attachment of gum-cement is very large in proportion to the size of the handle. The stone part is roughly trimmed, being somewhat in the shape of a core, and is embedded in the cement to about one-half of its own length, as far as one can judge. The very sinuous edge had been broken before burial probably, as indicated by the comparatively fresh fracture. My first impression was that this hafted tool was that of a ruler or medicine-man, somewhat on the lines of the Obdton de com- mandement, and I am inelined still to look upon it in that light, because as a serviceable cutting or graving tool it could not prove of much service. This mode of attachment which so greatly resembles that of the Australian Aborigines, must, however, have prevailed perhaps more commonly than we imagine. It would certainly facilitate the handling of the very small chips on which care has certainly been devoted. * It has also been resorted to in the three arrows represented in Fig. 142. On Cut 1 is a lozenge-shaped, flat lump of cement of apparently the same composition as that of Fig. 150; on one side there remains a small chip of white quartz inserted with the sharp edge outwards from the middle part of the cement to near the front, but without reaching it; the chip set on the opposite side would reach the very tip; and the other side being similarly provided would thus complete a sharp triangle, but it is missing, having dropped out of the fastening material. Cuts 2 and 3 are similarly fashioned, but only a tiny piece of white bottle-glass forms * Cf. Etheridge’s and Whitelegge’s ‘‘ Aboriginal Workshops on the Coast of New South Wales, and their Contents,’’ Rec. Austr. Mus., vi., 1907. 92 Annals of the South African Museum. the tip of Cut 2; and that of Cut 3, made also of a small piece of glass, is partially broken off. These examples should not, however, lead one to believe that all these moderately large scraper-knives or others were hafted in this manner. Many examples have been found in situations which, by the accompanying relics, are proved to be not old enough for the cement to have crumbled away from the action of time. A good instance of this contention are the objects represented in Fig. 137, and the tools used for producing the same. We can date the intro- duction of gun and spade in the district when the examples come ; and so recent is it that it is impossible that the gum-cement should not have been preserved had the small tools been hafted. (93) CHAPTER IX. TARDENOSIAN TYPE. ‘¢ Pyamy SCRAPERS OR DRILLS.” An indubitable fact is the tendency in the manufacture of the points, scrapers, or flake-knives towards a reduction in size. We may safely accept the explanation, that the white quartz or agate examples represented in Fig. 151 are intended for the side- pieces of arrow-heads. Those of Fig. 140 may likewise be considered as such. Some of them may have been points also, because the bulb of concussion still left is too small to prevent them from being held firmly in the gum-cement. But if we turn to Fig. 143, the use of these small tools so carefully trimmed on one side, tools which are not of the true crescent-shaped European type, whereas Cut 1 of Fig. 149 is nearly so, remains con- jectural. Not conjectural, however, is the purpose of the small, even, minute implements represented in Fig. 144. They, like the upper row of Fig. 146, are the tools used for boring holes in the tiny beads of ostrich egg-shell which are represented in Fig. 146, and were found together with the borers. I purposely term them borers or drills instead of punchers, because the perforation is executed from both sides. These flat beads are represented in all stages of manu- facture in Fig. 146, and all the borers show traces of secondary trimming apart from the serration- produced by use. But for the preliminary paring of the fragmented egg-shell, small, simple scrapers have also been used, because the examples in Fig. 141 are found where the aborigines set their workshop, and together with these scrapers are also found the nuclei, or cores, of Fig. 139. The examples here figured are from the sand-dune middens of the Cape Peninsula and other places on the coastal belt. But they are not restricted to that part of South Africa. These ‘‘pygmies”’ are found all over the country, and their 94 Annals of the South African Musewn. primary purpose is the paring and boring, in a word, the making of ostrich egg-shell discoidal beads. Few rock-shelters of inland districts are without them. The specimens in Fig. 149 are from the Orange Free State; in Fig. 145 from the northern parts of the Cape Colony, &c. In the enormously wide stretches of sand which form the coastal belt of German South-West Africa, the Bushman has still, among his most restricted belongings, borers like those of Fig. 147. Made of an almost intractable chalcedony rock, like that from which the Zambesi River scrapers, Figs. 119, 121, 122 in Pl. XV., and even paloliths, are manufactured, he has with infinite pains suc- ceeded in turning them roughly, perhaps, but effectually neverthe- less, into a drill for boring the egg-shell bead. But these borers are not always so uncouth. From Conception Bay, and round Walfish Bay, we have finished drills, some of jasper, some silicious, as perfect as any in Fig. 144 of Pl. XIX. Exposed, however, to periodical winds, that blow without intermission in the same direction and with the greatest violence for several months of the year, these tools have been reduced in spite of their minute size to the wind-worn shape known to Geologists as ‘‘ Drei Kanter.” Yet this wind-worn shape does not afford any clue as to the age, because the violence and the duration of the wind is such that the process might take only one year to produce.* But it is not there only that the violent winds smooth the faces of these small borers. Fig. 148 represent similar implements from the sand-dunes of Mossel Bay which have also been turned into “ Dreikanters.” And thus, these ‘‘pygmy’”’ implements of South Africa are con- nected with two industries, the one for tipping poisoned arrows, the other for ostrich egg-shell bead-making. As burins they might have also been used, but it must be remem- bered that the only ornamented bones or bone tools found hitherto are Cuts 1 and 2 of Fig. 194 in Pl. XXVI. This seems to indicate that graving on bone was not often resorted to. The problem presented by these ‘“‘pygmy”’ tools occasionally with, but often also without, the secondary chipping is as much complicated as that of the paleolithic bouchers. They are found in * It is in these sands that diamonds are found. Mr. W. M. Adrey, who pre- sented these specimens to the Museum, tells me that the noise produced by the small pebbles displaced by this wind is quite audible. It is not only the imple- ments which are reduced to the Dreikanter shape, but also all the jasper, agate, and other pebbles. The Stone Ages of South Africa. 95 Africa from South to North, and were first noticed, if I mistake not, in India. They occur in Europe, in Austria, England, France, &c. The resemblance of ours to the European is such that M. Cartailhac to whom I submitted some examples, wrote: ‘‘I admit that it is greatly surprising to find in your series small pieces akin to ours. Your minute flakes with secondary trimmings (petites lames a bord abattw) might be mistaken for (se confondraient avec) those of our French deposits ; but the neolithic of Japan affords surprises of a similar kind.” That some of the examples figured in Pls. XVIII.-XIX. were intended perhaps as burins or gravers, but more certainly as borers or drills, and also as parers, of ostrich egg-shell beads, or cutting- pieces to be fixed by gum-cement to poisoned arrows, I have already explained. But it is not certain—at least, the finds made hitherto do not substantiate the beliefi—that these pygmy implements from elsewhere than Africa served the purpose they did in South Africa. The ostrich was not found in early neolithic time in Europe, nor can I find, in the mural paintings or engravings of Central and Southern Europe, figures that can be said to resemble it ; moreover, no trace of this bird has been as yet discovered in those deposits which we ascribe to the early, middle, or late Pleistocene of Europe. It may, therefore, be taken for granted that the boring or paring tools were not fabricated in the Northern Hemisphere for the pur- pose of making ostrich-shell beads. But in the Paleolithic deposits of Europe, shells, marine or fluviatile, were perforated in a manner implying clearly that they were suspended or strung for ornament,* and teeth were bored for the same purpose; and how could they have been so treated but by the use of these very drilling tools. The deposits or ‘‘stations’’ where these European “ pygmies’’ occur are usually near the sea, along rivers, or close to lakes and ponds. It is claimed for them that some were used as hooks, others to barb harpoons. It is clearly proved that, as far as the French pygmy deposits are concerned, whenever culinary remains are pre- served, they show that the makers fed on fish and molluscs. The crescent-shaped implements of Pl. XVIII., Fig. 143, both edges of which are, however, very sharp, and thus differ from the quartier dorange shape of Fig. 149, are found close to the sea, * A kind of short petticoat covered or adorned with the shells of Nassa neritea was discovered with the skeletons of two young Troglodytes inhumed together in the upper floor of the Grimaldi caves; two found on the lower floor, which proved to be of the negroid type, bore round the head and the wrist, respectively strings of the same shell. 96 Annals of the South African Museum. or in the neighbourhood of “ vleis,”’ in the South-Western part of Cape Colony. Mr. J. P. Johnson has figured some “highly charac- teristic crescents,’ as he terms them, in what he ealls the river deposits of Riverton Island, on the Vaal River. It is not at all unlikely, with the method of gum-cement which we now know obtained, that these semi-crescent-shaped tools with one cutting edge or two were used here also for barbing arrows or per- haps harpoons. No evidence that they were used in that manner has so far been obtained, it is true, either here or in Europe, but that such a form of attachment existed here is proved by the arrows in Pl XViIli Big, 142: That the majority of our pygmies are borers or drills, or parers of ostrich egg-shell discs is patent in Algeria and in the Soudan,* whence many specimens are exact duplicates of ours, the “ pygmies”’ served there the same purposes, and were manufactured in all likeli- hood by the same race of men. These pygmies of Kurope are considered by some to belong to the initial period of the Neolithic Age. With us they have lasted to this day, and are most usually connected with the ! kwés, grooved mortars, bone tools, pottery, &c. In the Aurignacian period of Europe, however, we begin to find chips and nuclei which are also singularly like South African ones ; the published figures+ of the finds in the stations of Hundesteig, Krems, Southern Austria, correspond altogether with those of our Figs. 139 and 145, and many with those of Fig. 140. * Cf. Debruge, Les burins et les silex de forme géométrique de 1l’Atlas ‘“‘T’Vhomme préhistorique,’’ 1905. ‘‘Les stations sahariennes ont fourni a M. Foureau de nombreux fragments d’cenfs d’autruche les uns trés délicatement gravés, les autres taillés en forme de petites rondelles perforées a leur centre. Pour travailler ces coquilles d’cuf, louvrier eut fréquemment brisé son ceuvre avant de l'avoir achevée, et les petits silex a dos abattu répondaient fort bien a cet usage." Verneaux, in Foureau's Mission Saharienne, pp. 1109-1110. + Hoerns, ‘‘ Der diluviale Mensch.,” p. 119, Fig. 44. (97 ) CHAPTERe X: BonE Toots AND STONE SHARPENERS. Bone- and stag-horn tools were greatly in use in the Solutrian- Magedalenian times, but bone tools have not prevailed in the South African lithic epoch. This is the more surprising in view of the undoubted Aurignacian and Magdalenian facies of many of the stone pieces. The scarcity of the find seems to justify the assumption that the utility of bone as a material was not quite realised. These bone-tools can be arranged in three categories: Knives, awls, arrow-tips. No bone ornament, either bored for suspension or otherwise, has, as yet, been discovered that can be connected with the stone industry. KNIVvEs. The knives are very scarce. I know of four examples only, and all have been found in shell-mounds in the open, or in caye-shelters. They were possibly used for detaching the molluscs from univalve or bivalve shells. Cut 2 of Fig. 193, Pl. XXVI., is very blunt, and bears numerous traces of short, sub-transverse incisions made plainly by a stone scraper-knife. These incisions must be posterior to the shaping of this tool by polishing. Cut 1 of Fig. 172, which, like Cut 2 of Fig. 193, is from one of the Outeniqua-Tzitzikama Caves, has the edges plainly ground. Cut 172 may have been a knife, although it may also have been used as a spear-head. It is made from the rib of a ruminant, is thin and smoothed on one face only. I know of another exactly alike, but smaller, found with a number of bone awls and arrow shafts in a rock-shelter. Awus AND Arrow Pornts. Awls or bodkirs are more numerous than either knives or arrow-tips. 98 Annals of the South African Museum. This is the more surprising considering that these awls were used for stitching skins together—which is undoubtedly the case—for the aborigines had close at hand the hard, long thorns of several Acacia- trees, which would have proved quite as effective. But custom dies hard, as the whole account of our South African neolithic industry proves but too well. An examination of these awls or bodkins shows that those from the cave-shelters of the seaboard are of a better finish and also the Bre On, S >t? ° 6 zp 52 aby = ad 3 24 1 he Bre. an 1. LOOSE SANDY LIMESTONE. 