Ergebnisse Naturwissenschafilicher Forschungen auf Ceylon von Dr. Paul Sarasin und Dr. Fritz Sarasin Wiesbaden. C.W. Kreidel's Verlag. 1887-1893 und 1908 Vierter Band: Die Steinzeit auf Ceylon Wiesbaden. C.W. Kreidel's Verlag. 1908 Results of Natural History Research in Ceylon Volume Four: The Stone Age of Ceylon by Dr Paul Sarasin and Dr Fritz Sarasin Wiesbaden: C.W. Kreidel's Verlag, 1908. Translated from the German into English by David Bulbeck, June-August 2006 Preface [page V] When we wrote the preface, in February 1893, to the three volumes of our Ceylon work that had been completed by then, and finished the necessary preparation for our research into the island of Celebes [Sulawesi], we had not imagined that the results of our natural history research into Ceylon could be further extended. Each of our journeys to the Dutch possessions of the Indo-Malaysian archipelago had brought us through Ceylon, and so in early 1902 we paid our third visit to the Veddas of the Nilgala region. Nothing additional for our research into this group of people emerged at that time. However, following the discovery of the Stone Age in the Toala shelters of South Celebes at the end of that year, our attention has repeatedly been drawn back to Ceylon. Until then nothing had been written about any stone tools in the island of Ceylon, but we would have to say that if the Toalas of Celebes had passed through a Stone Age, in all probability their relatives in Ceylon, the Veddas, would also have experienced the same history, despite the absence at that time of any knowledge on the topic. On its own, this consideration would have hardly been enough to cause us to devote ourselves again to our Ceylon studies. However, despite all the anatomical, ecological and historical reasons against it, a view had appeared in the literature that the Veddas were simply neglected Sinhalese, nothing other than a degenerate limb of this native people. For the definitive dismantling of this hypothesis, we saw only one approach that could lead to certain success. This was the attempt to demonstrate the existence of an original culture in Ceylon belonging to a period prior to the arrival of the Sinhalese immigrants, as recorded in their indigenous history telling of the early historical period. Should the reality be unmasked of the pre-Sinhalese Stone Age referred to here, particularly in the caves of the Vedda lands, then all questions regarding the autochthonous status of the Veddas would dissolve. The motivation which brought us again to Ceylon in the winter of 1907 was therefore the hope to find such traces of a prehistoric people on the island. The results of our campaign constitute the contents of our fourth volume, which in our opinion is a welcome, indeed a necessary supplement to our Vedda monograph. Appropriate acknowledgments should be made for the images we present of the prehistoric Ceylon material. It seemed important to us to allow the reader a true judgment of the material in hand, and we gratefully thank the firm of Brunner & Co. in Zurich for their valuable efforts and work which were required for the reproduction of our photographic plates. Basel, July 1908.Contents of the Fourth Volume — The Stone Age of Ceylon Preface p.V. Exploration for Caves and Report on the Findings (F.S.) p.1. Lithies (P.S.) p.23 Literature for the Lithics Section, p.56 Artefacts from Organic Material (F.S.) p.58 a) Artefacts from Bone, p.58 ) Artefacts from Mollusc Shell, p.66 ©) Artefact from Wood, p.72 Postscript to our Work on the Toala Caves of Lamoncong, Celebes, p.73 ‘The Animal and Plant Remains from the Caves (F.S.), p.75 a) The Animal Remains from the Caves, p.75 b) Edible Plant Remains, p.86 Summary of the Animal Remains, p.87 ‘The Human Remains from the Nilgala Cave (F-S.), p.90 Cited Literature for the last three sections, p.93 {page 1] Exploration for Caves and Report on the Findings. F.S. By the year 1885, we had already had our attention drawn to many examples of caves in the hinterland of Ceylon, but we had not once made a discovery of the Stone Age there. This is because we had not been sufficiently methodical to undertake excavations, and because we were insufficiently prepared for our exploration to recognize the distinguishing marks of the stone implements. We had been expecting to find stone axes, or knives and points of flint similar to those in Europe. ‘As we earlier explained in our Vedda work (Die Weddas von Ceylon und die sie umgebenden Volkerschafien, Ergebnisse naturwissenschafilicher Forschungen auf Ceylon, Volume 3, 1892-93, page 380), the term “cave” is not quite correct for Ceylon. Because the island lacks any limestone formation, we do not find any solid and deeply coursing caves; rather, the correct term for caves in the Ceylon context is shelter, or rock shelter [abris sus rouche, in French]. Sinhalese and Veddas agree in calling these Galgé or stone houses, which is derived from Gala, which means stone or rock, and Gé, which means house or home, These stone homes are made from tablets or blocks of mountain gneiss that have fallen onto the ground in such a way as to create a rock shelter at their rear, or (in much rarer cases) from two of these blocks leaning against each other to create a tent-like room. Many inhabitable recesses can be created both from erosion and the well-known weathering of gneiss into sheets. Ceylon has an enormous number of such rock shelters, from the smallest slippages to major semi- domes which can be many metres high, wide and deep. In Europe it is essentially caves that hold the traces of prehistoric people. So for Ceylon we think of seeking the traces of the original people and indeed the Veddas in these [page 2] natural houses of stone, and see no reason why this should not have been true from early times. One severe difficulty that disrupts the examination of the caves of Ceylon is that many of the stone ‘homes have experienced major, secondary disturbance of their floor. In the period of efflorescence of Buddhism, a strong urge must have often been felt to flee into isolation and lead a holy life there, and 80 a proportion of the caves have been tured into living cloisters and Buddhist temples. Even today there are such stone temples; the most famous example involves the sequence of caves (with endless statues and frescoes) in the Dambulla Rock.The prior use of a cave as a priest’s residence or holy place can be recognized immediately by the existence of an engraved channel for the generation of dripping water. In these cases a channel engraved with sharp outer edges, from top to bottom, is present on the exterior surface of the gneiss cliff face. This is so the rain water cannot wash across the rock face and follow its contours into the interior of the grotto, but must drip from the edge vertically to the ground. This way the caves remain inhabitable and dry even during the wet season. In these cases there are occasionally rock inscriptions in Pali, and, if the cave lies on a high flank of rock, steps engraved in the rock surface leaving the cave. ‘Whenever rain channels or inscriptions occur on a cave, we also find installations. Frequently the grottos are closed off by a wall, reaching to the roof, built mostly from bricks and clay. These may have doors and windows, and the interior divided into small rooms through cross walls. Additionally on rare ‘occasions there may be a declining Buddha statue made of brick and mortar. During the laying of the foundations for the walls and for building the installations, the original floor is severely disturbed, while the exposed rock face is occasionally shaved off. There can also be a thick secondary layer of fallen bricks mixed with yellow clay and infested with termite nests. In combination these factors produce a huge obstacle io excavation, Moreover, we shall later talk of the disturbances to the cave floor created by burrowing animals. Plate 26 and Fig. 48 of our Vedda Atlas show a typical Ceylon rock shelter that we had explored in the year of 1885. It lies not far from the little place of Nilgala in the eastern hinterland in the middle of ithick forest (see our map of Ceylon in Plate 1 for the location of the place). Its name is Bisokotuwagalgé, and its existence is due to a major slippage [page 3] of stone blocks slanting against the surface of the gneiss. Soft earth from the parent rock, without any other inclusions, makes up the floor of the shelter. In the same year we also looked for caves in the forests to the north of Mahaoya. This village lies on the large strip of land running from Badulla as far as Batticalao (see the map). One of the caves has a roof made of an overhanging unhewn block of stone, which bears an inscription in Pali, while the other is made of two blocks leaning against each other, with a rain channel. Both had served as holy Buddhist places. From the testimony of the Ratamahatmaya these should be temples from the time of the kings of Walagambahu, in the last century before the Christian era. Excavations in both caves produced nothing other than a few pot shards. The results from a cave in the Omuna Massif, about 12 miles northwest of Mahaoya, were no better. This very lengthy shelter had obviously been used in recent times as a temple. In it we found a massive declining Buddha, half disturbed, of brick. In the soft dust (a foot deep) of the cave floor lay a stone Buddha statue without its head. The head had been lost when rocks fell from the wall against which the statue had originally been standing. At one time the cave had also been closed off by a wall. Despite the fact that we had once been very close to the discovery of the Stone Age, and had held artefacts in our hands, we did not recognize them. After we had stopped at a rest house somewhat northeast of Mahaoya for the purpose of an elephant hunt, on 28 February 1885 we undertook an expedition in a boat to the river feeding the lake, and then followed a channel upstream and came to a ‘gneiss outcrop knoll, up which led about a hundred small steps. At the top there lay brick rubble from a building in the shape of a cross, along with hewn quadrilateral pillars of gneiss. As our diary recorded it: “We found many splinters of quartz that have good points and are flat; it looks to our eyes like a natural formation, because to be useful as weapon points the quartz would need strike-alights, which are lacking”. We were only convinced of the fact that these stone points must have been stone implements during our following travels, those that we describe here. The surface of a gneiss outcrop very frequently breaks up into scaly tablets, and is often filled with rainwater whether from minor or ‘major downpours, permitting no rain or storm to pass through because of the manner in which thegneiss tablets lie. Stone implements could lie there for thousands of years without being washed away, exposed through the wind-bome removal of the sand and dust from the weathered gneissic material. From 1885 onwards the question of Ceylon’s prehistory lay dormant. So convinced were we of Ceylon’s likely lack of a Stone Age that, when we set on a new visit to the Nilgalaland with our friend Prof. Leopold Riitimeyer in 1902, we did not pursue his suggestion of a renewed cave excavation there, [page 4] Then, when the Stone Age was discovered in the Toala caves of Lamoncong in the year 1903, the situation changed. Even though no-one had made any further enquiries into the caves of Ceylon in the meantime, we decided to return to this question and undertake a new journey to Ceylon at the start of January 1907. Our plan was to investigate the caves in the various parts of the island, beginning in the south and proceeding northwards, and in this way cast a network of probes across the land. For our first probe we took a cave, near Kataragama (Katragam) in the southeast of the country, about which J. Davy (1821, An Account of the Interior of Ceylon etc.) had written the following (page 419): “ten miles from Katragram we came by a massive outcrop named by some as Gallegay (correctly Galgé) and by others as Kimegalle. The former name means stone house, and refers to some roomy caves in its flank where the traveller can obtain good protection, and the latter name means stone water, because of two deep holes on the summit which are natural reservoirs and which never lack water, even though this element [water] is often extremely rare in this arid land and can hardly be found elsewhere” [German translation of Davy’s English translated back into English]. This combination of caves with a reservoir seemed so propitious that we decided to investigate the place and begin our excavations there. In Kandy we procured the necessary recording equipment, shovels, picks and crowbars, and hired 12 coolies experienced in digging. The departure point of our trek (with our equipment transported by dray oxen) was Bandarawela, which was then the final station of the Ceylon mountain road on the southeast edge of the Central Massif (near Haputale on our map). From there led another route, a good overland path, down the southem mountain edge towards Wellawaya which was the first village in the lowlands (614 feet = 187 metres), and then through torrid heat and troublesome dust to Tellula. Here a farmer in the rest house told us that he knew of two caves in the region, one to the west known as Kalugalgé (black stone house), and the other to the east known as Demiyagalgé. Even though we had intended to start out study of caves in Katragam, we believed that this opportunity should not be foregone, especially as the site was quite near. This last observation is not an accurate statement, in that we needed a full three hours to reach the place, especially as our guide was not always sure of the path. The first part took an hour southwards along the main path, then we tured eastwards on a small path and crossed over the Kirindeoya, and then came a veritable bush track inhabited at occasional places by Sinhalese Roddies, then high forest to one of the