2. LAYER OF BROKEN SHELLS. 3. LOOSE SAND WITH MANY BRANCHING TUBULAR BODIES OF HARDER ROCK. 4, POSITION OF HUMAN BONES. Mr. A. W. Rogers, Director of the Cape Geological Survey, who kindly accompanied me to investigate the deposit and disinter the skeleton, could not express an opinion as to the actual age of the former, no shells or bones of extinct animals being found in it. The crest of that midden is very much higher than the present level of the sea. It was not dominated at the time of our visit by the crescentic sand-dunes, but the insidious creeping sand-waves were gathering round. One month afterwards all signs of its existence were absolutely obliterated. Separated from the seashore, where Mytilus Turbo and Patella shells abound, by a long stretch of undulating sand-hills, continually 138 Annals of the South African Museum. exposed at all seasons to the blast of that ever-moving sand, it does not seem probable that the shell-gatherers chose the top of this sand- dune to settle there even temporarily. Itis true, however, that under the layer of shells, and under the body, there was sand, but we found also there tubular lime-coated roots, or trunks of trees or bushes, of a size plainly indicating that vegetation of a robust kind had in previous years consolidated the original sand-heap.* I have come to the conclusion that this sepulture is very old, because the immense accumulation of sand postulates a great age. The extent of the midden must also have been considerable ; for we found about a mile from the place large potsherds exposed by the wind. MIDDENS OF THE CAPE FLATS. On the Cape Flats, beginning at about 5 miles from Cape Town and reaching almost to False Bay, near Strandfontein, is a succes- sion of longitudinal depressions in which the rain-water accumulates, and which are usually dominated on one side by sand-hills often of considerable magnitude, and as shifty as those occurring in a line more or less parallel to the seashore. When one, or more, of these dunes is removed bodily, although gradually—a thing which happens at times—one finds here and there on the floor heaps of accumulated sea-shells, mostly reduced to minute fragments, and with them a profusion of small pieces of ostrich egg-shell, as well as perforated discs of the same material manufactured into beads. These beads are irregularly rounded outwardly, and most of them are of much smaller size than those met with farther north. The very much reduced periphery, in proportion to the size of the perforation, required consummate skill in the making; the number of broken, and therefore spoiled, discs bears out the supposition as to the difficulty inherent to the making. Fig. 146 represents these egg- shell beads in all stages of manufacture. At the top of the Fig. are shown the borers, also found a siti, and on the right three glass beads of Huropean manufacture. Some of the implements represented in Fig. 144 were found there, * «The sand iscalcareous. As it shifts before the wind it in many places buries bushes growing near the shore. These die, and their stems buried in the sand, decay, and in doing so set free a certain amount of acid which brings about a solution and redeposition of caleareous matter in the sand. The sand immediately surrounding the stems is thus cemented into a solid mass which encrusts the remains of the bark. The wood decays away, and a pipe with a wall of cemented calcareous sand is the result”? (Moseley, ‘‘A Naturalist on the Challenger; Cape of Good Hope,”’’ p. 149). The Stone Ages of South Africa. 139 as well as the scrapers of Fig. 141, and most of the crescent pygmies in Fig. 143. With the stone implements and the fragments, more or less worked, of ostrich shells there were discovered a few brass buttons ; clay pipes, evidently of early Dutch manufacture, with the stem broken off close to the bowl; several lead musket-bullets, together with undecorated potsherds of native manufacture; also querns, and smithies with iron ore and stone hammers *—! kwés split into two, &e., showing that the dwellers, or the later dwellers, were acquainted with the use of iron, but had still retained the stone for certain purposes. These sand-dunes of the Cape Flats have not yielded many bone implements, or the grooved polishing stones intended for sharpening them into awls or arrow-tips.t In the neighbourhood of these sand-hills we found remains of the Rhinoceros keitloa and of the Elephant. The pottery is of the usual simple type, but at Strandfontein there were found sherds with a regular series of perforations of a unique kind (Pl. XXVI., Fig. 2). THE BLOEMBOSCH DEPOSIT. Strictly speaking, this deposit should not come under the denomi- nation of kitchen-midden, because we did not discover there any layer of accumulated shells. Except for that, however, the conditions appear to be the same as in the middens of the Cape Flats, only that we met for the first time paleontological evidence of the utmost importance, intimately connected with that deposit. Bloembosch is about 25 miles from Darling, Cape Colony, and 3 or 4 miles from the seashore, from which it is separated by intervening sand-hills. At one point is a “fontein,’’ which does not seem to be intermittent, but alongside of it, and threatening to cover it ultimately, at the time of our visit, was a huge sand-hill, disconnected for a long distance from the coastal ones. It is here that the remains of an extinct buffalo (Bubalus bainz), and horse (Hquus capensis), as well as of other large mammals still living were discovered. As has been seen in Chapter VII., dealing with the paleontological aspect of the Antiquarian question, it soon became evident that the animals had been brought piecemeal to the fountain by their slayers, because parts only, not whole skeletons, were found im siti. See Chapter XX. + No. 3 of Fig. 172, which is one of the very few bone implements found there, represents a knife or spear-head made of the rib of a mammal, and greatly resembling in shape the bone knives or spear-heads of the lacustrine deposits of Switzerland. 140 Annals of the South African Museum. On the floor of that partly removed sand-dune Mr. J. M. Bain and I discovered small cores, scrapers, and borers similar to those of Pl. XVIII., Figs. 139, 140, 141, 144 (some of the pieces are there included) ; also a few of the crescent-shaped pygmies (Fig. 143) ; perforated ostrich egg-shell beads, but of a somewhat larger size than those of Fig. 144, and not so carefully concentrically rimmed ; rough scrapers of the type of Figs. 118 and 129 (Pl. XV.), also made from the Cape Flats surface quartzite ; larger scrapers, but also of the Cape Flats type; a small mortar; nuclei of different kinds of rock ; broken ! kwés; a small grooved stone for sharpening awls, &e. Here also we found two brass buttons of the same pattern as those occasionally found in the Cape Flats middens, as well as a few pieces of Oriental china, derived probably from the wrecks of some Dutch East Indian merchantmen sunk in Saldanha Bay, a place but little distant from Bloembosch. When fresh water is as rare as it is in these parts, it is not at all surprising that successive generations of aborigines should have resorted to a locality where this necessity of life is found all the year round. But diligent search, three times repeated, failed to reveal any other kind of implement that could lay claim to great antiquity—no boucher or vestige of it was met with, no long knife-scraper showing sign of old age. Six months after the last search, that sand-hill had crept again to the floor on which the find of implements and bones was made. It is now fixed in position by a plantation of ‘‘marem ”’ grass (Psamma arenaria), to the great delight of the owner and the chagrin of the Antiquarian. From the account of this deposit, it will be seen that the domestic utensils or tools are the same as those occurring in the Cape Flats ; the pygmies are alike, and soare the largerimplements. The culture is therefore the same, and the race in all likelihood identical. CarpE Town MIDDENS. In Cape Town itself are the remains of what must have been an extremely large midden, extending from the very seashore close to Grainger’s Bay, covering part of what is now the golf links, and abutting on what was once the Green Point “vlei,” a board and somewhat deep depression in which, in former times, the accumula- tion of rain-water lasted during the greater part of the year. The whole of this midden has now been removed or levelled—my house is built on part of it—but there are still to be found, and actually on The Stone Ages of South Africa. 141 the very edge of the high-water mark, enormously large examples of marine shells, of limpets mostly Patella, but also of Turbo, many of them of a size and thickness that are certainly not now met with in the neighbourhood. Close to the Mouille Point Lighthouse (now disused) the removal of sand for building purposes left exposed lately a heap of calcined stones, the examination of which showed plainly that a kitchen had formerly been erected there under shelter of the higher sand-hills that protected it from the prevailing boisterous south-east winds which rage so violently in these parts for half the year. The imple- ments found near these calcined stones tell plainly of the occupa- tions of the dwellers or makers. They are querns or mortars, mullers or pounders, a few scraper-knives of a very rude manufac- ture, together with a thick accumulation of all kinds of local sea- shells, and also a sprinkling of undecorated potsherds. At a distance from this very spot were found close to each other six mortars,* one with an artificial depression on each side, and with the accompaniment of pounding utensils. Some of these mortars are so heavy as to preclude the possibility of their removal by nomadic people having no means of transport. They were thus permanent fixtures which the natives used while camped on the spot, or when they returned to it. (I know of a group of some sixteen such mortars near the edge of a small lagoon in the Cape Peninsula.) There is no evidence that this midden is very old, yet, singularly enough, it lies on a raised beach, which, however, unlike that of the Klein Brak River, has yielded me, so far, no traces of artefacts. The type of these mullers or mortars, !kwés and scrapers in nowise differs from those found in similar situations, but the absence of bone tools, and the relative paucity of scraper-knives might be taken to imply that the dwellers had the use of iron. I have, in truth, met with no smithies there, but it must be remembered that vestiges only are left of this once extensive midden, and also that the smithies of the Cape Flats are near by. This shell mound is not the only one met with in the vicinity of Cape Town. The road that skirts Table Mountain towards Hout Bay has been cut through three such shell-mounds now covered with vegetation, while at Hout Bay itself there is an extensive deposit at a small distance from the mouth of the small river. The same culture as evidenced by the remains of this Cape Town midden prevails along the western coast of South Africa from Cape * For the shape of these mortars or querns see Pl. XXII. 142 Annals of the South African Museum. Town to the mouth of the Orange River, as known to me, and east- ward from Cape Point to the Kei and Tugela Rivers. The shores of Saldanha Bay are huge shell-mounds. An earthen- ware pot still more conical than Fig. 180 (Pl. XXIV.), which comes from the sands of Port Nolloth, has been found at Langebaan. Through the kindness of Mr. John Wood I am able to give an illustration of such a shell mound, exposed by a railway cutting in INixes, YPY the vicinity of East London, Cape Colony. Its extreme length is close upon 90 yards, and it would have been originally 40 yards or more across. The maximum thickness is nearly 4 feet. From the examples here given, but taken at random, of these open-air shell mounds, it is manifest that the culture they display is generally uniform. The implements are of a Neolithic type ; not the Neolithic of Europe, America, or India, but a South African type corresponding in part to the former. (143 ) CHAPTER XV1. CAVES AND RockK-SHELTERS. Tur HawstTon CAVE. It is not known if to the north of Cape Town the caves or rock- shelters along the coast were once occupied by aborigines in the manner of those of Outeniqua or Tzitzikama. But rock-shelters or caves are found along the southern coast between Cape Town and Mossel Bay, Cape Colony. At the extremity of the Cape Peninsula, towards Cape Point, we know of cave-shelters, but so ransacked are they now that no information as to the mode of life of their former inmates is any longer obtainable. The Museum possesses numerous examples of the original Cape Point finds, including two human skulls. They consist of the usual middens types: mullers or braders, pounders, querns or mortars, !kwés and potsherds. Bone tools or ostrich egg-shell beads may, however, have been disregarded, and the accumulation of sea-shells not recorded. Close to the spot a conical earthenware pot was lately discovered. At a short distance from the hamlet of Hawston, on the coast of the Caledon Division, fortuitous circumstances revealed there lately evidence of occupation by man and beast of rock-shelters differing somewhat from those recorded hitherto. While removing slabs for building purposes from the roof of a moderate-sized cave having a broad opening, and well known to the inhabitants of the neighbourhood, mostly fisher-folk, there were exposed to view several long, tunnel-like fissures, which, on exami- nation, were found to be almost completely filled with bones of animals. These fissures, numbering six or seven, radiated from that part of the open chamber from which the slabs had been removed; the floor was covered with cave dust, sand mixed with a certain quantity of humus, and they allowed of the passage of a man for 144 Annals of the South African Museum. only a short distance. Therein were discovered three complete human skulls and two lower jaws, but no other human bones. Under the accumulation of animal remains that filled these fissures other human skulls, parts of skulls, and a small number of other human bones were discovered; also a few bits of calcined wood, and one small stone-scraper. There had been no percolation of water through the roof or sides, and we were told that the workman who discovered the entrance had found several skeletons, but that, hearing of our arrival to investigate the contents of the cave, he had reburied them no one knew where, and had disappeared. He cer- tainly was not to be found at the time of our visit, nor has he been heard of for the last three years. The person who brought me information regarding this cave had with him a number of bones and teeth, and among them the two implements figured in Pl. XXVII., Fig. 202, which are, or were, primarily the horn core of the Cape Eland (Tauwrotragus oryz). Several of the bones alluded to, including the horn cores, bore traces of chipping, of which traces it was very difficult to say whether they had been caused by cutting stone implements or by the gnawing of animals such as the hyzena, the teeth of which leave on the bones it has gnawed most characteristic traces. But a second examination of the cave settled that pomt. Marrow bones had been cut open both artificially and by crunching; a few of the former bore still traces of charring, whereas those crunched by either hyeenas or possibly also by lions showed on each side otf the fracture the marks of teeth. Moreover, many bones not containing marrow, such as those of the elephant, rhinoceros, and hippopotamus, had their ends plainly gnawed. Many of the smaller bones bore also the imprints of the teeth of small carnivora. The characteristic marks left by the hyzena, especially on thick long bones, so greatly resemble that produced by the slanting blow of a hatchet, that had no human remains been ultimately discovered in these fissures one might have been justified in considering these two horn cores as part of the animals devoured or gnawed by beasts of prey. But on close examination I found that the upper right part of Cut 1 of Fig. 202, from the top to the beginning of the curve, and a little before the middle, is not gnawed, but deeply bruised, evidently by pounding, and that the original contour of the opposite side bears no such marks. This core was thus used as a club, and it is indeed a very effective one. The other horn core exhibits very plain, slant- ing cuts at the end of its thicker part, but there is no corresponding traces of biting on the opposite side. The only conclusion we can The Stone Ages of South Africa. 145 come to is, therefore, that the slanting cuts have been produced by cutting stone implements, in helping to fashion this horn core into a club. Moreover, these clubs are, especially Cut 1 of Fig. 202, semi- fossilised, much heavier than those of recently killed animals, and their appearance is that of bone implements patinated, smoothed and hardened by long usage, whereas all the bone remains were either greatly bleached or had become very light. Many, especially those of the elephant and hippopotamus—these were mostly lying on or near the floor—were greatly decayed, some of the molars of the elephant especially. Incredible indeed was the quantity of animal remains. Most, if not the whole, of the South African fere nature were represented in these fissures. There lay together the bone remains of elephant, rhinoceros, hippopotamus ; those of the southern whale and seal, of the lion, leopard, and smaller cats; of the jackal, otter, and Viver- ridz ; of the porcupine and other large and small rodents. Among antelopes are the eland, koodoo, bontebok—and bushbuck and the hartebeeste ; the bush pig is also represented, and together with all the remains are also to be found the buffalo, the ox, and the fat- tailed sheep. The extreme abundance of animal remains seems to imply an alternate occupation of the shelter by man and beast. This accumu- lation of bone debris has a parallel in Europe. In a Mousterian cave of the Hte. Sadne, in France (Echinoz la Moline), there were found remains of no less than eight hundred cavern bears (Ursus speleus). In another ‘“ station ”’ relics of some six thousand horses were counted. In the first case it is doubtful if these terrible beasts had been laid low by primitive man; in the second they certainly had, In the dens of the hyzna not only whole skeletons of this animal are found, butits presence or occupation is revealed by its fossilised coproliths, as well as by the characteristic teeth-marks left on the bones it gnawed or cracked, characteristics already mentioned- Dens are also found in Europe where, doubtless, beast and hunter have replaced each other; the first occupier was the hyena, the second the bear, the third man. It becomes thus interesting to compare the result of the investigations of the Hawston cave and see if they fit or not with the authenticated conclusions arrived at in Europe. The human skulls are found near the entrance of the fissures which are themselves separated from the large, open-mouthed chamber ; there also are found a few human bones, under the accumulation of 10 146 Annals of the South African Museum. animals’ bones. The remains had been disturbed before a proper examination was made. When this took place, the humus of the crevices yielded very few human remains, but the skulls were either complete, or when incomplete (2), the parts had merely become separated; the sutures were intact. None of these remains has been interfered with by hyzenas. Therefore they must needs have been deposited there before the advent of the latter. They had by this time been so dessicated that they could prove no tempting morsel. The dismemberment of the human skeletons is probably caused by a ‘‘remaniement,”’ or disturbance of the cavern fissures, with their accumulation of animal remains. But these remains do not represent whole skeletons of animals. If they did, it would go to show that they belonged to animals that took refuge there to die. This absence corresponds to, or coincides with, what is known to have obtained in the caves or rock-shelters of Europe. There the hunter seems to have been especially fond of the head and marrow bones. He seldom brought to his shelter or cantonments the whole of his quarry, especially when it was a large one. He carried to it those parts only which he relished—the head and fleshy parts. Vertebre and ribs are seldom met with in these stations. In this Hawston cave ribs are rare, and vertebra scarce in pro- portion to the remains. Several of the skulls of animals of prey are fractured, and it is therefore probable that these beasts were slain by man, and that the skulking hyena or the daring jackal, taking advantage of man’s absence, entered the shelter and feasted on his leavings Two frontal parts—one that of a young ox, the other that of an antelope—are so clearly severed that it could have been done by man only, and this with a cleaving instrument. It is therefore probable that the cave was used first as a place of burial, in the manner described further on, in the Outeniqua caves. Game being more abundant, or edible shells less common, the accumulation of the debris did not form a shell mound. The dead would be buried or deposited in the fissures. The presence, how- ever, of domesticated animals can only be explained by the assump- tion of a reoccupation by man at a later date. At no very great distance from the Hawston cave another was found, consisting of one chamber without traces of fissures, but the only things met with there were two ostrich egg-shell bead-disks and a small scraper-knife. This cave, situated in proximity to the other, has a parallel in the The Stone Ages of South Africa, 147 discovery of the Coldstream cave, of which an account is given further on ; but there the case is reversed. At Coldstream, besides the one containing the sepultures, there was discovered, about a mile or so from it, another cave that ‘‘ contained no skeleton but a quantity of animal bones. It seemed a sort of shooting or hunting box.” This fact would add strength to the theory that the accumulation of the animal bones in the Hawston cave is due to the act of man. About one mile from the Hawston cave, and close to the lagoon, there was found a large boucher. It may, of course, be a mere coincidence, but it is well worth noting that the clean, slanting cut so apparent at the end of the horn core, in Cut 2 of Fig. 202, was certainly not obtained by means of a knife-scraper; it appears to have been produced by a heavy cleaving implement, such as a boucher. Ture Caves AND Rock-SHELTERS OF THE TZITZIKAMA AND OUTENIQUA. Let us now see if the cultural relics of the troglodytic, or cavern dwellings, Aborigines of the littoral differ materially from those of the open-air midden makers, or of the cave- or rock-shelter dwellers of the inland districts. The caves known to have been inhabited, or in which relics of primitive man have been found, extend from the coast of the Cape Division to that of Humansdorp. It does not follow, how- ever, that they are restricted to that part of the coast, but others have not been discovered hitherto either to the west or to the east of that area. They appear to be more numerous between Knysna and Humansdorp. In the words of Mr. H. D. R. Kingston * “the coast for many miles about Cape Seal (Rob Berg) is rocky and abrupt, open or sandy spaces of shore being few and far between. The land falls suddenly and often almost precipitously from the ‘flats,’ a tableland or shelf at the foot of the Outeniqua Mountains, with a face to the sea of some 300 or 400 feet in height. This is scored or intersected by a number of small rivers which have cut their way deep into the land, forming narrow and densely wooded kloofs. The caves, or rather fissures, are usually found midway between the top, where the rough vegetation of the veld above fringes over, and the trembling sea below.”’ It may be added that most of these recesses are very difficult of access ; and this point is of great importance because it shows that * “Notes on some Caves in the Tzitzikama or Outeniqua District, and the Objects found therein’’ (Journ. Anthrop. Inst., xxx., 1900). 148 Annals of the South African Museum. the inmates took shelter there as a protection from wild beasts, and possibly also from men. Thus Leith,* speaking of the caves at Mossel Bay: ‘It could only be reached by scrambling up a very steep talus of rubbish at the risk of slipping downwards into deep water, or from the top by a still more dangerous path.” Ibid.: ‘‘ 1 found the entrance to the cave about 15 feet above sea- level, in a cliff about 200 feet high. The cave is reached from the top by climbing down a very steep and rocky path, at the foot of which a 3Q-foot ladder takes one to the entrance. As this could not possibly have been the pathway used by its former occupants, I looked about for another, but in vain.” Kingston states: ‘To our cave we scramble and climb from the shore below and enter by the roughly rounded ;window-like opening to the west.” Ibid. : ‘Our better fortune was to find another cave at a greater height above the sea, but so difficult of access that it had evidently remained intact.” Writing of the Coldstream cave just discovered, Mr. J. 8. Henkel, of the Forest Department, who has very kindly set inquiries on foot in his Conservancy for the discovery of such caves and their contents, writes: ‘‘The cave was very difficult of approach. This cave, so called, an overhanging mass of rocks, is about 100 yards distant from the mouth of the river, and the floor of the cave about 50 or 60 feet above sea-level.’’ An examination showed that there were no signs of any human beings having, in recent times, entered or in any way disturbed the cave.” These caves or recesses are filled with debris of shells, even up to a short distance from the roof; occasionally bat guano helps to complete. Intermingled with the debris, which is often of con- siderable depth, are found, either on the surface or in layers, stone and bone implements, as well as bones of fish, mammals, birds, ornaments made of sea-shells, and skeletons so numerous, that the question arises, Were not the caves and shelters where skeletons abound, used as necropoles rather than as habitations for the living ? They may prove to have been both. In the open-air middens of the sea-coast I have recorded the find at Blaawberg of the skeleton of a woman under a layer of shell debris that had been a midden; but there is nothing to prove that the place, or that particular spot, was abandoned either temporarily * G. Leith, ‘‘On the Caves, Shell Mounds, and Stone Implements of South Africa” (Journ. Anthrop. Instit., xix., 1899). The Stone Ages of South Africa. 149 or for good, or that the dwellers had selected another more or less distant spot to resume their mode of life. The evidence of the shell mounds of the San Francisco Bay region might be comparable to ours. ‘‘These people buried their dead in the collection of shells and other debris in the neighbourhood of their dwellings.” * But there is no suggestion that they resumed their abode above their dead. In these caves or shelters of the Tzitzikama we, however, meet with evidence of a nature not hitherto recorded, and the parallel of which has also not been found hitherto, because in several cases the dwellers seem, in spite of some precautions against the possible return of the dead, to have either continued to live in the necropolis, or to have returned to it for the resumption of their mode of life.+ Most of these caves have been disturbed and their contents sifted and removed, because the farmers in these parts utilise the mixture of dark loam, decomposed shells, wood ashes, and especially bat guano which fill the caves, as fertiliser for their lands. With this debris, which has been known to be 20 feet thick, have been found coarse flake-knives, some of them of large size and made mostly of a glassy quartzite (Pl. XIII., Fig. 104). But before discussing their affinity with the South African relics, I shall quote here original accounts given me, which will, I am sure, prove of interest. The first of Mr. R. E. Dumbleton’s communications, relating his search for skeletons, was printed in the Cape Town Dvuocesain College and School Magazine, 1892, but he has since supplemented it on several points. “The cave is situated not far from the mouth of the Touw River in the George District, Cape Colony, and contains a very thick layer of guano. At various times people digging in this have dis- covered bones, shells, ashes, and other unmistakable traces of human habitation ; but it was not until about a year ago that any human remains were discovered. Then a farmer, anxious to turn the guano deposit to account, removed a quantity of it, and in so doing discovered a skeleton, apparently of a young man. However, either from curiosity, superstition, or some other cause, the skull was broken in and some of the bones were lost. For the purpose of * N.C. Nelson, ‘‘ Shell Mounds of the San Francisco Bay Region.” + Sparrman, who visited Tzitzikama in 1775, says of some fugitive Hottentots which he met on a farm in the ‘‘ Lange Kloof,’’ that they confessed having come from the Outeniqua by crossing the mountains ; that they had there a good master, but that they preferred returning to their country, especially because the death of one of them made it a law that they should change their abode. 150 Annals of the South African Museum. procuring one or more skeletons, I proceeded to this cave during the recent vacation. I commenced my search near the middle, and close to where the former discovery had been made. After digging for a short time a skull appeared at a depth of about 2 feet from the surface. I removed the layer of earth carefully, so as to discover the exact position of the skeleton. When this was done, I found that it was completely enveloped in a thick casing of dry sea-weed (Zostera maritima), which was still in a perfect state of preservation. Inside this again was the hair of a bushbuck skin, which had evidently been wrapped round the body. The skin itself had entirely rotted away, but the hair was still in good state of preservation. After removing all this very carefully, the position of the whole body was made quite clear. It was lying on the left side, facing towards the back of the cave, with the knees and hands doubled up to near the chin. Having ascertained this, I proceeded to take up the bones. On coming to the head I discovered imme- diately in front of the face two tortoise-shells, which, however, fell to pieces on being touched. With these there was the lumbar vertebra of a large ruminant, several flint scrapers, and also a peculiar instrument consisting of a piece of flint fixed in gum- cement, in which was inserted a piece of wood about 4 inches long, serving as handle. The latter, unfortunately, was perfectly rotten, and broke off short. On raising the skull I found that it was resting on a third tortoise-shell, which also fell to pieces. On examining the teeth, I found that they were all very much worn and several were decayed, while one or two were missing altogether.” * Encouraged by the result of his find, Mr. Dumbleton began excavations anew, with the following result :— ‘« We commenced operations round about where the previous two skeletons had been exhumed, and found nothing but layers of ash and shell with a few bones of animals here and there. Among these bones were some of large fish, birds, bucks of different sizes, and buffaloes. “Then we tried a slightly higher part at the back of the cave. Here we found a large quantity of dried grass spread out over a considerable space, which was evidently used as a bed. This was quite near the surface, and below were the usual layers of ash and shells. We next turned our attention to the mouth of the cave, where the soil is considerably deeper, and here we soon found some * This skeleton, which proved to be that of an adult male, was presented to the Museum, and is now mounted. The implement, now also ours, with the wooden handle, is figured in Pl. XIX., Fig. 150. The Stone Ages of South Africa. 151 dry sea-weed at about 3 feet from the surface. Knowing that there would be a skeleton below this, we removed the earth from above it very carefully, and then took off the covering of sea- weed, leaving the bones exposed. We at once saw that it was the skeleton of a small child having still a full set of milk teeth. It was placed in a crouching position, face downwards, and with arms and legs doubled up behind the body, probably with a view of putting it in as small a hole as possible. In taking up the bones I discovered a number of bits of ostrich egg-shell, each of which had two holes in it.* “These had evidently been strung together for a necklace and buried with the child, but there were no other ornaments to be found. The skeleton itself was in a somewhat dilapidated condition, the skull being in several pieces, and the lower jaw broken in half, so that it would appear to have been buried at a very much earlier period than either of the other two skeletons, which were in perfect condition when taken out. The position, at the mouth of the cave, and the depth at which it was buried also point to the same conclu- sion. There seems to have been no sort of ceremony about this funeral, such as placing a tortoise-shell for the head to rest on, nor did I find any implements about the head as was the case in the last discovery.” At my urgent request Mr. Dumbleton resumed his search, and he investigated five caves at Rob Berg, Knysna District, and found there several skeletons which he sent us with the following account of his search. (See sketch map.) ‘Nos. 1 and 2 are somewhat mutilated skulls which have been disturbed from their original resting-places but left lying in cave No. 1. They are only included for such measurements as may be obtained from them, and are probably of little further use. ‘No. 3 was also found in cave No. 1, very far back, and at about 3 feet from the surface lying just above the second of three distinct layers of ash and shell. All the bones found with it have been included with this specimen, but obviously many are still missing, and no trace of them could be found by sifting. A large water-worn stone—evidently used as a mill—(mortar) with a smooth groove on one side was found in close proximity to this skull. “Nos. 4 and 5.—These two skeletons of a full-grown person and child were both found in cave No. 3, which, luckily, had never been * IT have lately seen a disk of the same material with two holes; and in the district of Willowmore were found bits of Unio shells each also with two perfora- tions (cf. Pl. XXV., Fig. 187). 152 Annals of the South African Museum. disturbed by guano-seekers or others. This cave is very lofty but not deep, and the skeletons were practically on the surface, No. 4 being only covered by a few inches of cave dirt, and No. 5 by so little that a portion of the skull was actually protruding when found. No. 4 was lying about 3 yards from the back of the cave, and No. 5 almost touching the rock. The graves were evidently only holes scratched in the ashes just large enough to contain the bodies, which as usual lay all hunched wp on their sides. Both these skeletons were removed with great care. They have been packed up exactly as found, without any attempt at cleaning, as most appear to be partly decayed and are very brittle. BLT EINE AGS. INDIAN OCEAN 0 1 2 RES (Ce ee = al J SKETCH MAP SHOWING POSITION OF BUSHMAN CAVES AT ROB BERG, KNYSNA DISTRICT. “Very few implements of any sort were found with these speci- mens, but a few stones which may have been scrapers,* and a bit of pottery found on the sand-hills near cave No. 1 have been included in the box. “Four caves (No. 2 on sketch) close together, one being very deep, were also examined, but all of these had been ransacked by guano hunters, and nothing of any value was obtained from them. ‘All the caves contain immense quantities of shells and layers of ash. No paintings were found in them. “A spot, some 20 yards in front of cave No. 1, was pointed out * The usual uncouth scraper, of small size. The Stone Ages of South Africa. 153 where numerous human bones have been found and dug out, but this place is now under cultivation, and consequently not worth searching over.” Mr. H. A. Fourcade, of Storms River, not very far from Rob Berg, adds his quota of observations, which, however, differ in a material point, z.e., the posture of the skeleton, with that of other searchers. He wrote :— “As to your questions, I answer them seriatim. “1, The skeletons were wrapped in a cover of ‘sea-grass,’ probably Zostera maritima. I do not think that the skin of any animal was added as a wrap; traces of it would have been left. ‘«2. The one or two skeletons, the original position of which could be ascertained, were inhumed in a vertical position, as if they had been sitting, and with the knees tucked up. ‘“‘ 3. I have not noticed any implement the situation of which showed indubitably that they had been inhumed with the body. It seemed, however, that the grave had been covered with a flat stone, covered in turn with breccia debris.” The ‘sitting posture’ of at least one of the skeletons here mentioned does not thus concord with the horizontal one noticed by other observers. But Mr. Fourcade emphasises it, and it must be remembered that the so-called up-country Bush ‘people are said to have been buried in a sitting position.” The news of Mr. R. E. Dumbleton’s last successful search became known, and some residents of Knysna, prompted by the fabulous prices which rumour had proclaimed were obtainable for Bush skeletons, set to work to investigate all the caves and shelters that could be discovered. Several skeletons were found, but obtained in a manner that forbids any subsequent attempt at ascertaining the stratification of the caves. Mr. R. Atkinson, from whom we obtained two such relics, supplied me kindly, at my request, with the following information, which to a great extent corroborates that of Mr. Dumbleton, as well as that of Mr. J. S. Henkel. ; * See L. Péringuey, ‘‘ Rock Engravings of Animals and the Human Figure ”’ (Trans. 