old reservoirs (Radgamawewa) so characteristic of Ceylon’s hinterland, completely overgrown and inhabited by white herons. Another hour [took us} through forest and parkland, Small open places with short grass lay [page 5] hemmed in by bushland where some higher trees sprung up with crooked, knotted and often dead branches, grey stems and a light canopy, interchanging with higher stretches of forest rich with lianas. Finally after long hours in the wilderness we stood in front of a gneiss knoll, up which led a row of stairs, There we found several caves with the remains of built walls, which evidently had to do with Sinhalese dwellings, temples or hermitages. The floor was made of a hard yellow earth studded with strong quadratic fire stones. As our coolies were already tired by the long march and time was short, we gave up the excavation after a ‘metre without coming upon signs that the floor of Sinhalese rubble had come to an end. From the top of the outcrop we gained a glorious view of an unbroken sea of forest, as far as the Central Massif to the north, while to the east emerged a chain of hills from the forest wildemess. We did not investigate Kalugalgé, which is the other cave. Our onward march from Tellula next brought us to Tanalmalwila (not on our map, as is also true of the following places) lying on the Kirindeoya. The banks of this river are shadowed by enormous trees, 4whereas the forest farther away from the water, through which our overland route led, consisted of miles and miles of unattractive, low bush with a half dried-out undergrowth. Then over Wirawila, whose half-collapsed reservoir, overgrown with grass and white flowering lotuses, offered waterfowl nourishment and a place to stay, to Tissamaharama (whose highest point lay at 162 feet = 49 metres above sea). The small village of this name is the place of an ancient wealthy city dating to 300 or 400 years BC, whose extended ruins make a great impression on the visitor. From here one can, within a short half day, proceed through desolate bushland growing on a stony and sandy soil, to Katragam, famous as a holy Brahmin place. However, this is not the place to describe this temple, and we continued towards our destination which is the cave lying 814 English miles northwards of Katragam. Ina remote area here, where elephants abound, a long gneiss ridge rose to a height of 30-40 metres. Deep inside it, basin-shaped cavities without a roof contained ponds of rain water, where cyclopids, ‘water mites, insect larvae, and frogs (the common Rana tigrina Daud) lived. These frogs which keep their young in a foam nest (just like Rhacophorus) must have this body of water. This we observed ‘many times from the visible foam of their nests stuck against the rock above the water. The nest lets the tadpoles fall directly into the pool underneath (for an explanation, refer to our Vedda work, page 60). On the northwest side of this block of gneiss there is a 15 metre long shelter, neatly defined by a layout of soft strata. It does not have the same roof height for its entire length, but instead both ends incorporate [page 6] deep cavities, which are joined together by a small overhanging middle piece. The accompanying plate, Fig. 1, shows the northerly example of the two deep shelter roofs, built from tablets which had fallen southwards. (The floor is somewhat raised in the picture, as a result of our excavation.) The southern cave was very similar, except that the gneiss strata dipped northwards, and the direction of collapse turned back on itself after a short stretch of 15 metres. [Fig. 1. Northem section of the Galgé of Katagram] We then decided to set up quarters in the southern of the two caves. We covered the powdery floor with a canvas to take our field equipment, and began our excavation in the northern stone house, which is the one illustrated here. It is 24 metres high, 5 long, 4 metres deep, and opens towards the northwest. Traces of smoke on the walls, fresh ash, and blackened hearth stones showed that this cave was frequently used by passers-by. It lay on the route which to the north led to the Temple of Katragam and onward to Tissamaharama, and south towards the coast to a district which (understandably) had no village anywhere, During our stay of six nights we often encountered evening travellers who wished to overnight here. We then made a trench over two metres long in the said cave, The deposit was coloured grey from ash in the upper layers. There were also hearth stones (suitable as whetstones when exposed on the surface) which could have served for preparing curry even today, numerous (page 7] fragments of ceramic containers, and animal bones of recent appearance. These belonged to both types of hind (4xis and sambar) and also included a monkey canine (Semnopithecus). Further down the deposit became browner and harder. Broken pieces of gneiss frequently protruded from the sides of the trench; these were often bulky tablets which could only be moved after being reduced to rubble, and this made progress heavy work. We proceeded about 1.5 metres deeper in the manner described, with our major observations being that the pot shards became thicker with depth and were more decorated towards the top. All of the recovered shards appeared to have come from containers made on a potter’s wheel. We can certainly denote this layer of around 1.5 metres thickness as a Sinhalese culture layer. Its contents obviously reflect the use of the cave as night quarters for travellers, since there were no traces of its use as a holy Buddhist location or hermitage. This indicates that the path running past this shelter must have a great age, and could reach back to the time when the state of Tissamaharama had been 5founded. Therefore this layer of 1.5 metres thickness would go back approximately 2200 years. The weathering of the gneiss rock must have proceeded rapidly, which can be easily explained by the presence of cavities with water in the overlying surface of the gneiss, and the water seeping through all of the crevasses. Below 1.5 metres there was a zone of around 30 cm thickness, without any discemible traces of a human presence. Then we came upon a thin layer outlined in the shape of a nest, which contained some bone splinters of dark brown colour, mixed with quartz chips, whose appearance suggested they had been knapped. That is, these knife-shaped objects showed themselves to be artefacts. Immediately beneath this, at approximately 2 metres of depth, occurred a layer of gneiss blocks which had fallen diagonally in a row. This did not appear to be bedrock, because 60 centimetres further down we came upon a soft mass of cheesy, snowy powder, whose source is not clearly known. We now vacated the cave that we had been inhabiting till then and moved into the one that had been probed, so that we could start an excavation in the former. Here as well the superficial layer was made of loose grey ash; at around 60 cm depth the deposit became browner and firmer and richly studded with fallen blocks of gneiss. Pot shards continued to a depth of 1.5 metres; then followed same old- looking pieces of broken bone and some flaked quartz, till at 1.7 metres we again reached the weathered bedrock, whereupon the excavation (naturally) was closed. ‘As already remarked, the gneiss layers dipped towards the middle from both the north and the south to make these elongated shelters. This point suggested that there should be a greater depth to the floor of the cave [in the middle]. We therefore lay out a third excavation in between the two deeper caves, beneath the shelter roof which projected only slightly. However, the large [page 8] blocks of gneiss that had fallen down there were truly laborious to pull out. In this excavation at a depth of 1.7 metres we lifted out two large hearth stones, and pot shards continued to 1.8 metres. Then we again hit a small zone without any finds and underneath it, between the depths of 2 and 2.3 metres, a rich assemblage of flaked pieces of quartz. and some charcoal, After this there was no sheet of bedrock, but instead a layer of thick blocks, interlaced with bands of soft weathered material, so that further work appeared to have no prospects. We informally observed that the gneiss blocks in this layer were fresh and hard to a depth of | metre; farther underneath they became more rotten and brittle, tll they had the consistency of cheesy snow below 2 metres’ depth. With the advantage of hindsight, we would not have doubted for a moment that the quartz pieces, of which truly many had been extracted from the depths during the third excavation, were veritable artefacts. Although a few pieces had the thoroughly typical form of knives and points, and core-like pieces with flaking scars were likewise present, the material was very difficult to judge at that time. Had it been made of flint, we would have been sure of having discovered the Stone Age, but the properties of quartz were unknown to us, and we did not know whether nature unaided by human activity could produce such objects. The major point of difficulty was that the gneiss in our caves itself included veins of quartz, meaning that it would not have been necessary for humans to have brought pieces of quartz into the caves. It is true that we were able to determine that the quartz. which weathered from the gneiss, and which littered the floor of the caves in abundance, was of much lower quality than the quartz used for implements. The latter frequently resembled crystalline quartz, and moreover the vein quartz weathered into formless pieces, of a cloudy colour, which no-one could have used for artefacts. The observation which further encouraged us to believe in a Stone Age is that all three holes had yielded the flakes of quartz beneath the Sinhalese culture layer which is characterized by pot shards and nothing in the way of quartz except raw, weathered pieces. The final clue was a find that we could readily trust in, a smail core of broken chert, which obviously must have brought into the shelter by people. But — and when people wish to doubt, there is always a “but” — could this piece not ultimately be a stray piece of Sinhalese hearth stone? Could it not have dropped a great depth though the frequent mouse holes in the floor, and been mixed in at the depth of the quartz splinters? 6While for us it was more than probable that we had discovered a true Stone Age in the Galgé of Katragam, we were still not prepared to risk communicating this result by letter. So after five days we decided to satisfy ourselves with another excavation program, and, to confirm [page 9] our expectations, in some other place where a completely undeniable proof might emerge. ‘We now wended our way northwards with the intention to penetrate the true Vedda lands in the district of Nilgala. In two short days of trekking we reached the little place of Buttala, lying on a large path which is not yet entered on our map, and which led eastwards from Wellawaya to the coast. Our inquiries here had produced reports of good caves in this neighbourhood, so we did not wish to overlook this opportunity. A farmer from the district then led us a mile northwards to the footpath which leads over Alupota as far as Passara, and then turns northeast. After a further mile through bushland and dry rice fields, we reached the hamlet of Andanpahowa at the foot of a stony, wooded chain of hills called Meminnahela (mongoose hills). A short four hour climb through tall forest brought us to a large grotto made through blocks of stone heaped randomly together. Its floor revealed shallow depressions, the sleeping places of sloth bears. ‘One unfortunate circumstance (as we could easily and immediately see) is that the rainwater could penetrate the cave and mingle with the ample earth. The excavation through the tablets of gneiss that had fallen thickly together here proved to be truly laborious. There was not a single pot shard, and after approximately one metre’s depth we reached bedrock with its rind of weathered product. In the direct vicinity there appeared a second grotto, very beautifull and roomy, opening south-west- west, which had been used as a priest’s dwelling as shown by its hewn water channel. Here as well the work, carried out at two locations, was made more difficult by large blocks of stone in the floor of the excavation. At around 50 centimetres depth we found a rusted Sinhalese knife-pick of iron, an iron ring and a clay offering dish. All in all, the shelter floor was rich with pot shards, but unfortunately of very limited depth, for our first hole stopped at bedrock at 80 cm and our second at 1.1 metres. It was entirely an unalloyed Sinhalese cultural layer, rich in shards. Water could penetrate the cave through a hole in the rock, and as the earth also contained pot shards, it was clear that somewhere higher in the hill another, earlier inhabited cave would have to be found. In fact this phenomenon lay open for exploration directly above. It contained remains of walls and large masoned stone blocks, showing that it was an old stone temple; we have not explored its floor as we were discouraged by our previous inauspicious results. It is worthwhile remarking that whenever negative retums come from an excavation, the pieces of quartz are never like those we had found in the Katragam cave, but instead are formless pieces weathered out of gneiss. This seems to us a further indirect proof of the artificial nature of the [Katagram-style pieces], for they should have been here too if they were natural products. This was accordingly a minor benefit from our second effort at excavation. {page 10] The next investigated cave lay due north at the Yakunehela, or Mountain of Demons, between Bibile and Nilgala. This name appeared very promising to us as an indicator of old habitation by Veddas, who are known in the Mahawansa as Yaka or demons. The said stony contour line was exceptionally picturesque in its open parkland; at its eastern foot (ten miles from Bibile) a large cave revealed itself in its wild rocky landscape. Unfortunately there were remains of a wall here too, and the excavation, which soon came upon bedrock, produced nothing other than shards and bones. In front of the cave a small deposit had been built up along the drip line, and here we found some water-rolled pot shards as well as some good flakes of quartz and crystalline quartz. These obviously must have come down from a cave lying higher in the mountain, On this matter our guide indicated that he knew nothing, and to look for it alone in the mountains and in the thick undergrowth did not appear to hold out any chance of success.‘So we proceeded further to Nilgala (height about 200 metres); the chain of gneiss with its dome-shaped ‘summits is represented in the introduction to our Vedda work (page 14) as a painted background of the place when looking southwards, An old farmer here was happy to show us a cave that he knew as Gangodedeniyagalgé. We walked over the Galoya or Patipaluru River, which runs past Nilgala as a wide and surging stream, then hiked for half an hour further to the stony hills of this name somewhat southwest of the path to Medagama, and then turned northwards and reached the south-south-west precipice of the Bulupitiyahela mountain of stone, at whose northeast foot lies Nilgala; the Galoya flows past this mountain bending northwards. [Fig. 2. The Nilgala cave] [page 11] A mighty block of gneiss here had fallen from the main mountain and come to rest so as to form a grotto opening to the southeast. The text figure (Fig. 2) gives a good impression of this romantic stone house hidden in a copse. The stately height of the semi-dome can be judged from our Sinhalese guide standing in the picture. In the vicinity was a permanent spring which, before disappearing under a nearby rock, served as a watering spot for the numerous game animals which broke their journey here. Above the large block which created the cave, one reached a natural sea of stone made of rubble fallen together. Here there lay a second, much smaller cave with the remains of a wall, which rainwater could bypass, to one side. The large cave, which we shall call the Nilgala cave for short, finally gave us the longed for, undeniable proof of the existence of a Stone Age in Ceylon. The first probe into the cave floor had already yielded the desired result. The upper layer consisted of grey dust which contained much ash along with numerous pot shards and some bones. But already at 35-40 cm depth quartz artefacts began to appear, and, as we particularly welcomed, red and yellow chert of a material exotic to the cave, and indisputable cores, as well as flakes and unambiguous points and knives, mixed with splintered and charcoal-stained bones of game animals. During the next week we systematically removed a good proportion of the cave, but without preventing our successors from checking our results. In particular, at the front of the cave we left an undisturbed baulk which undoubtedly contains a quantity of finds. In the cave we undertook several scientifically exact excavations, proceeding layer by layer and recording the finds by their position in the cave. However, we soon came to the realization that this was pointless, because the character of the artefacts ‘was always the same. Also, many sections of the cave were disturbed, which we knew from finding Sinhalese pot shards mixed up with the stone instruments to a depth of one metre. The disturbance factor here did not seem to be human activity, as in the other caves, but instead animal burrows, particularly from porcupines, pangolins and sloth bears, and otherwise termites. The whole floor was, clearly shot through with termites; overall the amount of deposit we came upon that was not riddled with termite nests or not empty was the size of a head. In search of nourishment the termites, even without help from the sloth bears, had performed substantial earthworks, with the result that objects must have slid around in the depths. ‘The floor of the cave proved to be unevenly deep and productive in different places. The poorest and least impressive contents came from the rear of the cave, where bedrock was soon reached; elsewhere [page 12] they were deeper and richer towards the rim of the overhanging block; often there was an intermixture of bands laden with finds and small blocks of unproductive inclusions. The deepest reached location lay at approximately 2 metres’ depth, where we came upon a layer made of a large, weathered block of stone. This proved to be a solid floor that prevented further work. Stone implements were constantly found down to this layer.In general the following can be said about the excavation. To the depth of 35 cm the floor was grey from ash and contained only pot shards, charcoal, bones, hearth stones, and formless raw quartz ‘weathered from the gneiss; the Sinhalese culture layer reached down to this level. Farther down the floor became firmer and more yellow, but interlaced at various depths with grey bands which were generally rich in finds, so that each handful of earth contained stone artefacts and splinters, bone, teeth and snail shells; often these items were found clustered together. This layer of rich finds lay between 40 and 80 cm depth and occasionally reached to one metre’s depth. Below that to two metres” depth (as already noted) there was a yellow layer without any ash colouring and where artefacts and bones were Jess common. The flakes of quartz, crystalline quartz and chert occurred in such huge quantities that we only collected those that possessed a known form and had been clearly manufactured as implements by removing flakes. Approximately 5-10% of the stone artefacts that came to hand were collected. Ithardly needs to be mentioned that we found pieces of gneiss broken off from the walls at all levels. Here however the fragmentation of gneiss must have been proceeding longer than in the Galgé of Katagram where the Sinhalese culture layer reached to 1.5 metres’ thickness compared to only 35 cm here. The block of gneiss that makes the cave was still sound rock, free of water penetration until today. This is because its strongly inclined surface prevented any damming of the water, and all rain water immediately drained away. The cave had certainly been used for a long time; until today it is used by Sinhalese passers-by, and also as quarters by wild Veddas (as people told us). It is particularly fortunate that the floor is absolutely dry, whereas in most caves in Ceylon the powerful rains drive ‘water in along the roof or through the sides. The latter namely applies to all of the caves along the foot of the mountain, which are weathered by heavy streams of water breaking through gaps in the rock. However, the rock which capped our cave is the high-mountain variety, so that the drainage water did not reach it, Even during the cyclones which devastated the region on the night between 9 and 10 March, and made the water flow in torrents for hours, the floor of the cave remained dry, even though the surrounding copse was ravaged. Two further caves were investigated at Nilgala. One of them, {page 13] Ballawalabokagalgé, lay at the southem end of the Danigala line of hills. To reach there, one follows for 40 minutes the path leading to Patipal on the peaks of the Ullhelagala ridge which ascends from Nilgala south-east-east, then one turns eastwards from the high path and crosses the Bambarakandura stream (15 minutes), and finally through high grass to the cave, This stone house as well is formed by an overhanging block and appears very pretty and picturesque. As is generally the case, Sinhalese walls had once upon a time been built to make a solitary dwelling. The rock face was partly covered by a lime wash and revealed a rain runnel exteriorly. The excavation was partly outside of the house wall but always undemeath the shelter roof. The yellow earth was riddled with bricks and fragments of clay, and at 50-70 cm bedrock was reached. No stone implements were found; the sole, probable prehistoric object was a carbonized fragment of a swine canine with use marks; in addition, very little broken bone tured up. The cave appeared to be of quite a young age and seemed to owe its existence, as was so often the case, to a block of stone that had fallen to its current position. Northwards of this cave lay a small one to the right flank, oriented across the direction of the Danigala ridge. To reach here from Nilgala we needed 1% hours. The circumstances were difficult, because the cyclone had created a lot of disturbance in the forest and built up entire walls of fallen trees; otherwise an hour’s march would have been sufficient to get there. The cave, Mietigeha Aregalgé, lay under a thick rock beneath which a little spring bubbled up; it is open towards the east. Even in this remote rock shelter in the wildest forest there had been a little Sinhalese dwelling, which further had a rainwater runnel and attractive smoothing of the rock face. Here the elephant rubs itself and covers itself with yellow powder. The floor of the shelter showed a lot of disturbance, as was to be expected. As the cave is damp, someone had poured large stones into it to make a kind of terrace; our excavation into the deposit 9encountered pieces of four-sided stone arrows and many pot shards. Were it not for the Sinhalese installations, without doubt a good booty from the Stone Age would have been exhumed, because in the disturbed ground we found a typical quartz core and also some broken detached flakes of chert. The cave is still visited by Veddas until now; above the surface of the earth lay the broken-off end of a Vedda arrow, sharpened at that place with an iron blade (see Volume 3, page 428), and near the spring a fiesh hearth with burnt tortoise scutes was found, Near it lay a broken chert core, bearing fresh percussion marks, which the Veddas had clearly used as a strike-a-light, It [page 14] was a perfectly typical Stone Age core, perhaps left by the Veddas themselves in the cave. With this we closed off the search for caves in the Nilgala region and left the place on 23" March. A brief communication about the discovered Stone Age had already been sent to Dr L. Riitimeyer in Basel on 7 March, and soon appeared in the Globus on 25 April 1907. ‘The finds in the Nilgala caves now filled us with certainty that the quartz objects earlier found in the cave near Katragam truly were implements made by people. Therefore we had now discovered two Stone Age stations at a considerable distance from each other. We now wanted to travel farther north to undertake a probe, to gain an insight into the extent of the indigenous inhabitants of the island. In fact, we had in our view the already known caves north of Mahaoya, as mentioned in the introduction [to this chapter]. We next tumed back towards Bibile (240 metres), prepared ourselves again, and walked for three days over Ekiriyankumbura and Kaloday towards Mahaoya. Here we had the misfortune that a great Buddhist temple festival lasting a week for the inauguration of a Dagoba had begun, and as a result no porters could be found. Of our own 12 porters, half of them ‘were no longer fit for heavy work as a result of various maladies, mostly malaria. We therefore turned back towards Kaloday (Pailegama on our map) to see if a statement could also be made about caves in the neighbourhood there. ‘A Vedda of mixed descent, who still occasionally used caves as his night quarters, offered himself as, our guide. The path led from the rest house south-eastwards in the direction of Marangala-Nilgala, ‘After about half an hour, the guide turned off this path into one running north-eastwards, to cross over the small Talayapitaoya stream after ten minutes. After a further 40 minutes the cave was reached, which was a low stone house no more than a man’s height; its name is Wauluwelagalgé. Inside it was a row of stone tablets laid so as to make a kind of sleeping platform, the harsh night camp of the Vedda accompanying us. Near it lay a small clay pipe roughly made by his hands, and a strangely double- pointed, wooden arrow coloured yellow and black. Our Vedda said that he had carried one like this in his belt, when hunting, out of veneration to the Sinhalese earth god Bakirawa (?). In the cave itself there was no cultural layer to find, only naked bedrock. A scree slope descended at the front, and we excavated it through its dense rock fall. To the depth of around 40 centimetres it consisted of black humus and contained nothing other than a few pot shards. Then there followed a red coloured, crumbly earth, which was quite full of quartz pieces for about 20 centimetres’ depth, without any accompanying shards. Further down the quartz also completely disappeared, even though the black covering humus layer continued. The cave was apparently of relatively little antiquity. At the beginning of its [page 15] period of habitation it harboured Stone Age people who tossed out the refuse from their activities; then came the historical period with its clay vessels. ‘The following day there was a variation to the yield from this rich find spot, in that a core and a chip of green quartzite came to hand. Then we searched out a second quite low cave, Kanadtawelagalgé, lying on a knoll approximately 20 minutes from the previous one. Our Vedda had occasionally sheltered under its roof as well; he spoke of a water container made in a Sinhalese form but without traces of wheel production, lying near his fire place, as his own work. The excavation in this cave, where one 10had to bend over, was made difficult by numerous stones and roots, and by the ferocious heat as the sun fully ascended. After a while, our people gave up on this excavation out of tiredness, before any results were produced. This did not however change the fact that the first cave had demonstrated the existence of the Stone Age, of the same character as further south, at the northern end of the current haunts of the Veddas. By