8. Afric. Phil. Soc., xviii., 1909, p. 4). ‘““Some years ago, in digging a grave, the workmen came across a human skeleton at the depth of 3 feet. Judging from the manner of burial, in a sitting position, we supposed it to be a Bushman grave” (Rey. Westphal). ‘**T was told by an old Bushman that it was not the custom to bury the body deep. It was generally put in a sitting position, the legs being doubled up, pressed against the body, and the head bent forward before rigor mortis set in” (J. G. Connan). 154 Annals of the South African Museum. ““ When in search of Bushman remains we explored about a dozen caves between the Knysna Heads and Plettenberg Bay, and though we found several skeletons, they were all more or less in a bad state of preservation, owing to the caves being more or less wet from the overhead drip. Eventually we got to Cape Seal (Rob Berg), when about a quarter of a mile east of the one in which R. E. Dumbleton found the skeletons he sent you, we found a fairly dry cave. In this we started work systematically to a depth of from 4 to 6 feet, and after fourteen days’ work we had found as many skeletons, con- sisting of two infants, one half-grown female child, one adult woman, and the rest adult males, all with the exception of the children, in a good state of preservation ; they were lying on their sides, some- times the right and sometimes the left, with knees drawn up towards the chin. The debris in which we found them consisted of bones of birds and animals, some sea-weed, ashes, and tons of every. kind of sea-shell from the adjacent ocean. The cave was about 40 feet from the sea and about 30 feet above it. Everything in it was more or less impregnated with salt, which, in my opinion, served to pre- serve the bones, &c. Though the cave had a good many ash-heaps in it, we found no signs of pottery or other utensils (excepting some bone and stone knives) nor had any of the shells been subjected to the action of fire; many still had sea-weed adhering to them. I consider that the great amount of shells and other debris to a depth of 6 feet in places must have taken very long to accumulate.” The description of a cave quite recently discovered is sent me by Mr. 8. J. Henkel, of the Forest Department * :— “The cave, so-called, is an overhanging mass of rock. It is about 100 yards distant from the mouth of the river, and the floor of the cave about 50 or 60 feet above sea-level. The cave is a fissure in Table Mountain sandstone which here had asteep dip. The sketches on page 155 will give an idea of the appearance of the cave. “In order to ascertain the nature of the deposits Mr. Witcher, sen., dug a trench about the middle of the cave, 1 foot wide, across the de- posit. When a depth of about 18 inches was reached the skull of a human being was found in a fair state of preservation. Mr. Witcher was much interested in this discovery, and decided to commence operations at the southern extremity and carefully remove a section. ‘“The material, consisting of shells, ash, &c., appeared useful as manure, and with a view to utilising it a road and footpath were constructed. The material, covering a section of about 9 square * A short account of the excavations carried out by us in this cave will appear in an Appendix. The Stone Ayes of South Africa. 155 yards, was carefully searched and removed, in most cases by him. The excavation was commenced in the southern part of the cave 24. PLAN OF FLOOR OF CAVE, WHEN DISCOVERED. Fia. ELEVATION. __pide sketch. When a depth of about 2 feet was reached human remains were discovered more or less decayed owing to dampness. The skeletons were complete, but the skulls were usually found 156 Annals of the South African Museum. to be broken. It was soon discovered that each skeleton had three or four flat stones placed directly over the remains. At about 2 feet 6 inches the skeleton of what appeared to be a woman was discovered ; next to it was the skeleton of a child approximately two to three years old. The skeleton of the child had two flat stones above it. It was lying on its right side. The left hand was lying across the neck. The legs were doubled up, the knees nearly touching the chin. The debris was carefully removed, and sur- rounding the wrists and ankles were pieces of perforated bone or ostrich-shell averaging 5%, of an inch in diameter. The ‘ beads’ were strung on a cord, which, however, crumbled away as soon as touched. One hundred and forty-two ‘beads’ were collected, but this does not represent the total number, as some were lost. The perforations appear to have been made in the same manner as the holes in the ‘stone hammers’ (! kwés)—that is to say, they were conical in shape, commenced from either side, meeting in the centre. The beads are in a good state of preservation. “At the same elevation and underneath a flat stone, in a cavity 8 or 9 inches deep, were discovered a number of shells averaging about 1 inch in diameter, each being perforated so as to admit of being threaded, apparently used as a wrist or waist ornament (Pl. XXVII., Fig. 208). No cord was found. Ninety-one shells in good preservation were found—about a dozen or fifteen had perished.* “Excavation was continued to a depth of about 9 feet, and up to the time of my visit in July, 1909, seventeen skeletons were dis- covered. In every case flat stones were found to cover the skeleton. These stones appear to have been placed directly upon the body, as shown by the evidence of the bones in some cases adhering to the stones, and other indications produced by decomposition of the flesh. In every case the skeletons were doubled up in a manner some- what similar to that of the child already described. All the skeletons discovered by Mr. Witcher were lying on the right side. His son, how- ever, asserts that one skeleton he removed was lying on the left side. ‘In the course of excavation three stones were found on which drawings were executed in black. “On No. 1 are four drawings of human beings with prominent calves and buttocks. The sizes of the figures are respectively 44, 44, 44, and 14 inches (Pl. XXVII., Fig. 199). ‘‘On No. 2 is to be seen the figure of a human being (Pl. XXVIL., Fig. 201). * They are small examples of Cassis achatina, Lam. The Stone Ages of South Africa. 157 ‘No. 3 contained four drawings of animals; the largest animal appears to represent a wildebeeste, two smaller animals drawn facing one another appear to be steinboks, and the fourth a bluebuck. ‘‘Unfortunately the drawings on stone No. 3 have disappeared, as it was left exposed to the weather. “On the walls of the cave, which are discoloured by smoke, no drawings have hitherto been discovered. In the course of excava- tion different coloured bits of clay were found. The drawings on the flat stones are made with charcoal. ‘«« As already intimated, the debris in the cave was systematically removed in layers, bagged and used as manure, on lands in the neighbourhood. The straight perpendicular wall of the northern face of the section shows clearly the method in which the debris accumulated. “The latter consists principally of various species of shells, ash, and fragments of bone, both of animals and fish. Interspersed here and there are chips of stone, water-worn stones, and household tools consisting of stone plates,* stone hammers, scrapers, bone needles, &e. “This debris appears to have accumulated entirely by human agency, that is to say, it is the waste products of articles of food, such as shells and bones of animals and fish used, the ash from fires, &c. The fires appear generally to have been made some distance from the walls of the cave—say 10 to 15 feet—because little or no ash is found near the walls or the looser texture of the debris. It is interesting to note that no attempt appears to have been made by the inhabitants of the cave to remove any waste material. There are, however, indications—though further inves- tigation is required—that fires were made above the graves. There are no indications that the graves were of any depth; on the con- trary, from the number of skeletons found at varying depths it would appear that only shallow excavations were made sufficient to admit a corpse and to permit of the flat stones being placed a few inches below or flush with the then floor of the cave. ‘Tn order to verify the statement made by Mr. Witcher, I carefully examined the section of the cave exposed by the excavations up to the 14th of July, 1909—the date of my visit. I found an excava- tion about 9 yards square and a depth of 9 feet. In my presence a further layer was removed and a skeleton discovered. It had above it three flat stones. It was lying on its right side with knees drawn up to the chin. The arms were also bent. The skull was * We shall see later on that these are painting palettes. 158 Annals of the South African Musewm. broken, probably owing to the weight of material above it. The skeleton appeared to be that of a full-grown man and much decayed. ‘In another part of the cave, about 2 feet below the present surface, the skeletons of two children were found. In these two cases only one flat stone was placed above the body. In one case, part of the skeleton was found adhering to the stone, showing clearly that the stone was placed directly above the body. The stones removed from the graves were carefully examined, but no indications of paintings on them were found. “The cave still contains a great deal of debris, and it is estimated that at least fifty skeletons may still be found. “Up to the present over twenty skeletons have been discovered in this cave.” * A communication of this importance requires, however, careful perusal, and I had less hesitation in asking Mr. Henkel for more details because I enjoy his friendship, and I have good reasons also for appreciating his power of observation. This is his answer to my further queries: ‘‘ I notice in your note of the . . . that you lay emphasis on the position of the skeleton. All I can say is, the three I took out were distinctly lying on their sides. W. says that all those he had taken out, about seventeen, were lying on the right side. From the position of the stones it is quite evident that the bodies were not buried in an upright position. You will observe from my account that there was no direct evidence to connect the stones decorated with drawings with skeletons, although in the first instance I was led to believe that this was so.t Further with regard to your letter of . . ., the skeletons I removed showed no evidence of a bushbuck skin. Fine remains of organic matter were discovered, but nothing to show that it was other than the human remains. No tortoise-shell, no implement as sketched by you.| Numerous round perforated stones (! kwés) some showing many chips, and seemingly used for a long time, were also discovered.” The evidence obtained from this Coldstream cave throws a good deal of light on the mode of life of the people that inhabited it. New facts of very great importance are revealed; others corroborate discoveries already made in similar localities, and it is interesting to * The relics mentioned in that account of the discovery of the Coldstream cave are now in the Museum Collection. + Subsequent excavations carried out by us have proved that these stones with paintings were deposited on the skeleton. + That of Pl. XIX., Fig. 150. The Stone Ages of South Africa. 159 compare these results carefully. Let us therefore examine in detail the domestic utensils found in these Outeniqua-Tzitzikama caves. Of stone implements we have large scraper-knives, several are figured (Pl. XIII., Fig. 104); some have worn edges, in others that part is sharp. None of these, however, seems to have been buried with the corpse, whereas the ornaments of the defunct in several cases were.