foot we reached Badulla from Ekiriyan in four days, and continued on to the cavalry outpost of Bandarawela, the starting point of our journey. This place lies at ca. 1230 metres (or 4036 feet in the imperial system) at the highest point of the Uwa Mountains, whose surrounding lands are nicely represented in a photograph that we purchased in Colombo (Fig. 3). In this region, when you get above acertain height, you see a huge number of long rounded hills that often look like breasts, covered by a dense grey-green vegetation of bushland copses interspersed with grass. The green bands are the only strips of forest in the central mountains, and follow veins of water, at the bottom of the travelling fields, which soon come upon a narrow stream that broadens into a lake, In this landscape of grey grass, these forested strips are like oases in the wildemess, and help to emphasize the essential impression here of the play to light that one sees in a desert. [Fig. 3. The Uwa Mountains] As the hills were made entirely of gneiss, the surface weathers to a soft material. [page 16] The flat crowns of the hills prepared one more big surprise for us. Whenever we climbed on top of one of these, ‘we would suddenly come upon splinters of quartz and crystalline quartz, which had the same form as those we had excavated from the floors of the caves. Our astonishment grew as we found knives and points of red and yellow chert along with the cores from which these pieces had been detached. It was easy to see that all the quartz and chert had had to be carried by people to the hilltops and ridges from another location and that none of it could in any way derive from the stone that constituted the hill. These types of rocks were missing from the base of the hill, and from the entire, steeply increasing slope, all the way to the top of the hill (where no rain flow could reach). The reader can see the gleaming white quartz strewn between the grass clumps on one of the hilltops in our photograph (Fig. 4). The crown of the hills is either constituted directly by weathered gneiss or presents a thin surface of ‘humus; on one hill this was 15 cm thick and all of it contained splinters of quartz and chert. As already observed, the gneiss of the hills is riddled with veins of quartz. whose exposed surface [page 17] weathers. This vein quartz however breaks into sharply angled, broken, opaque pieces, which are completely different from the slender quartz flakes on the hilltops. [Fig. 4. Prehistoric quartz implements on a hilltop at Bandarawela] Not all of the hilltops that we investigated in the region of Bandarawela were equally rich in artefacts. The best example in this regard is the elongated hill which rises up at the middle of the small reservoir of Bandarawela (Fig. 5); at its peak there stands a wooden cross, held up by a stone cairn, which is also visible on our picture. Here amongst the other artefacts and the human-made splinters we found a small, round hammer with percussion marks, very similar to the one we had found in the Nilgala cave. We did not observe any pot shards on the hills, although they would have to be there if the Stone Age people had had any pottery vessels. These finds on the ridges clearly identify them as open-air prehistoric sites, and the quantity of artefacts indicates a long period of habitation by Stone Age people. One even gets the impression that the use of caves had only a secondary role compared to that of the open-air sites. MNA visit to the Veddas of Danigalakette in March 1907 had instructed us to think in terms of such open sites. In the unbroken forest that cloaks these mountains a high, bare cliff rises up. On its rounded ridge there stood a low, quite tiny hut which, so as not to be blown away by the wind, {page 18] was anchored by a rope of bast looped round a block of stone. It was a crude copy of the Sinhalese building style and had a main room and a small building in front of it. ts walls were made of plain planks of wood with yawning gaps that let the wind pass freely through. The original casing of bark had mainly collapsed, and the roof was nothing but a grass covering. In this hut six men cowered close to each other, along with three women and half a dozen children in the room in front. How all these people ‘managed to sleep in this tiny hut was hard to think; they must have slept at night in shifts. Beneath a nearby block of stone was a fireplace, which was used [as a sleeping place], though most [of the residents] slept in the open, next to a fire at the foot of a tree that gave some protection against the wind (Gee Vedda work, page 379). We had already received this information from a Danigala Vedda in 1885; he had called the rock which we visited Galkabala, and said that the people lived sometimes together and sometimes alone. Most of the time, he slept alone by a fire in the forest. [Fig. 5. The hill at Bandarawela] This was more or less how the prehistoric “housing estates” of the ridges of Uwa appeared to us. Now without doubt we had [evidence of] huts from the old system of simple shelters (Primitive huts, page 382), in the same way that the present-day Veddas still occasionally camp during their hunting treks. Were elephants in the vicinity, so would their sleeping places have been erected in the trees, as we {page 19] can see even today. Here is a noteworthy sentence to mention from J.W. Bennett’s report (Ceylon and its Capabilities, London 1843, page 253) over a horde of Veddas in the eastern province [translated into English from the German translation of the original English). “Asked whether any of them ever lived in trees, they shook their heads and, pointing towards the mountains of the Kandy district, they said: Those in that high and very remote district do so.” This is evidence worthy of note, because here could lie a tradition from an earlier occupation of the mountain lands by Veddas, of which the foremost, definite traces are the stone implements yielded at Bandarawela. It is accepted as known that the hills of Uwa, where the Stone Age people lived, used to be cloaked in forest, and that the Sinhalese destroyed the forests when they penetrated the highlands. The implements were then weathered by rain and wind (which removed all the light earth) from the humus of the burnt forest, so that today they occur on the upper surface of the knolls. Doubtless the hill ridges offered the advantage for habitation of greater dryness compared to the slopes and valley bottoms, as well as a better view over the landscape. Finally, they were healthier than the lowlands, because the malarial mosquitoes could reach there only with difficulty or not at all, although the original people would not have known that. Of course, we were not satisfied with the proof of the Stone Age at the height of Uwa, without addressing the omission of not having searched for similar sites in the region of Kandy (1696 feet = 517 metres). For a long time our troubles were in vain. (We found] isolated, intriguing splinters of quartz and crystal on the hills around the lake of Kandy, but many of these hills had been disturbed through housing construction and gardening, and others were completely barren of stone objects. The first indubitable implements were found on a hill named Bahirarawakanda, which one reaches from Kandy in around a quarter of an hour by following the road to Peradiniya and then turning northwards to climb its ridge for 20 minutes. This place had been cleared for a rubber plantation, and large numbers of quartz and crystal implements lay on the bare ground, albeit in smaller numbers than at Bandarawela, Nonetheless here was a Stone Age habitation in the central highlands, Then our people brought artefacts from a hill named Wattapolla in the Kandy district, and a few from the heights of Kadugannawa, 12In addition, during a visit we made to the Peradeniya Botanical Garden with Mr E. Emest Green, we happened to talk of our finds of lithics in the Vedda caves. Mr Green then imparted the information that he and a man named Mr J. Pole, in the district of Maskeliya, had long ago noted such point- and knife- shaped objects of quartz and crystalline quartz. However, they could not determine with certainty whether they were truly dealing with implements, especially as their field workers (who had sent them these pieces) expressed considerable doubt as to their artificial nature, so they had never published anything on them. [page 20] Despite this, Messrs Pole and Green deserve the recognition of being the first to have an awareness of Stone Age tools in the central highlands of Ceylon. They now had the proof that these were certainly artefacts, based on our discoveries in the caves, where these primitive quartz implements occurred with midden remains, and were accompanied by appropriate tools made of stone foreign to the caves. Without these cave finds we also could not have been confident about the artefacts that we had found on the hills of Bandarawela. On another day Mr E.E. Green guided us to a hill which rises on the left bank of the Mahaweliganga, close to the Peradeniya Station. It belongs to the Botanical Garden and is planted with Rizinus and mulberry trees for the nourishment of silkworms which are bred in a building that crowns the hill. At the foot of the hill we found only crude, broken quartz weathered out of the gneiss. Above half way we found several good, flaked pieces of quartz. These had obviously eroded out from above, as the rounded summit was rich in instruments of quartz and crystalline quartz, some of very good manufacture, even though chert was lacking. Holes dug there for the plants showed that the flakes pertained to the uppermost layer. According to Mr Green’s report, artefacts have been found in the Eton Estate at the height of 4200 feet (= 1280 metres). This plantation lies in the valley of Pundhuloya, in the district of Ramboda, west of Pedurutallagala, On our request Mr John Pole (already mentioned) then sent us the quartz and crystal flakes he had collected, for comparison with ours, for which we gave him our sincerest thanks. The first of these was a collection of 40 crystal artefacts, in part very typical, from Scarborough Estate (4400 feet = 1340 metres) in the Maskeliya district east of Adam’s Peak; then another from the district of Nawalapitiya (a railway station at 1913 feet = 583 metres), south of Kandy, another from Matale north of Kandy, and a regrettably dubious collection from the Horton Plains above Bogawantalawa. If this last collection had been incontestable it would have been of great interest, as it would have shown that Stone Age people lived or at least occasionally visited the higher mountains (1800-2400 metres), whereas the sites we have mentioned so far do not go above 1340 metres. In addition there were some pieces from a southern coastal beach in the district of Ambalantota (west of Hambantota) and the west coast of Puttalam (northwards of Colombo). Mr John Still made his own discovery during his excavations at Anaradhapura, which we should discuss with care. There he came upon a layer full of flakes of quartz crystal and carnelian at a considerable depth. In other circumstances, this finding would have been interpreted as prehistoric, but here it demonstrated [page 21] a Sinhalese industry of breaking up crystal and camelian, Both finished and half-finished pieces from jewellery stone, amidst relic containers shaped like pagodas, showed without doubt that this was Sinhalese manufacture. Mr Still added that in the Northem Province he had also found flakes of quartz and crystalline quartz of the same style as the [Anaradhapura] industrial waste, five miles east of Puliyankulum and around 20 miles north of Lake Padawiya, On these rocks a circle of crude unhewn stone 17 inches in diameter had been formed, looking over the installation of a reservoir; inside the circle next to some bricks he had found knapped chips of crystal and flint, as also in similar stone circles at other places in the vicinity. We have not seen these chips, and Mr Bruce Foot declared their artificial nature to be dubious. Certainly one may observe that when it is clearly a case of prehistoric instruments, they do not have any relationship with stone circles but instead are older. This is nicely illustrated by the existence of brick fragments inside one of the excavated stone circles. 13Through the discoveries of the gentlemen named above, in combination with our own, it can now be taken as certain that the whole island had had Stone Age people, from the lowlands to the central uplands, although we do not yet know up to which altitude this situation prevailed. Proof is still missing from the southwest, but we could not afford to spend any more time waiting for that. Without hesitation, we relate the stone artefacts from the caves and mountains to the ancestors of the present-day Veddas, and indeed to the proto-Veddas whose autochthonous status is in agreement with the accounts from the Sinhalese chronicle, the Mahawansa, We have proven not only the existence of campsites clearly beneath the Iron Age culture of the Sinhalese, but also, as the following sections will review, consistent associations with recent fauna, which show that the remains are relatively young. In the Ceylon highlands there are no longer any Veddas, and this raises the question as to how long they ‘were able to hold out there, and when their last descendants died out or were destroyed in the highlands. We cannot give an answer here, but our Stone Age finds give meaning back to the traditions which speak of Veddas in various parts of the island where there is no longer any trace of them. ‘The Mahawansa (compare here our chapter: On old stories of the Veddas etc, page 577) relates how the Sinhalese, on their arrival, found the northern and eastern lowlands to be populated by Yaka or Demons, by which Veddas are without a doubt to be understood, and how their main location was in the district of Bintenne, where even today some remnants of these people may be found. The central highlands were known as the [page 22] source of the deities, meaning either demons (again) or gods; at other places these deities were simply called Pulinda, which means hill or mountain people. There has been just enough anthropological study of the different names to let us know that these Pulinda should be understood as the original Vedda inhabitants of the mountains, whose Stone Age remains today cover the hill ridges. It is quite clear that the Stone Age population of the mountains must be the same people as those who left their relics in the caves of the lowlands, because their tools are made from the same materials, of the same dimensions and with the same techniques. On 5 May we returned by ship from Colombo to Europe, extremely content that the long awaited result ofa prehistoric Stone Age had been demonstrated, one of the richest across the lands of the earth, ‘whereas until then not a single stone instrument from there had been described in the literature. [page 23] Lithies. P.S. ‘When we glance over the prehistoric finds from our last campaign in Ceylon, we are struck by the crudeness of the great majority of the white quartz flakes of the assembled stone industry. This impression confronted us particularly when we contemplated working through the mass of finds. The majority by far consist of roughly knapped flakes, splinters of stone which, having been thrown away as useless, could never testify to any working activities. From the start this impression was so strong that we continually discussed the possibility, from the first to the last of the finds, as to whether we could be dealing with natural objects, or “geofacts” [Jsifakte] as I call them, in contrast to artefacts. Even after we had found stone flakes in the cave floors, namely in the Galgé cave near Katragam, they were so niggardly and so rarely purposeful in their form, that it took us a long time to claim their artificial nature, as discussed in the first chapter. The massifs of gneiss, which are scattered across the south-eastern lowlands of Ceylon in the form of rounded, domed knolls and mountains, have their surface split through by fissures. Often, huge blocks of stone have broken away from these fissures and rolled off the hilltops, perhaps (in many cases) because earthquakes had vigorously shaken them till they fell on top of each other. The depressions have deeply invasive systems of crevasses — fissures, perhaps also split apart through earthquakes, which immediately filled up with milky white quartz conspicuous to the eye. In addition the quartz had cemented together the blocks of stone. The formation of such a system of crevasses can be clearly seen in the accompanying Figure 6, even if how it came about still involves some guesswork. 4From time to time we saw such cracks full of quartz extending from the rock surface to the sides of the grottos (abris) at the foot of a hill. In these cases it would appear certain that rubble from this [page 24] quartz must have fallen out of the crevasses into the sediment on the cave floor. Therefore we carefully examined the disintegration products of these veins of quartz, Large amounts of this quartz cover the substrate of the lowlands in many places where the substrate consists of weathered gneiss. However, the small fragments into which the quartz disintegrates through ‘weathering consistently reveal sharply angled and polygonal forms. We could never establish cases of this quartz disintegrating into the form of the flaked splinters that the cave floors had yielded in such large amounts from our excavations. In addition, along with the splinters of white quartz we found similar splinters of quartz crystal and chert. Hence, these would have had to be part of the formation of the quartz masses filling the cracks of the rock matrix, but we very rarely came upon this (except in. holes etched into riverbeds). Accordingly we established with certainty that our quartz splinters are not “geofacts” but artefacts, as, shown by the unquestionably worked artificial nature of the rich assemblage from the Nilgala cave. The discussion turns not on the question of which are the real artefacts, but on comparing the revealed stone industry with that of Europe. Plates I to VII, which present a wide-ranging selection of our yield, decidedly convey the impression of the ad hoc status of these examples of worked stone, a nature that persists unmitigated throughout our collection, We recognize them in terms of well-known prehistoric types such as knives, points, scrapers and so on in the following description. However, as with all prehistoric industries, the only definite traces left on the stone tools involve the odd patch of gum from bark-rich trees, and as the plants can be identified from their gum, so we judge the value of the prehistoric stone industry from this relatively sparse testimony of intentional working. The great majority of the stone artefacts, from the most ancient Palaeolithic to the Neolithic, are unused discarded flakes, whose fundamentals as knapped products do not change at all for the majority of the collected stone industry from the Chellean to the ‘Neolithic; for the [page 25] judgment of a stone industry, they are therefore worthless. The used, worked flakes or, as I would call them, “teleomorphs” — an expression I would recommend — must be the leading artefacts for the judgment of the present stone industry. (Fig. 6. Quartz veins in gneiss] On this occasion I would propose replacing the work “industry” with the expression lithics [Lithoglyphic}; even better would be lithology [Lithurgie], a fine Greek word, except that it already has another meaning which naturally should not be changed. I therefore speak of Palacolithic and Neolithic lithics, and also of Chellean, Mousterian, and Magdalenian types and so on. The expression of an industry applies only to the particular places where apparently the purpose of its production had really pertained, as for example at Grand Pressigny. 1 further remark here that the concept of lithics should in no way be confused with life-ways [Ergologie], which include the totality of mental or cerebral living nature, that is, the entire theoretical basis for activity, as was earlier described in unmistakable terms (this work, Volume III, page 375). This [confusion] would be a mistake as life-ways are already identified with physiology or with ethnology. So we now turn to a detailed consideration of the proto-Vedda lithics represented in the plates, before ‘we tur to more general questions. 15First, as to the raw material, this was generally quartz in its various forms of opaque white vein quartz with a typically greasy sheen, then chert coloured red, yellow, brown or black, and finally transparent quartz crystal. True flint, or silex, occurs very rarely on the island, and the proto-Veddas made do with the material available to them, as we had also found with the lithics of the proto-Toala. The material used by the proto-Vedda is found in the rivers in the form of rolled pebbles, and in particular in the case of the Nilgala cave we could show that its rolled pebbles of crystal quartz, white quartz and attractive chert derived from the beds of the nearby Galoya (Patipalar). ‘These circumstances explain why the general appearance of the Ceylon lithics creates an insalubrious impression. Knapping of white quartz, chert and quartz crystal mostly results in poorly formed flakes; only as an exception would a knife-shaped flake or delicate point be obtained through knapping, in contrast to flint, which can form the most beautiful knives from flaking as an outcome of its crystalline structure. When we cast a glance at the stone knives in Plate I, we should therefore conclude that their unstylized form is not the result of a particular absence of style from the finished products. [page 26] The following picture can show this; it presents a knife of white quartz (Fig. 7), placed for comparative purposes next to an example of the silex flakes (Fig. 8) from the collection of Aurignacian lithics from the famous Cro-Magnon cave in Les Eyzies. The difference in form between the two flakes clearly reflects the discrepancy in constitution between these two varieties of silica. The knife of white quartz looks so crudely flaked, and that of silex so elegant, yet both are the product of a similar style of manual production. [Figures 7 and 8 have the caption: Stone knives from the Cro-Magnon cave; Fig. 7 of white quartz, Fig, 8 of flint.] The knife-shaped quartz flakes of Ceylon only exceptionally have the large, clear form of Figure 1 in Plate 1. Very rarely are they detached in such a beautifully flaked form with longitudinal ridges, and most end up in fragments. The illustrated piece is made of translucent white quartz and has a striking platform at its thick base, certainly not beautifully shaped but nonetheless a clear indication of a type of handle. Such clearly shaped striking platforms, which are so characteristic of silex, are created only occasionally on crude quartz. More frequent than that depicted knife bladeare the large flake-shaped flakes shown in Figures 2 to 5. In view of their width, these take the form of scrapers, as with the wide, thick cutting tool in Figure 2. The flakes in Figures 3 to 5 and 7 consist of impure quartz crystal; a crude striking platform, which is not well displayed in the pictures, constitutes the handle-shaped base. The provenance of these flakes varies somewhat. Those shown by Figures 1 and 3 come from the Nilgala cave, and the others from the hills of the central mountains, as explained in the captions to the figures, and so is not a topic that the following text will return to. Knife-shaped flakes of average size are shown in Figures 6 to 26 of Plate 1. Of these, the specimens in Figure 8 and 9 are the only cases of elongated flakes, made respectively of white transparent quartz with a greasy sheen (Figure 8) and of opaque red-brown quartz (Figure 9). The former has many similarities with the piece from the Cro-Magnon cave illustrated in the text figure above. (page 27] As the captions explain, Figure 6 displays an aesthetically pleasing, retouched knife of translucent quartz crystal, Figure 10 a flake fragment of red chert, and Figure 1 a pointed blade of yellow chert; brown chert for the Figure 17 blade fragment, honey-yellow chert for the pointed piece in Figure 18, red-brown chert for the pointed flake in Figure 26, and quartz crystal for the piece in Figure 13 with a striking platform at one end, while the pieces in Figures 12 and 14 to 16 were struck stylishly from white quartz; the captions explain this because the photographs themselves display the surface appearance only imperfectly. Transparent quartz. constitutes the attractive, formed knives in Figures 20 to 22, which have a similar shape to the knives of yellow and red chert shown to the side (Figures 24 16and 25). The pieces in Figures 19 and 23, of which the latter is a particularly delicate piece, were detached from quartz crystal as clear as water. ‘The miniature knives of the third size tier are almost always flakes of quartz crystal, as a glance at Figures 27 to 36 immediately shows. The appeared like ice and glinted particularly strongly when brought out to the light of day from the dusty grey floor of the caves; the most exquisite are flakes of water-clear crystal as shown in Figures 34 to 36. All these miniature knives have a stumpy base and a sharp tip. Most have a clearly thickened bulb of percussion, shaped like a handle, whose formation is assisted by quartz crystal somewhat better than by quartz, The small knives of quartz crystal in Figures 82 to 88 of Plate III are a still smaller version that we shall discuss below. These knife flakes, large and small, are all double-edged. One-edged knives, as we had found in one of the Toala caves (38, Plate 1, Figs 6 and 7), are lacking from the Ceylon lithics. Further, these knife blades could not have served for working hard materials such as wood or bone, but would have been used for butchering game, just as Spencer and Gillen (45, page 654) demonstrated in some observations on the use of Australian stone knives. They wrote “As for the use of these stone instruments, primarily because of the nature of the stone, their physical properties would have been entirely unsuitable for hard materials, for example wood. They could have only served the purpose of cutting flesh, and in fact that was the purpose for making them” [translated back into English from the German translation]. If we again cast a quick look over these knife flakes as a whole, we find not one piece with true secondary working along its margin (the so-called retouch or, as I have translated it (38, page 12), after-improvements) of the type where the edges reveal many short, stout flakes of a pointed and scraper form detached from one face, as is especially typical of Mousterian lithies. The Vedda blades more often have the unaltered form of stone flakes that had been detached in the form of a knife, resembling the majority of the knife flakes [page 28] of the upper Palaeolithic and the Neolithic. I shall return to this fact again. We now tum to the points, which are depicted in Plate I in rows. As with the knives, they can be divided into large, middle-sized and small points. Those largest in all dimensions are so thick that one can dispute as to whether their form had been intended or whether they are unused waste flakes that just happened to have a pointed form, as is the case at least with the point in Fig. 37. However, those in Figures 38 and 39 clearly show preparatory retouch along their back, and the example in Fig. 40, made of a rough red-brown quartzite, shows a thickened, neck-shaped and rounded base and the slight beginning of wings. Such rough points were probably fastened to heavy spears to kill large game. ‘The following figures depict middle-sized and small points. Those in Figures 42, 43 and 47 show intentional preparation