* The two exceptions known are the hafted tool (Fig 150, Pl. XIX.) from the Touw River cave discovered by Mr. Dumbleton, and another from Rob Berg, acquired lately by the Graham’s Town Museum. The knife-scrapers are of the same material and shape as those found at Cape St. Blaize, and may be of any age, but we possess a lanciform one which is as much worn by exposure as any I have selected in order to demonstrate the antiquity of these artefacts. The other stone implements are pounders and grinders of the usual shape (Figs. 168 and 170, Pl. XXIII.), some are somewhat cylindrical but they concord more or less with those found in the open-air middens. Then, we have the !kwés, but although it may safely be said that two of these perforated, or partly perforated, stones are never alike, we notice in those occurring in the Tzitzikama caves a great difference. None is as large or as thick as those found up-country or outside the shelters. There is a tendency to flatness, and even to reduce the natural flatness of the stone selected; Cut 2 of Fig. 156, Pl. XX. is only 1 em. thick, and No. 3, 2 cm. in thickness. I know of examples one face of which has plainly been ground to reduce the thickness, and the convexity of the other side is very little pronounced. The median orifice of those flat disks is very small, cf. Cut 2 of Fig. 156, Pl. XX. This disk is so thin that it might be taken for a spindle-weight, whereas it is doubtless an ornament. One side of Cut 3 of Fig. 156 shows plain marks of it having been used as a grinding tool. Even the larger ! kwés, perforated or not, are somewhat depressed (Cut 1 of Fig. 155), and in this they correspond to a good many examples found in the Cape Peninsula and also in the inland districts (cf. Nos. 1, 3, Od; Of Big. 153, Pl. XX.). These flat disks with a very small perforation are not, however, absolutely restricted to these caves. Cut 1 of Fig. 156 was found in the Stellenbosch District while ploughing a field. But it is the only implement of this kind found outside the Tzitzikama area. * See also Appendix. 160 Annals of the South African Museum. In this Coldstream cave were also found plates of slate not more than 5 mm. in thickness (see text-fig. 25, and also Pl. XXVILI., Fig. 196). Their use was explained quite lately only by the examination of a similarly shaped laminate stone, resembling the left specimen of Fic. 25. x 4. Fig. 196, and covered still with a red pigment—ochre mixed with a fatty substance, which shone more and was thicker on the edges. These implements are therefore proved to have been used as palettes, not palettes from which the Bush artists picked their paint, but an implement by means of which they besmeared their body or parts of the body. The Stone Ages of South Africa. 161 This type of palette is not, however, restricted to the caves of the littoral. The specimen that has still retained the pigment on one side was discovered in the Oliphants River valley, Cape Colony, in a rock-shelter with rude paintings. The example on the right of Fig. 196 is from the Beaufort West District of Cape Colony. So far I know of these six implements only, but others may have been overlooked. An interesting point is that these besmearing palettes are identical, some in shape, but all in thickness, to those discovered in the dolmens (Menhirs, Cromlechs) of Aveyron, in Southern France, and also in prehistoric Egyptian tombs.” Had not the preservation of this red pigment indicated plainly the use of these implements there would have been justification in con- sidering them as braying tools, for they fit the hand wonderfully well for this purpose, and there is also no reason for their not having served as such. Our troglodytes brayed skins; of this there is little doubt, and, if there was any as to their wearing them as clothes or protection against the rigour of the cold season, this doubt will be easily set aside by a glance at the bone bodkins occurring in these caves. Cut 5 of Fig. 172, Pl. XXIII., was found in the rock-shelter where the doubly perforated ornaments (Fig. 187 of Pl. XXV.), made of nacreous shells, were discovered; Cut 4 of the same Fig. (the lower one) from one of the Knysna caves, has counterparts in Cut 4 of Fig. 172, Pl. XXIII. Fig. 194, Pl.. XXVI. represents wing-bones of sea-birds turned into awls or needles. The great interest attaching to two of the examples here figured is their ornamentation, which in Cut 2 of the Fig. consists of fine vertical, parallel lines, and of others crossing each other at such an angle as to produce lozenge-shaped intervals. In the middle of Cut 1 of the same Fig. we have a chevron pattern, somewhat rough, but it is the first and probably the sole example of its kind known. + But there are in the Tzitzikama-Outeniqua caves bone implements other than awls or bodkins. Cut 2 of Fig 193 appears to be the bone tip of an arrow; the bone-point (Cut 6 of Fig. 193), might certainly have served as an arrow-head. Then, in addition to these, we have bone knives. Fig. 1 of Pl. XXIII. is a very * Cartailhac, ‘‘ Les palettes des dolmens Aveyronnais et des tombes Egyptiennes ”’ (Bull. Soc. Arch. Midi., 1906). + Among the Bantu races, especially the Bechuanas, this chevron pattern is in general use. 11 162 Amnals of the South African Musewm. effective instrument for detaching shell-fish. Judging from the length and slant of the cutting part its maker and presumably owner, was left-handed. Cut 2 of Fig. 193 is an ivory knife that has seen much service. It is a stout one, bearing on both faces the unmistakable traces of the small stone implements which helped to fashion it into its present shape. Cut 1 of the same Fig. is a strong borer, also made of bone. All these bone tools bear that unmistakable patina that long use at man’s hand imparts. The ornaments that have been found in the Tzitzikama caves, seem to consist mostly of shells perforated so as to be strung, or worn singly. Thus in Fig. 203, Pl. XXVII., the perforation of the shell (Conus rosaceus) is transverse, and was caused, therefore, by a cutting implement; Fig. 204, a limpet (Patella cochlear), is, on the other hand, perforated by a piercer, and so are ali the shells of Cassis achatina (Fig. 205), which formed either a necklace or a girdle. These examples were found on the skeletons or adhering to them, and anklets of ostrich egg-shell beads, recalling the form and size of those encircling the lower part of the trunk of the semi- fossilised skeleton of the Blaauwberg midden, were discovered on the ankles and arms of one of the inhumed children in the Coldstream cave and the Touw’s River caves. This would show that although it may not have been a general practice the ornaments were buried with the body. And now the question arises, Do these traces of culture, simple though they are, denote a more advanced stage in the troglodytic Strand Looper, for Strand Loopers they indubitably were—than in those that lived in the open-air middens, or in the up-country caves or rock-shelters ? The answer is in the negative. In one of my papers on Rock-engravings of animals, the human figure, &c.,* I have by inadvertence allowed to be printed the statement that it was very doubtful if the Strand Looper belonged to the artistic race with which the Bushman is connected, because so far no relic of art in any form had been found in their shelters or sepultures. I had the less justification in allowing this state- ment, which was really intended to disprove the connection between rock-gravings and Bushmen as the aborigines that produced them, to go to press, that for many years we had in our Collection a fragment of rock found in one of the Knysna caves, by Mr. Cheva- lier, somewhere in the early seventies. This picture (Fig. 200, Pl. XX VIL.), is not of parietal, or wall nature; it was executed on a fallen rock fragment. In view of what we * Trans. S. Afric. Phil. Soc., xviii., 1909, p. 418. The Stone Ages of South Africa. 163 know now regarding the position in which skeletons lay, it is per- haps permissible to consider the scene as not men dancing, but men inhumed or going to be inhumed. For a long time this example of the troglodytic art remained unique, and it was necessary to wait for new discoveries before asserting that the dwellers in caves and those living in the open had the same artistic temperament and the same ability to express it. The fact may be said to be clearly established now. In the Coldstream cave were found three charcoal paintings. The first Fig. 200, figured in Pl. XXVII., may have been parietal, yet it is difficult to accept the theory that the fragment fell from the rock as it now stands. But the doubt is not permissible as far as Fig. 199 is concerned. This is not a parietal painting. The scene is executed on what was once a quern or mortar, with only a moderately deep artificial depression. I know of no other such occurrence here, but Mr. J. M. Bain informs me that near the mouth of the Orange River he found a scene painted on the shoulder-blade of what must have been a very large seal. Both these representations were found in the stratification of the Coldstream cave; there is evidence, now, that they were laid on the skeletons. They are fading so rapidly that we had a great deal of difficulty in reproducing them, and the ‘“ touching’”’ of the negatives however skilfully made by Mr. A. R. Walker, one of our Museum assistants, may prove to be very slightly inaccurate. Mr. Henkel reports that the third one became obliterated as soon as exposed to light. The representation on this quern (Fig. 199) is that of a dance executed by women. The steatopygia, the extremely swollen calf of the legs,* and the hanging mamme go to prove it. The head of the child and his position are somewhat realistic. There is another feature which is of extreme importance; it is the half- moon-like circle surmounting the heads of the dancers.t In a large -hunting scene from the Cedar Mountains (Cape Colony) a similar half-moon, or rather crescent, surmounts the head of the hunters who, armed with a club, go to attack a herd of elephants. The same head-dress is also to be found on rock-paintings in the Piquetberg District of the Cape Colony. Could this be a “totem ”’ ? But that it is a distinctive badge, there can, I think, be no doubt. * J have good reason to believe that this extraordinarily large representation of the legs is due to the leather or grass rings worn formerly by these natives, especially the females. + I must, however, explain that on the partly obliterated original these crescents are not quite so plain as shown in the illustration, but after a very careful examina- tion I am satisfied that they were meant to be as now represented. 164 Annals of the South African Museum. And on the whole these painted scenes, and also the “ mobilier,’’ or domestic implements, go to show that these cave-dwellers were members of a race that was not necessarily troglodytic and sedentary ; in all likelihood they also sallied forth as their wants or desires prompted. It might be urged that, safe from foes in their dwellings, they led there a quiet, perhaps simple, certainly monotonous life; that there they were born and thereafter buried. We have, however, in addition to the one adduced, other evidence that such was not the case, and that although there may have been among the shelter- dwellers some “stay at homes,’’ others went abroad. One of my reasons for making this assertion is the find in the Coldstream cave of two crocodile teeth, pertaining to an animal of no mean S1Ze. Although the rising of that part of the coast is not only far from improbable but most likely, yet it would be rash indeed to assert that the climate had altered so much since man’s occupation of the cave that saurians, known to require tropical conditions, could have existed in the Coldstream River. It is simpler to conclude that these teeth were brought to the cave from distant parts. We have also pictorial evidence of a similar character. Fig. 198, Pl. XXVII., is the reproduction of a ‘‘ Bushman ”’ paint- ing, from the “ Lange Kloof,” not very far from Knysna and Humans- dorp. The interest attaching to this painted scene is not the human head on the left, executed posteriorly probably by a shepherd, nor the delineation of the seals, also on the left, but of that of what is unquestionably a giraffe.“ Like the crocodile, the giraffe requires tropical conditions and also a very dry climate—conditions not obtainable in the Tzitzikama-Outeniqua Districts. We are thus faced with two alternatives: either the climatic con- ditions have changed since the occupation of the caves by these primitive aboriginals, a proposition which postulates an extreme antiquity, or—and this is much more likely—these troglodytes were * Difficult as it now is to endeavour to explain the meaning of some of these Bush paintings, several figures of this scene speak for themselves. In the centre is a hive-hunter smoking bees. The seals are typical—some painted red, others white. At the bottom is a dog chasing a jackal. The use of dogs by Bush-people must of necessity, have been a late innovation, and therefore the age of the paint- ing must be deduced from this fact. The hand, painted red, is plainly that of a human being, bedaubed with paint, and clumsily pressed on the rock. Had it not been done so clumsily it might have afforded valuable information. Abbé Breuil has recorded several impressions of the same kind in the now famous Altamira cayes, and it is of interest to mention that in some cases, one of the fingers was mutilated. This mutilation was by no means uncommon among the Hottentot- Bush people. The Stone Ages of South Africa. 