particularly clearly through retouch along their back. The point in Fig. 58 is a round, well-retouched fragment, those in Figures 41 and 44 are clearly points, while the tools i Figures 45, 53 and 55, and to a lesser degree Fig. 66, have their point sharpened through secondary flaking from the dorsum. The fig. 46 point is made of red-brown chert, and the Fig, 57 point from a red chert with a shellac sheen; the points in Figures 48, 52, 56, 60 and 65 are made of quite transparent quartz crystal. While the instrument in Fig. 64 could have served as a knife, some of the pointed knives could have been formed into points — there are all kinds of transitional forms — as shown by the small, delicate points of crystal in Figures 62 and 63, about which we shall leam more under the category of “miniature lances”. The Fig. 62 specimen has its left margin clearly retouched, as shown on the back and tip of Fig. 60, and the unusual point in Fig, 63 shows the back margin reshaped a little like a ‘winged arrowhead. Although it is known that we can talk of arrowheads amongst the finds of the Toala caves, and so the proto-Toala must have had archery, we must hesitate before declaring that the large and small points found in the Vedda caves had any purpose other than spearheads, notwithstanding the several examples whose intentional wings, even if rough, are unmistakably not makeshift. Such cases are illustrated in Figures 67 to 80 in Plate III, with the wings immediately evident on the pieces in 17Figures 67, 69, 70 and 75, and less clearly so on Figures 71, 72 and 74. These winged points all had their base prepared for being roughly bound to shafts. However, incisions to take binding to fix the points to the shaft with bast snares are recognizable on some pieces only in a very crude execution. We see examples of this with the points in Figures 68, 71, 73, 76 (very uncertain), [page 29] 77, 78 (uncertain), 79 and 80; in addition the notch on the Fig. 47 piece in Plate II can perhaps be brought forward as an example of a notch for binding. Most show one of the notches deeper than the other, as for example in Figures 68 and 79, and further it is sometimes the case that one of the notches to take binding is higher than the other, which we see with the points in Figures 68, 70, 71, 72, 73 (here hardly noticeable), 76, 77, 78, 79 and 80. Therefore, in almost all cases, the binding was not at right angles to the shaft but aslant to it. In these designs of shaft attachments to winged, tongued and notched points we can draw one of two inferences. Either they represent the establishment of the arrowhead techniques of the upper Neolithic, or they reflect the haphazard imitation of this higher technology by a group of people at an earlier stage of Stone Age culture. I take the latter to be more probable. The proto-Veddas became aware of more advanced forms of points; the idea of these spread itself and led to a crude, provisional emulation. I interpret these crude, winged and notched points as a Neolithic influence on the Palaeolithic of the proto-Veddas, and leave it as an open question whether these are spearheads or arrowheads. However, I suggest that the former is probable, for I see archery as a Neolithic invention and therefore not possible for the proto-Veddas (but without prejudging the empirical question regarding the cultural levels to which future finds of archery belong). ‘The points converted into scraper and borer forms, in Figures 109 to 119 of Plate IV, present a distinctive appearance which definitely appears intentional. It may be tied up somehow with a particular use, perhaps to inflict a strong internal injury on encountered game. The illustrated examples, which represent our entire collection, are made from quartz crystal or white quartz, with the exception of that shown in Fig. 116 which had been produced from a honey-yellow chert. These curved points have their analogues in the European Upper Palaeolithic. As already described for the knives, so is it also emphasized for the points that to a large measure they involved the unplanned knapping of brittle material (vein quartz, quartz crystal and chert). As with the knives, so the points show a slighter version of the preparation of similar products that occur amongst European lithics. Worked points of quartz crystal are fairly common in both the Palaeolithic and the Neolithic of Europe, but only the Neolithic shows the predominance of crude, unstylistic rejuvenation, albeit in silex, which the proto-Veddas could not surpass. We depict two such points [Text Figures 9 and 10}, derived from the Neolithic long houses of Robenhausen and Moosseedorf, for comparison with those from the Vedda caves. [Figures 9 and 10, on page 30, have the caption: Crystal quartz points from Neolithic long houses; Fig. 9 from Moosseedorf, Fig. 10 from Robenhausen.] Under the category of scrapers we bring forward a further proof of the fact that the European Neolithic as a general rule did not come up with better results using the brittle material of quartz crystal than did the Palaeolithic inhabitants [page 30] of Ceylon. According to a verbal communication from Dr Heierli, there are well-made Neolithic crystal points with wings, respectable index artefacts from the Neolithic efflorescence, but these are very rare exceptions, There are both knives and points of very small dimensions, which we have illustrated in Figures 79 to 85 of Plate III, and which may even be miniatures, of which some are represented in Figures 86 to 108. They are so delicate that they could hardly have been used for butchering game or as spearheads, but possibly as arrows. They hardly represent accidentally knapped flakes as even a superficial contemplation of the figures instructs us. To start with we observe the knife shapes (admittedly of unfortunate dimensions) in Figures 82 to 85, then proceed through the very delicate artefacts in Figures 1886 to 88, and the miniature knives of honey-yellow chert in Figures 92 and 108; the photographs portray them well, making a detailed description unnecessary. A particular issue with the points is that the larger ones of fine structure could be overlooked as unremarkable, but they have obviously been worked. It has already been noted that the points in Figures 79 and 80 must have served as spearheads or arrowheads, when we referred to the notches used for binding these points. The others illustrated on that plate speak for themselves. Special examples include the very carefully worked point of red chert (Figure 91), the delicate example of quartz crystal (Figure 97), the examples from red and yellow chert that resembles jasper (Figures 98, 99 and 107), and the white quartz flake (Figure 103). The remainder are mostly made of crystalline quartz. There are also rarer examples of miniature cross knives worked in this delicate manner, but with the cutting edge shown towards the bottom (Figures 89, 96 and 104). The special characteristic of these is that the thickened, bowed back has been carefully shaped with the tiniest flaking scars (not well shown by the photographic representations of this glassy pieces), a form of retouch very similar to that on the fine knives and points of the so-called Tardenoisian, and evidently a difficult task on a material as brittle as quartz crystal, At least in our collection these designed cross knives are all of crystalline quartz, whereas the microliths known from the Tardenoisian are made of silex. So it may now be proposed as a first take that these delicate points represent arrowheads, perhaps for hunting birds or flying foxes, for which purpose the similar [page 31], undoubtedly intentionally worked arrowheads of the proto-Toala (38, pages 15 and Figures 25 to 27) had served. However, the following interpretation of the proto-Vedda miniature points is also possible. In his treatise on the Andamanese, Man (24, page 379, notes) wrote the following [translated back into English from the German translation). “As evidence of stone working they also use flakes and chips for cutting hair, tattooing and scarification. They call these quartz teeth. These chips are either fragments of opaque whitish vein quartz or opaque crystalline quartz, or, if they obtain opaque, bluish white quartz pebbles with a greasy sheen, thin translucent chips from their edges.” This category of Andamanese “quartz teeth” fits perfectly with the Ceylon microliths of quartz and crystalline quartz, and despite its meagre size the illustration (23, Plate 12, Fig. 61) shows a form of point somewhat similar to our Figures 93 and 99. In the caption to that illustration it is further stated (ibid, page 462): “These white quartz chips were earlier used for cutting hair and tattooing when they had no other available material. Today these operations are performed with glass”. He further adds (24, page 380): “These chips were not used more than once, indeed most were worn out with each operation; those with a sharp blade were used for cutting hair, while others with a fine point were required for tattooing and scarification”. “Moreover (24, page 331) they were hand held, without being attached to a handle, with the unfastened chip held between thumb and index finger. After use they were disposed of in a midden, or taken far away, so that no-one would be injured through carelessly treading on them. The manufacture of these blades was an obligation of the women, who performed all scarification and hair cutting; the difficult task of tattooing was however handed over to their husbands.” [Translations back into English from the German translations,] Jagor speaks very similarly (16, pages 47 and 50): “Many women were occupied with cutting the hair of the children or themselves, and making other small cuts in the skin. Many individuals had almost their entire body covered with the scars of such cuts, not only for decoration, but also as a remedy against unwholesomeness. The custom appears widespread, I found it amongst various wild people of the Philippines, and in Aden in Somalia I saw the long rows of small scars from the traces of the cuts which they make in the skin with a razor knife in times of illness. The Andamanese turn glass chips or, better yet, glass flakes into these knives; from their trade with Europeans they appear to have glass, not flint or chert, for use”. 19‘When Jagor could find no such flakes in the Andamanese shell middens, despite his concerted efforts, this can perhaps be understood by suggesting they had been discarded in specific rubbish dumps after they could no longer be used, as indicated by Man in the previous quotation, [page 32] From the statements of Jagor (1887) and Man (1878 and 1883) it follows that, before they could avail themselves of glass, the Andamanese previously used chips of quartz, crystalline quartz and chert, some in the form of small knives and some in the form of small points, for surgical purposes, of which those that are now diligently pursued are cutting hair, tattooing and scarification, We may therefore certainly suspect that the proto-Veddas’ microlithic knives and points that we found in Ceylon had served similar purposes, for which reason I shall call them miniature lances. The living relics of the Veddas appear to have completely lost these surgical operations; at least all that we observed is exceptional cases of cutting scars, and nothing of hair cutting or tattooing. The poverty of the Vedda life-ways are perhaps based in part on their impoverishment; as they have irrecoverably lost their stone-working techniques, without however acquiring iron-working technology, in as much as they must trade for their odds and sods of iron tools from the Sinhalese, so they could well have given up on certain customs such as the surgery mentioned previously. Similar miniature knife-lances to those of the Andamanese and the proto-Veddas were also used by Australians in following their custom of making deep cuts in the skin. Brough Smyth (44, I, page 381) illustrates examples of very similar appearance to those in Ceylon, only somewhat larger and made from thick basalt, and the same is true of the stone knives whose use is described as follows by Spencer and Gillen (45, page 654) [translated back into English from the German translation. “The very numerous scars one sees on the body of most of the Aborigines were produced through incisions with stone knives, and during a burial ceremony it is the duty of every man related to the deceased to slash his thighs. These incisions, called ceremonial incisions, were constantly executed, and additionally there were the customs of circumcision and subincision for which stone knives were used.” Those authors admittedly do not distinguish miniature lance-knives for surgical purposes from large knife blades, but in a collection of stone tools from Australia I have observed an array of microlithic knives and points, which could well have been specially designed for surgical operations of the type described here, although naturally their use could have occasionally been replaced with light cuts from large blades. Tf one believes that the proto-Veddas really had archery, then the miniature knives in Figures 89, 96 and 104 could have served as cross-piece arrowheads, as we demonstrate here (Figures 1] and 12) with two such comparable pieces from the work of John Evans (9, pages 369 and 409). [Figures 11 and 12 have the caption: Cross-piece arrowheads, after John Evans.] [page 33] We now proceed to considering examples of knapping which still include trimmed points as part of their form but which, on closer inspection, appear to indicate carefully prepared boring instruments. Figures 120 to 127 in Plate IV depict the perfect burins that we have found; these show with particular clarity their intentional preparation, although it is also certain that other stone flakes would have been used for boring. The best examples are Figures 121, 122, 126 and 127 which reveal preparation of the flake through the elongated corners of the boring point. The last one reveals the most careful retouch of the mentioned pieces; it is entirely triangular, albeit not inverted back as found on particularly good European pieces; such preparation would have been prevented by the stone material, The point of the burins of quartz crystal (Figures 122, 126 and 127) is hard and firm, while that of the white quartz piece (Fig. 121) is slim, first-class and prepared with great effort. The Fig. 126 piece is made of a marvellous crystal as clear as ice. On the other hand the Fig. 123 burin, of ruddy granular chert, has no real reshaping; this flake had originally been detached in this form, and therefore cannot be unambiguously declared to have been a burin. 