165 great rovers, and had seen in their wanderings towards the North both crocodiles and giraffes. They, or at least one of them, retained such an impression of the latter as to be able to retrace one, awkwardly it is true, but none the less realistically, on his return to the South. Tue Intanp Districts Rock-SHELTERS OR CAVES. If, leaving the littoral, we proceed to investigate the rock- shelters of the inland districts, we find there, also, in addition to the characteristic rock-paintings, relics of a stone and bone industry. ; As long as the cave-dwellers’ centres are not found far from the littoral, there is some similarity in the domestic implements—not in the implements themselves, but in their execution. Thus, at Montagu, Cape Colony, in a cave-shelter filled with bat guano, was found under several feet of this material the wonderful club repre- sented in Pl. XIX., Fig. 153; with it was discovered the stone bead, (No.3 of Fig. 186, Pl. XXV.). At the entrance of the rock-shelter were a few painted figures. It is interesting to note that this club is really a phallus, bearing plainly the intaglios of a stone scraper- knife. Made of one of our hardest indigenous woods (Olea spec), it has been thinned at the end so as to allow of a small hand, like the one we know to prevail among our present Bush races, to grasp it firmly. The shape and composition of the bead will be treated in the chapter dealing with ornaments. Human bones are said to have been found in the Montagu cave, but I have seen no proof of their occurrence. That the food of the aborigines of the littoral was mainly shell- fish must be readily admitted. But the inland districts aboriginal did certainly appreciate fresh-water molluscs, and their remains are often found in rock-shelters. The contents of one sent us lately by Dr. C. E. Piers, occurring in the Beaufort West District of the Cape Colony, include nacreous pieces of the large fresh-water bivalve Mutela wahlbergi, as well as pieces of the sea-shells Patella and Halioits. It is thus probable that these rock-shelter dwellers resorted to the sea, for a change of diet, prompted, maybe, by a craving for salt or salted food. But so far as our knowledge of the relics left by the up-country cave-dweller goes, it may safely be stated that the more removed from the littoral, the more primitive is the culture, even when his talent at portraying is superior. The stone implements found in 166 Annals of the South African Musewm. the caves or rock-shelters of West- or East-Griqualand, Rhodesia, &c., are very primitive indeed, whether these caves contain paint- ings or not (cf. Figs. 133, 134, 1385, 136, Pl. XVII.). In Griqua- land-West we have—but not well authenticated—a few pounders and !kwés, in Rhodesia none has been found. They may be dis- covered ultimately, but they are not known, so far, to occur there. In the Orange River Colony, however, we find a connection with the culture of the littoral. The bone bodkins (Fig. 195, Pl. XXVI.) are not unlike those from the Tzitzikama-Outeniqua caves, but they are not so well finished ; the flattened appearance of the two examples on the right of the Fig., which show a great deal of usage, point, however, to another pursuit—that of mat-making. The ornaments are also of another style (Figs. 206, 207, Pl. XX VIT.), &e. There is a distinction, perhaps more apparent than real ; but nevertheless, on the whole, the culture shows retrogression. The following is the inventory of the domestic appliances of a rock-shelter cave found between Wepener and Dewetsdorp, in the Orange River Colony. They were presented to the Museum by Mr. John Wood, and are illustrated in Pl. XXV., Figs. 185, 188, 190, Eto? Two bone points (Nos. 6 and 7 in Fig. 185), are made of ostrich bone: fig. 7 is blunt at tip, but had plainly been sharp once ; fig. 8 is pointed, and the point corresponds with the groove in the small fragment of stone (No. 2); there is another fragment (No. 13) bearing two grooves, one of which was intended for producing a sharp point, judging from its smaller diameter. Nos. 8 and 10 are tiny scraper-knives, and 9 is an agate core from which No. 10 was produced. Nos. 4 and 12 represent horn cores, the tip of bush-buck horns; in addition there is a partly charred stick of the thick- ness of a pencil. No. 5 is a very small muller or brader, and No. 1 a naturally rounded pebble that shows no trace of usage. No. 11 is undoubtedly the most effective implement among these domestic appliances. It is a piece of flat stone of which the crescent-shaped upper part has been artificially ground into a sharp cutting edge. Whether due to exposure or to age, the horn cores and also the bone awls show signs of extreme decay; the awls having become very brittle and the horn cores very much split and grooved. Pottery of four kinds was also found in this rock-shelter. Fig. 190 represents two fragments with decorative patterns. In No. 1 the incisions could have been produced by means of the The Stone Ages of South Africa. 167 bone awls (Nos. 6 and 7 of Fig. 185), or the scraper (No. 8 of the same Fig.) and the elliptical impressions on No. 2 also by the same tool. These fragments of pottery are not sufficiently large to enable us to restore the original contour and to decide whether they were conical or not. In the same shelter were found undecorated pieces of enormously thick earthenware (No. 2 of Fig. 192), as well as some of a thinner kind (No. 1 of the same Fig.), and also a large fragment (Fig. 191), which is the flat bottom of an evidently massive earthen pot. The connection of that kind of pottery with the mobilier here described is, however, doubtful. Of the bill of fare of the inhabitants of this shelter we have also traces. There were found several valves of the fresh-water shell Unio caffer (Fig. 188, Cut 2); part of the lower jaw of the poreupine (Cut 3); bits of ostrich egg-shells (Cuts 5 and 6), that probably might have been eventually turned into perforated disks. Cut 4 is a piece of the leg-bone of an antelope blackened with age, Cut 1 is that of an ostrich and deserves special mention. In the inner part is an artificial groove made in such a way that either part when eventually detached, would be already adapted partly for the grooved sharpeners (Fig. 171 of Pl. XXIII), by means of which it would become ultimately the bone head tipping the arrows (Fig. 142, Cut 1). Small and few are the household goods of the occupier of this shelter: his industry is indeed primitive, for there is no reason to believe that he has removed the greater part of his mobilier. He is a nomad; in his pursuit of game, perhaps also to avoid being pursued, he must not be hampered in his movements which perforce must be rapid. Hence the paucity of utensils, hence also their ‘reduced size. Discarding the pottery he will take with him his tiny tools—certainly a not overwhelming burden in his search of pastures new. I have selected for illustration the contents of this up-country rock-shelter from among many known to me, or the contents of which are stored in the Museum, but the diminutive size of most of the implements found in the caves or rock-shelters other than those of the littoral is apparent throughout. Whether occur- ring in place with paintings, from which we would conjecture that these shelters had been occupied in somewhat more than a spasmodic manner by a clan or a number of people, or in places from which the number of these utensils implies occupation by either a single, or at most a few families ; or, again, whether these 168 Annals of the South African Musewm. shelters show or not remains or traces of mural decorations in the shape of paintings, small implements only are met with. Fig. 133 represents eleven scraper-knives from Matoppos, Southern Rhodesia, where parietal paintings occur; Fig. 134 of eleven scraper-knives from Modderpoort, Orange Free State, where paintings are also found ; Fig. 135 implements from a rock-shelter which, like the one I have dealt with, is also from Smithfield, O.F.S., and without any painting; and Fig. 136 similar pieces from the neighbourhood of a large cave full of frescoes in the Prieska District, Cape Colony. Fig. 137 in Pl. XVIII. is very instructive. It represents a gun and a spade—terribly symbolic of the fate of the race—cut or fashioned of wood, and discovered in a small rock-shelter in the District of Victoria West, Cape Colony, with traces of fire, &c. Together were found the three small knife-scraper tools used for the fabrica- tion of these wooden representations (Nos. 3, 4 and 5). That they are of very recent age is obvious; but we have no evidence of these pygmy tools having been hafted or held in a handle by gum-cement. Yet had they in this instance been so treated traces of that mode of attachment would have certainly survived the comparatively few years that have elapsed since they were manufactured. The conclusion is that the man who used these shelters trusted to and relied on his bow and poisoned arrows; the rest is of no great consequence to him. Any stone will do to cut his quarry; has he not his firesticks with him? are pots so difficult to make, roots so difficult to grind? But his arrow has to be tipped with flakes with a sharp edge—stones of a kind, however, which are not found everywhere; hence, the small agate nuclei he carries with him, the small grinder: with which to prepare the poisoned paste he uses for his arrows. He may have known the bouchers, and perhaps the men that used them. What matters to him? he has now the dreaded weapon before which every other aboriginal flees. Boucher-maker probably he never was, and had he been it, what would he do now with these ponderous weapon? Is not the small chip of agate sufficient to bring his quarry down ? Is not any stone he may pick sufficient to break the marrow bones without fashion- ing it at great trouble into an amygdaloidal shape and carrying it wherever he roams! Is this man the most primitive of mankind? Are his stone implements, then, the true eoliths, the precursors of the Chellean- Mousterian bouchers or other tools of those periods ? No, plainly not. He is an example of regression as far as the lithic culture or industry is concerned, but this regression in The Stone Ages of South Africa. 169 one direction implies a singular case of advance in another—an advance which permitted of his survival. He, no longer a fabricator of large tools, contrived to outlive those who were, and for aught we know—and it is not such an improbable hypothesis—he took their place, having mastered them through the efficacy of that tiny weapon which he invented, namely, the poisoned arrow. Craft versus brute force. ( 170 ) CHAPTER Sovak: THE SEPULTURES. If we compare the relics, as now revealed, of the open-air middens, caves, or rock-shelters of the littoral, as well as the few yielded by the inland caves or rock-shelters, one thing stands out, namely, that the culture, as indicated by the stone- coupled with the bone- industry, has many points of affinity, if it is not identical. But the style of burial which we know now prevailed, or was followed in the George, Outeniqua, zitzikama caves, differs so much from what is known to obtain elsewhere in South Africa, that it was quite justifiable to expect that there should be found with the skeletons, and in addition to scraper-knives, some Paleolithic implements, owing to this style of sepulture being probably more ancient. But such is not the case, and it is the more surprising that these large paleoliths are almost beyond counting in South Africa, and they are also far from uncommon on the tableland at the foot of the Outeniqua, called the ‘ flats,’’ not far from the cave recesses. Sparrman, the most credible of the early travellers in South Africa, visited in 1775 the very parts of the Cape Colony where the burial caves or rock-shelters occur, and he tells us most emphatically that the Hottentots bury their dead naked or wrapped in their ‘‘kaross’’* in the ground, or thrust them into crevices of rock. He, unfortunately, does not mention the attitude imparted to the body after death. Kolben, who preceded him (1705), goes into more details, and although his veracity has been often impugned, I have found that his statements were far from being quite devoid of truth. But he certainly exaggerated the facts. + * A covering or mantle made of sheep or any other skin. + In the posthumous account of his visit to the Cape by Abbé de la Caille (1751), whose ‘‘ severe probity was much shocked at the want of veracity of Kolben,” it is admitted that the latter’s statements were really culled from, or supplied by, Mynheer Grevenbroek, Secretary of the Council of Policy, who had recorded in writing what the Hottentots whom he had met had told him in answer to his questions. The Stone Ages of South Africa. L71 Speaking of a Hottentot who breathed his last, he states that his friends “ bend his body, in such a manner that his head is brought between his legs, so that it assumes the foetal form.” All the accounts of the excavations agree on one point, and that is that the skeletons were lying in the fetal posture, but Mr. Henkel has made it quite clear that in the Coldstream cave flat stones were placed over the body. These stones are some 18 inches long