20Burins whose anterior edge terminates in the boring point are shown in Figures 120, 124 and 125. The Fig. 124 piece, of red chert, is particularly interesting, for the retroversion of the termination of the point had been achieved through retouch; the reshaping has been performed as though to invert the point back with elongated flakes perpendicular to it. To illustrate the retroversion of this stone even better, we provide an illustration from the side of this unique piece (Fig. 13). [Fig. 13. The borer from Figure 124 in Plate IV from the side, to show the spiral axis of the piece] ‘The pieces in Figures 120 and 125 show quite short and tough bore tips with clearly intentional preparation, as can be well recognized on Fig. 120 notwithstanding the strongly glinting surface of this instrament made from crystal; the Fig. 125 piece, made of white quartz, evinces particularly clearly the careful preparation of the bore tip. Perfect borers are well known from all cultural levels of the Upper Palaeolithic, but they are rare. Their existence is also claimed for some earlier times; they are truly frequent amongst the yield of aeoliths in the early Pleistocene and the Tertiary as far back as the middle Oligocene, at which astonishingly early period a lively tradition of making borers had been established (36, page 28). We now turn to the scrapers. Like the burins, these are also rare in comparison to their occurrence amongst European lithics, where they [page 34] play an extremely important role. However, they compare well with our yield from the Toala caves, where perfect scrapers occur equally rarely. The reason for the scarcity of scrapers in the prehistory of places with a warmer climate is suggested by Hamy (14, page 387) to be that the scrapers were particularly important for people in cold climates in working with pelts, a craft that is lacking amongst the natural inhabitants of the tropics and sub-tropics. This would explain the rarity of scrapers which we discovered in the Celebes and now in Ceylon, and which Hamy described from the cave near Konakry in French Guinea as “mal défini” [poorly defined] compared to their European counterparts. Naturally, the poor quality of the [stone] material is also relevant to our [artefacts], as we shall consider later on. ‘We now address the pieces illustrated in Figures 128 to 137 in Plate IV, which are oriented in such a way that the scraping edge is directed towards the top of the page. The first prominent characteristic is that the entire body of the objective piece shows careful preparation through skilfully executed flaking movements. However, the working edge of most of the pieces has the same form as when the flake was detached, and is not buttressed or supported through reshaping as in Europe, where invasive retouching of the scrapers during the Upper Palaeolithic and the Neolithic represents an echo of the Mousterian (one of the hangovers from that older and more primitive cultural period into later periods). The scraping edge of the Ceylon scraper is accordingly sharp; the solid handle or body of the instrument is however fully characteristic. Most are made of white quartz, the Fig. 132 piece from transparent crystalline quartz, and the Fig, 135 example from brown-yellow chert. The crystal scraper in Fig. 137 shows careful preparation through reshaping the working edge; three skilful flakes have produced two backs, and the upper part of the scraping edge is carefully rounded. Fig. 130 portrays a scraper, of quartz crystal, whose transverse axis (running parallel to the scraping edge) is longer than its longitudinal axis. The piece is fully worked, with skilful rounding of the upper margin towards the scraping edge, and any doubt as to whether the piece is perhaps a point or a knife instead of a scraper can hardly be entertained. Several scrapers have the attribute of a crescent-shaped indentation along one of the longer margins, so that they could have served equally well as typical scrapers or as concave scrapers, corresponding very well to European finds of the same variety. Most of the illustrated examples (Figures 138 to 147 in Plate V) need no further explanation; the indented margin of the Fig. 146 crystal scraper is particularly carefully worked. With the Fig. 144 piece the indentation occurs near the pointed anterior end, with the left scraping edge running straight to the tip, and with the base thickened. 21‘As would never be doubted, these concave scrapers served for smoothing the shafts of spears and arrows, assuming in the latter case that archery was practised, which as [page 35] already indicated remains an open question for the proto-Vedda. To be sure, the indentations of the presented scrapers also have widened flanks as would be necessary for arrow shafts; however, this sheds no light on that question. As already indicated for the knife blades and the points, the same must be said for the concave scrapers, that the considerable difference in their finish compared to their European counterparts must be ascribed to the brittle material and not to skill deficiencies amongst the proto-Veddas. To show this we illustrate a scraper, made of quartz crystal, from the Neolithic site of Moosseedorf (Fig. 14), which is distinctly similar to the described Vedda scrapers and concave scrapers. [Fig. 14 has the caption: Scraper of quartz crystal from Mooseedorf.] We now come to a consideration of an idiosyncratic form, which sometimes shows working on one margin and sometimes along all the margins. The latter examples appear like broken slugs, for which reason we shall denote these pieces as slugs [Schuppen]. They are illustrated in Figures 148 to 158 of Plate V with those worked on one side in Figures 148 to 153, and those worked on all sides in Figures 154 to 158. We first address those worked on one side: on any one slug, as shown by the typical example in Fig. 151, the top edge in the illustration is the blunted striking platform from the core, while the edge opposite the striking platform (the bottom edge as illustrated) shows two barbs. We also recognize two barbs in Fig, 149, while in Fig. 152 the barbs jut out from both ends of the straight cutting margin. In Fig. 150, both ends of the cutting edge are rounded. The slugs in Figures 148 and 153 have a single barb. These slugs, which we can call barbed slugs in contrast to the round slugs which follow, prompt the thought that they must have served a particular purpose, and it is entirely possible that they had served as side barbs on spears, as still occasionally practised by Australians, We illustrate a case of such an Australian spear from Brough Smyth (44, 1, page 304) (Fig. 15). [Fig. 15 has the caption: The point of an Australian spear with inset barbs, after Brough Smyth.] This author says this about them (ibid.): “The mongile is a double-sided spear. The Aborigines use a piece of quartz to cut a slot in both sides of the end near the tip, and insert small flakes of hard basalt or another suitable stone, fastened with a pitch-like resin” [translated back into English from the German translation). [page 36] We have found stone slugs of similar form, but with a distinctive serration, in the Toala caves and for certain reasons called them club barbs (38, pages 16 and 17); they could perhaps also at least served this use amongst the proto-Vedda, even though their living descendants do not have clubs as the Toala do. On the other hand the rather large number of serrated slugs worked along one side by the proto-Toala suggests that these could have served as mounts for mongiles or spears (as well as club barbs) even though these weapons are no longer found today amongst either the Veddas or the Toalas. The round slugs illustrated in Figures 154 to 158 differ from the barbed slugs in that the original thickness of the striking platform has not been retained, but instead has been reshaped to make a cutting edge, so forming a broken slug. These stone slugs are conspicuous for their particularly careful finish, and prompt the thought that, even though they could have served as spear mounts, this idea would not be satisfactory. The round slug in Fig. 156 and the beautifully retouched piece in Fig. 158 are made from quartz crystal, the carefully worked piece in Fig. 155 from white quartz, and the daintily finished pieces in Figures 154 and 157 from brown-yellow chert. They have the shape of delicate discs; but they could not be interpreted as sling stones (which we usually associate with discs) and their real purpose is unknown. 22‘We now tum to a particular form which, as Figures 159 to 168 show, has the form of a double cone, as ‘we shall accordingly call them, to use a neutral expression. They reveal many small scars that would have required considerable effort and indeed bother on the part of the maker, to give these stones (in the form of a double cone) a sharp surrounding edge. The more lightly worked versions of this form are illustrated in Figures 159 and 160, 162 and 163, 164 and 165, and 167 and 168, showing the same stone both from the surface and from the side; the first is made of quartz crystal, and the others from white quartz. The greasy sheen of these last pieces does not allow their careful preparation to be shown with particular clarity, especially on the upper surface of the pictures. The stone in Fig. 172 is also a double cone, but it can be identified as a core and transferred to that category. The described double cones were probably created originally as sling stones, to be propelled either with a sling or by a freely swinging hand. This appears very likely if we accept that a skilled hand is known to impart a rapid rotation to a sharply edged stone in a sling, making it a dangerous missile. The impressive capacity of stones of this kind to cut when rapidly rotating in flight has been observed on visits to the Alps, where sling stones rushing through the air cut right through a party of mountaineers. Further, itis reported that the Australians [page 37] occasionally slew kangaroos with a stone slung by hand; so writes Basedow (4, page 24): “A good proportion of the chase is accomplished with the simple help of stones and sticks, which they use as catapults for the smaller game animals”, and Klaatsch (20, page 666) saw an Aborigine from northwest Australia slay a kangaroo by throwing a stone. These double cones lead on to the pointed cores, of which Fig. 171 in Plate VI is a particularly typical example. The artifice whereby the lower surface runs to a conical tip, making them all pointed undemeath to a greater or lesser degree, is not really the essential feature distinguishing them from the usual cores of so-called artichoke form. Rather, the upper surface of the described stones retains the unaltered natural surface, even where it appears retouched, as is the case with the specimens in Figures 166 and 172. These pieces are reminiscent of the European Solutrean pieces described as pointed cores, which as a rule are understood to be core scrapers (compare for example the pieces illustrated by von Piette, 32, page 537, with ours). However, I cannot rightly believe in this interpretation, as it would imply handle-like extensions close to these illustrated scrapers, whereas these scraper cores lack an end to fit in a handle; moreover, the conical tip must have hindered the use of these pieces as scrapers. The use of these stones is unknown, but perhaps they also served as sling stones. It would be wrong to see the occurrence of this Solutrean type as denoting our Ceylon lithics as Solutrean, a point that I shall return to later. The true nuclei show on their upper surface only the original upper surface of the stone pieces detached from the margin of the core. Particularly typical examples are represented in Figures 169, 170, 173, 181 and 183. Some also readily show scars on their upper surface from where flakes had been detached all around in an unsystematic manner, particularly the Fig. 177 piece of whitish yellow chert, and the smaller pieces, of quartz crystal, in Figures 174, 176 and 178 to 180. These cores were not however then used anew as tools, as is the case with the pointed cores, although the irregularly fracturing material provided a favourable edge to guide the knapping of tools from the core. True artichoke cores, which dominate the cores of the European Palaeolithic where use of flint was the order of the day, and which characterize the Neolithic technical efflorescence, were very rarely produced, with their scarcity due to the poor flaking quality of the white quartz and crystalline quartz material. Perhaps also the technique for the removal of [bending] flakes with hard stone splinters, fastened to bone or wood, through skilful, lightly rotating taps, “as though the flakes had been cut with a knife from beetroot”, is [page 38] a post-Palacolithic invention. Flaking with hammer stones is known to have been the original method, and its products depart visually from the beautiful results of grand, regular flaking, In Australia both these methods are applied, including their combination to produce the serrated spear points in northwest Australia (2 and 20, page 667). In addition both methods were applied one after the 23other during the European Neolithic, as I suspect on various grounds. Our proto-Veddas had