and wide, and from 3 to 4 inches thick. At Storms River, Mr. Foureade, in search of human remains, says that it seemed as if the grave had been covered with a flat stone. Mr. Dumbleton and Mr. Atkinson have not met with the stones, but, not being aware of the fact, it is quite possible that they did not notice them. We have thus proof of a burial rite, which, as far as we now know, is peculiar to the littoral troglodytes, but which does not appear to have been followed by all. At Touw River, Rob Berg, and Storms River, the skeletons are wrapped in sea-weed ; one is, in addition, shrouded in a skin. With it were discovered remains such as tortoise-shells, that crumble to shreds as soon as handled or exposed to the air, and a hafted tool, the scarcity of which in these sepultures is not easily accounted for except on the ground that, in other cases, the wooden hafting decayed without leaving any trace. In the Tzitzikama cave sea-weed does not seem to have been used in the burial, but we have there placed over the body stones, some of which still retain traces of fire on the upper side. Is it permissible to draw from these discrepancies the conclusion that the people that buried their dead in these caves had different practices, or that some were of greater antiquity than others? That they belonged to the same race is proved by the study of their skulls ; but there is so far nothing to show that some of the burials are more ancient than the others, nor do the skulls differ from those of the surviving Hottentots in the manner of those of the Neanderthal and of the Cro-Magnon. In most cases the caves contained several skeletons; at Cold- stream, in the superficial debris received from there, I find remains of nine adults; to all appearance the cave will yield a great many more; twenty are said to have been exhumed. Some, if not most, of these remains from the caves undoubtedly show signs of great age. The vertebra and ribs especially are as light as if they were made of papier maché. Many of the bones are hygrometric ; most of the relics, bone or shell, are strongly impregnated with salt—a fact which, as Mr. R. Atkinson suggests, may account for their 172 Annals of the South African Museum. state of preservation—and all the human remains are extremely brittle. But in spite of that there is nothing to give us a possible clue to their relative age. These caves and rock-shelters may, and probably have been, resorts of aboriginals until very recently. In a cave-shelter where the bodkin (Cut 4 of Fig. 172, Pl. XXIII.) and the double-pierced nacreous beads (Fig. 187, Pl. XVI.), were dis- covered together in the Willowmore District, there was also found the small rusty iron blade of what may have been a knife or a spear-head. In the Coldstream cave was found, 2 feet below the surface, a portion of what appears to be the steel striker of a tinder- box, and also a piece of copper, probably a portion of a wreck. These finds are on a par with the piece of Oriental china and brass button at Bloembosch, of the glass beads or clay pipes and brass buttons of European manufacture in the middens of the Cape Flats, &e. Buried in the sands of Blaauwberg was found the ornamental brass side of a military bridle, together with the skeleton of a Strand Looper. The pattern of this bridle is said to be Dutch. At Hagel Kraal, in the Bredasdorp District, Cape Colony, two tiny brass bangles, 440 mm. in diameter, one round, the other flat, were discovered in the sand-dunes, «e. These finds point plainly to a continuance of occupation of sites that lasted until, and probably long after, the advent of the colonists. The lamentable account given by Sparrman, which I quote in the chapter on ! kwés, makes it also probable that the wretched Hottentots whom he mentions had taken shelter from their enemies in places difficult of access, as these caves prove to be. These fugitives were provided with !kwés or perforated stones, similar to some found in this cave debris, and also in the neighbourhood of the shelters. But interesting as these finds prove, they throw no light on the relative age of the men who were cave-dwellers, who manufactured the stone and the bone implements, and who executed the paintings. Their culture is strikingly like the Magdalenian ; but a similarity of culture does not imply a similarity of races nor a similarity of date. In spite of this proposition, the mode of sepulture of these aborigines, the cave-dwellers especially, have so many points of re- semblance with what we know occurs elsewhere that it is permissible to refer to these occurrences for purposes of comparison. In South Africa, whether buried in a sitting position or lying on one side, the position is that of the ‘child in the womb.” The Stone Ages of South Africa. 173 This rite has been observed in many parts of the world, and three causes are assigned to it: symbolic attitude (the dead returning to the bosom of mother earth); greater facility for burying the body in a restricted space, especially when a rough crypt had to be made; and lastly, when ligaments were used—which is not the case in South African sepultures—due to a sense of terror, inspired by the dead, in the survivors. But one reason which has not, to my knowledge, been adduced to explain this position may be found in that assumed naturally by the Hottentot-Bush aborigines of the present day when sleeping, and especially when ill, and that posture is identical with that of the skeletons. This fact had not escaped the notice of Sparrman. Endeavouring (1775) to induce a young Hottentot to enter his service, he crawled into his hut, and “found him lying under his skin-cloak in the manner of his countrymen which I have already mentioned, the knees drawn to his nose, as a fetus in the mother’s womb.” It should not necessarily be inferred from the above remarks that this method of sepulture originated with the Hottentot races. Yet if was in common use among the neolithic people of Egypt with whom the Hottentots or ‘‘ Sans ” have undoubtedly come in contact, if they are not of them. It is also observed among some pre- historic European races,* even when the small stone coffins, or ‘‘ stone cists,’ themselves much smaller than the body and making therefore a reduction of length compulsory, had come into vogue. But whereas the connection of the Hottentot and prehistoric Egyptian is probable, the proposition that the same existed with the Neolithic races that inhabited Central or Southern Europe might seem hazardous, except that the discovery in the “‘ Grimaldi Caves” near Mentone, of a small race with negroid characters, associated with australoid, throws a new light on the possible dispersion of a race which inhabited at all events the shores of the Mediterranean, and possibly also Central Europe—a race that preceded the Cro-Magnon, and that was in turn posterior to the Neanderthal, which, in all likeli- hood, is the middle pleistocene race of man that occupied part of Europe. It is interesting to compare with the sepultures of the southern littoral of the Cape the discoveries of the ‘‘ Grimaldi” cave, with reference to this negroid type of which two skeletons of small size have been found in the lower stratum. One is that of an old * In Bohemia the Neolithic race that inhabited that country is designed as ‘‘ people of the tucked skeletons ”’ (Pié Starozinotsti Zemé beské, i., apud Déchelette). 174 Annals of the South African Musewm. woman (1 m. 54 em.); the other of an adolescent from fifteen to seventeen years of age, and both are in the tucked position. : ii ‘ ve es 7 . i po . hy ‘ . i) — Ann.5.Afr Mus.Vol. VIL. veelad@. 74 x 3 West, Newman collo. co x Ann.S. Afr Mus.Vol. VIL. 76 nat size 77 nat size. 144 FISHHOEK cAPE z oz 4 G< 50 ce = On o ° 85 nat size. 358 x 125 x 67™- 84 90 xf West, Newman collo. : 1194. DA. KANNENEYER” _SNITHFIELD.. OR Pil Te fewmen callo. I Ann.S.Afr Mus.Vol. VII. + N2ix 36)0um: 105x24"™™- O36 "2m: 107 106 ole UNE 109 West, Newman collo, s ; ~ d : 7 Ann.S. Afr Mus. Vol. VU. » Pixie 3 4 HS wnat.size. 2 3 4 117 rat size. West, Newmar collo. x 3h, 19 Ann.S. Afr Mus. Vol. VIIL. OO —— West, Newman collo. © CRADOCK.CE. | a SMS EN ec WPOUV HQ o Ann.S. Afr Mus.Vol. VUI. 127 nat. size. West, Newnan collo. 4 size. 125 na 126 nat. size. (Ze. vu S Da 129 iw ie ir ba) a is < Sir i aan irs 8 - “s * s 7 - Ann.S. Afr Mus.Vol. VII. mC 4s 133 137 nat. size. PIEXVIE 136 x ho West, Newman collo. = - : : a i 7 n = ® I. Ann.5. Afr. Mus.Vol.V PI. XVIIL. J COOP CORE eD DOR NTIY RYTNYY ee6 West, Newmar collo. Ann.5.Afr.Mus.Vol VII. Ce aie ) iis oe BORD resin Sia aise Ig 20 30 31 oe 33 144 x 4/5 147 nat size. 149 Approx. nat. size. hdGe ade Poddodolede... eceqgd eC O COC COCO CGO OO oEOceegoeec €fo0 fo COICO eoGCo CVC eEeecoaoo0e00o0rvr00gn 3 eeogdro~eQg@ocoo0retaoo00axaovogqg0oo0:’ CHOOBCECOoto0ECONOO0QG0K GOGO at Cp Agee Coec0oOeoeoReC Cones oaooeoeegeac aco Lie eee 2009OSC7OOG00QO000GQQ000006 eoGOoodooegoogoov00aggoaodgscago eodod0occsco C°c0eCOeOnRODOO 90860006 OFGASGOCOLECHOHGV0OFLICAaGD 3006 Daeksavgescusssenndg SGEOASbGHhseacenas & seo ; NR / hs t *. \ ‘ac RT ye Pll lg \ { 152 Approx. Ya 50% /5 I5| neat srze. West, Newman colle Ann.S. Afr Mus. Vol. VIIL 155 Pl Oe i] : | West, Newman col/o, it Ann.S. Afr Mus.Vol. VIII. Pl 162 West, Newman colo. gee ‘ Ann.S. Afr. Mus.Vol. VIL. Piao 166 167 Approx “5 West, Newman collo. West, Newman colo. Ann .S. Afr. Mus.Vol. VIII. {3 Zale ee /3 x 178 igo x/s Pe. West, Newman co//o. te Ann.S.Afr.Mus.Vol. VIL 184 sat. size. 186 nat. size. 190 nat. size. 92 xe Nest Newmen collo. ee ey Ure AG vs Py ay ee Sper a 7 ‘a? 4 ‘yer ~< i * . 5 - ; a : 7 aT. ,] “aN " e Z 7 A ; ‘ = * : ° oa = ; ‘ : : oe 4 7 : ag = —_ oa) . ; F cha ; 7 es.) 3: . we ene - ’ ’ a 7 ¥ 7 i red ’ 7 md - 7 2 i > ay oe o 7 od f | | => : : | A @ i mB % . : : | | . : i : - > . | anh a Po ae 9 > te S _ ‘ # ay ied : * = | 4)? cs 7 by | 7 . ‘a ¥ * . | 7 : . ; ' a rg ‘ F ’ ‘@ : 1. -_ ; | e: a : ia ‘ 7 | | | 7 | i = y . | a] : : | | | q an i : | | | ry = a ¥ ¥ ~ 4 . . . Var we, | | | ar | => . | | : - : = : , | 1 bd : | : = i ay i 7 Ay - t i | . ate Ann.S. Afr Mus. Vol. VII. 193 x 196 xo s/zeé. 194 nat. TR RRR Aree RP RE leas aa 3 West, Newman collo J 25 y 198 x 202 x's ee @ All|. 189 x¥ as) ‘N G 207 nat (Zé. 7S 203 na size 206 nat size 204 nat West, Newman collo. > "Oo > vp) | - & = 2) q e bs : , ] CAS aS; ) gy . “ob ee de SiiY a se ? ti ues ANNALS OF THE ‘SOUTH AFRICAN MUSEUM. VOLUME VI. PART © contawning::— 1.— The Stowe Ages of South Africa as represented in-the Collection of the South African Museum. By li. Psrineuny, D.Se., Dorector. (With Plates XXVIII. and 26 text-figures.} A Chapter on the Sources of Rock for the Manufacture of Sione Implements. By A. L. Du Tort, of the Cape Geological Survay. A Chapter on Bushmiim Craniclogy. By FC. SHRUBSALL. (With 9 text-figures.} ISSUED JULY Sth, 1911. PRICE. 40s. PRINTED FOR THE TRUSTRES OF THE SOUTH AFRICAN MUSEUM By West, Newman & Co,, Lowpon. PARTS Cf THE ANNALS PREVIOUSLY ISSUED: Voi. l.—Part 1, 7/6: Part 2.10/—; Part 3, 5/-; complete £1 2s. 6d. Vol. 1Y,—Part 1, 2/6; Part 2, 5/5: Part: 8.) 1/43 Part 4, 2/63 Part Sy 1/+3 Part 6, 2/6; Dare 7, L/+» Patt 8; 2/6} Parh 9," 1=/; Part 10,6/-; Part 11, 2/6; Index, &e.,1/—; complete £1 8s, 6d. Vol. Iil,—Part 1, 2/-; Part 2,1/-; Part 8, 5/—; Part 4, 2/6; Part 6,.5/-; Part 6, °6/—; Part. 7, f=; Part 8, 2/65 Part 9; 1/-; Listen, Drtla, Ken Lin ig i j . complete £1 Gs. oe Vol. £V. (containing Paleontological papers published in conjunction weth the Geological Survey).— Part 1, 10/-; Part 2, 6/-; Part 3, 4/-; Part 4, 4/--; Part 5, 2/-;) Part 6, 4/-; Part'7, 12/6; Part 8; F/-- ; ¢ - consplete £2 9s 6a. Vol. V.—-Part 1, 4/+; Part 2, 7/6; Part 8, Q)--; Part 4, 1/-; Part 4,'1/6; Part 6, 4/6; i Pari 7%, 2/6; Part 8, 4/--; Part 9, 4/-. Vol, VL —-Part J, 12/-; Part 2,.4/-; Part 8, 3/-; Part 4,:27/-. Vol. VIE. (containing Paleontological papers published iu conjunction with the Geologica] Survey).-— Parti: (AIG: 2099 Vio Part 8, 4/63 art 4, T/+. Vol. VithaBact dope, Co ih eee : Vol. X.-—Part dy 2/6, Vol. XL-—Part 1, 3/-. ; The Annals of the South African Museum will be issued at irregular intervals, as matter for publication is available, Copies may be obtained from~— Mussrs. WEST, NEWMAN & Co., 54, Harron Garpen, Lonnon. Mussrs. WILGIAM WESLEY & SON, 28, Essex Street, Stranp, Lonpon. Musers. hive de & Co., Cann Serasce, Bran, Ur, THE LIBRARIAN, Sours Arzican Muszum, Capr ‘llown. “an ta a ¢ > a BL/WHOI LIBRARY OCC ee AA A NARANARAR AS AAr aa PA ae. Ane WA ogee ake : af AA ARAR A. ae A PPS a, OAARBA AANA Anny alaata) aia A AAAAN RY AAARAAAAA ANRAAAIAN anaean a fad A A “WAan \ APR AA ARAA : AANA a annanaarranarry Rn ya [NE Na\e / Nin Al a A AAA A Az A RAAR A ; AA A A aN ~wAL $6202 05Ae Rahana.an oe Al LAADAAS Ya aa AARON, A A ye NA a ‘: v 5 ex Ag ot a MEG ORE MERC, Ce COR Ee me € CC COSCO f 6 CO CLECune C & NA aN ea, a: oe ~ 2% aft . . ARM AAR - fim S| [x : FAS J f YA tN ry OR NAAAN NAA: Yala WARAAAAAAS CEC OO CE & (Ge q ® 4 CC +S CO HOMME Ke COC CO & Oil, Ot Cc MO OGL: Ge eos