knapped the stone flakes directly from the core with hammer stones, as we shall see. We have already had reason to make comparisons with the stone techniques of the Andamanese, and do so here again. According to Man (24, page 380) they obtained their fine stone flakes for surgical operations through the following method; first they heated a piece of white quartz serving as the core, then let it cool down, and knocked off the flakes with another quartz pebble. Our Vedda cores do not give the impression of having been heated to any great temperature. They show not the slightest difference from naturally occurring stone. However, glowing flint takes on a chalky dull appearance and can easily fragment into small angular particles, as my experiments have shown me. During our excavation in the Nilgala cave we constantly ran into pebbles, ranging from very large to those the size of potatoes, which at first we paid no heed to. However, when we washed one clean of its covering of grey ash, we noticed that at places it had a granular surface instead of the original patina. Then we knew that these round pebbles were hammer stones (percussion stones) that were the tools the proto-Veddas had used to detach their stone flakes from their cores. Figures 189 to 199 in Plate VIL show these small percussion stones, of which we found approximately 60, lined up in rows. Alll these stones are water-rolled river pebbles, and almost all are made of white quartz, many with a red or yellowish patina, as can be clearly seen in the places where they do not bear pit marks. Percussion stones of hard grey granite or grey quartzite cropped up as the exceptions, and they too were river pebbles collected from stream beds. The pockmarked area left from the knapping activities is occasionally found at one end of a long or egg-shaped stone, as in Figure 195. On this piece the percussion work covers the whole top; the flaking surface contains as it were a scar smorgasbord, and (based on this) we shall speak of it as the pockmarked flaking scar. As this small percussion stone shows us, small hands were used to produce the stone tools, {page 39] Somewhat larger is the elongated stone in Fig. 194, whose white flaking scars can be clearly seen in the picture against the backdrop of the red patina. A smaller version, with pockmarked flaking scars at both ends of an elongated percussion stone, is illustrated in Figures 191 and 198, The hammer stones with a more rounded or discoid shape mostly show the flaking scar in one part, but now and then around the entire periphery. The former case is illustrated in Figures 190 and 197, and the latter in Figures 196 and 199, The Fig. 193 piece, which we had found on a hilltop at Bandarawela, is a particularly important piece, a disc-shaped hammer stone exactly like those from the Nilgala cave, and for its part the perfect proof of the status of the hilltop flakes as artefacts, The majority of the described hammer stones show the cracks that our work had originally described as. pick marks [from the excavation]; however, they occur more frequently than would have chanced under those circumstances, because these hard pebbles lying in the soft cultural layer would have only very occasionally been hit by a pick. Furthermore these same cracks occur on the stone from the Bandarawela hilltop (Fig. 193), on the rear side (which is not visible in the illustration). Without a doubt these cracks would have been created through use, from occasional, very heavy pounding action. Ina few cases the stone had even split in two. Very similar percussion stones of such a small kind must be used by the Andamanese until today, as the following passages from the literature demonstrate with certainty, Jagor (16, page 47) wrote: “Instead of knives they use small glass chips or even better glass flakes, which they break off old bottle bases with a round stone shaped like a small potato”. In the above cited passages, Man said that the 24people used two pieces of white quartz to create small knives and points, by hitting one (undoubtedly corresponding to the percussion stone mentioned by Jagor) against the other. However, these authors have not provided a more detailed description of these small percussion stones, apart from noticing their indubitable [page 40] flaking scars. There are also such percussion stones found by Jagor (16, page 43) in the kitchen middens of the Great Andaman Island, although he did not correctly identify their role as tools. He wrote only that, amongst the other stones he found, “these appear to have been used for breaking up mussels and splintering bone”, Along with the other accounts of the named authors, this suggests that a revision of the Andamanese kitchen middens would he desirable. Under the circumstances whereby the Andamanese, who are small people, correspondingly used hammer stones merely the size of small potatoes, it would follow that the Ceylon hammer stones would have the size and form corresponding to the tools of small people, particularly as the entire body of Ceylon lithics, as already noted, is characterized by small dimensions. There still remains the task to address a couple of stones which are idiosyncratic both for their size and their form. The first to consider is the disc-shaped stone in Fig. 187 (Plate V1). It exhibits a flat front and a bulbous back and is carefully knapped all around, but not above or below. There is a second, somewhat smaller version, also made of brown-yellow chert; in addition a still smaller piece of quartz crystal can be referred here. We may call these stones discs, and for comparison we illustrate a very similar piece ~ whose form and size visibly stamp it as a stone implement — from the collection of stone tools from the renowned cave of Le Moustier (Fig. 16). [Fig. 16. Discoid stone from the Le Moustier cave] It is important to note that all these stones have a slightly oblong form and thickened base (shown to the right in the figure), whereas the upper margin next to it appears sharpened, even denticulated, through the detachment of slug-shaped flakes. We can see a very similar piece found by Pietty (32, page 538) in the Brassempouy cave, whose lateral view also illustrates this thickening; he assigned this layer to the Solutrean, Perhaps they represent a kind of fist-driven wedge, or frechand axe; however, their purpose is quite uncertain. The so-called discs from the Chellean and Acheulean have a more rounded form. Another mystery artefact is illustrated in Figures 186 and 188 of Plate VI, [page 41] apparently made of ‘gneiss, and cut into an ellipse, of which the Fig. 188 piece is particularly clear (although both are unambiguous). One of these small elliptical slabs (Fig. 186) is 23 mm thick, and the other (Fig. 188) 12.5 mm thick. I believe that an analogue to these small, strange elliptical discs of stone has been found by us in the cultural layer of a cave in Kaltbrunnental, near Basel, from the Reindeer Period. It is clearly a small, hewn slab of sandstone, admittedly larger than the Ceylon examples, but with the same recognizable form. Such a piece, unfortunately only a fragment, is illustrated here for comparison. The thickness measures 28.5 mm, and its girth is shown in the picture (Fig. 17). [Fig. 17 has the caption: ‘Mystery stone from a Palaeolithic cave from the Swiss Jura Mountains. Fig. 18, on the same page, shows the side view of the stone of unknown use from the Nilgala cave.] [page 42] The purpose of these stones is a riddle; were they perhaps involved in religious performances, perhaps “soul stones”, as with the churingas made of stone? (Notice such examples in Spencer and Gillen, 45, pages 729 f.; compare also Carnegie, 7, page 20; Brough Smyth, 44, I, page 386, and II, page 399; for similar stones from Tasmania, Art Ling Roth, 35, page 57.) It should be further related that coarse tuberous stones and pieces of chert were found, which had obviously provided the raw material to be knapped for extraction of the smaller pieces used as cores, and finally we must address yet one more unique piece. This is a lugubrious piece with one end blunted across, in the shape of a sliced pointed sausage (Fig. 18). This blunted end is artificial without the slightest doubt, so the stone is an artefact. It has a smooth upper surface but is not polished, No light 25cain be shed on its use; we can not even dismiss the possibility that it might have originated from the Sinhalese layer above and worked its way into the depths, as has happened to some pot shards. However, if we accept that it represents the proto-Vedda life-ways, may I express the suspicion that it had served as a pounding stone to make objects from bark, for which purpose the Veddas of today use their axes, as we recently discovered. The stone fits well in the hand for such a purpose. Following this overview of proto-Vedda lithics, the task presents itself to characterize the assemblage, and to seek out the cultural development (based on the known European finds) to which it can be assigned, For this enterprise we would pay less attention to the finds which are at hand than to the diagnostic artefacts which are missing; crudely worked flakes, knives, points, scrapers, borers and so on are just as well represented in the Neolithic as in the Palaeolithic; indeed, when we consider the pictures on pages 26, 30 and 35 of European, Palaeolithic and Neolithic artefacts made of poorly flaking white quartz and crystalline quartz, we would have to come to the conclusion that the proto- Vedda lithics would have a very flattering appearance if the natives of Ceylon could have availed themselves of good silex. Then it would have been easier to draw a fixed conclusion on the general character of the body of lithics based on the form of the stone implements. ‘What is conspicuous at the first take is the lack of particular tools which distinguishes the Ceylon lithics, and here the most salient point is that a true stone axe is completely lacking (when we exclude this possible purpose for the extremely rare discs). We find no Chellean or Acheulean hand-axes, no Micoquian wedges or Mousterian scrapers, not even the roughly hewn, elliptical axes of the Mesolithic orthe polished axes of the Neolithic stage. The absence of these latter tools weighs even more [page 43] heavily when we remember that axe blades are the most important tools for the Veddas living today, and they use them to carry out the technically most diverse acts, and also as their most important ‘weapon against attacks by animals, as we have shown in our ethnology (this work, III, pages 418 ff.). Admittedly these are made of metal, but we know that the metallic axe blades had developed from their stone counterparts and are nothing but a metallic improvement on their stone predecessors. Accordingly, the metal wedges we observed were bored through or not bored through with equal probability. ‘We have completely failed to find these stone predecessors of the metal axes anywhere, whether searching in the caves or the mountain ridges, or otherwise making inquiries amongst the Sinhalese farmers, priests and antique dealers, or even checking the votive gifts on the temple altars. We had brought a stone axe from Europe and no-one who was shown it understood it, and yet {in Europe] people on the land find stone axes remarkable, and interpret them as supernatural, as thunder stones, and carry them around as lucky talismans. As today’s Veddas do not know how to make for themselves these axe blades that are so important to them, but must obtain them through trading forest produce with the Sinhalese village blacksmith, this stands as virtual proof that they were brought to Ceylon by the Sinhalese during their immigration, and the current situation soon eventuated whereby the Veddas could not do without these new implements. This newly introduced iron axe completely distorted the orientation of their entire lithic industry, and gave them many advantages in the hand, but brought them into dependency on the new immigrants from India, because they could not manufacture these metal instruments but could no longer survive without them. Soon, apparently, the Vedda life-ways were completely changed through disruptions from the Indian culture of the Sinhalese, as can be seen on their shafts used with weapons; the fairly rich lithic industry disappeared irretrievably; the axe was turned to new tasks as never before; in places, the spear met with the higher invention of archery. To understand proto-Vedda life-ways, the lack of ceramics teams up with the lack of stone axes as a noteworthy consideration, a point which requires some further remarks. Never once on the mountain ridges strewn with stone artefacts did we encounter pot shards, even though they are able to withstand 26the atmosphere just as well as stone; so it is certain that the people who had inhabited those mountain and hilltops knew nothing of ceramics. The question is more difficult in the many caves which have occasionally been inhabited by Sinhalese from earlier times to the present, particularly by travellers who have needed overnight shelter. These people have left behind many shards of pottery, which were not always restricted cleanly to their layer in the deposit of those caves where stone artefacts were also found. However, thorough analysis [page 44] leads to the conclusion that these Sinhalese pot shards must have subsequently moved into the depths through the Sinhalese earthworks for their walls, postholes or protecting roofs, and perhaps also through treasure secking, as undertaken in caves by European farmers, Further, we saw burrowing animals at work, such as rats, pangolins and porcupines; sloth bears themselves burrow a camping spot in the cave floors; finally, we found very deep and broad termite nests in the cave deposits, which must have transported pottery shards from the surface to below (see pages 2 and 11). Finally, almost all the recovered shards had a fresh, well-fired appearance, in short the characteristics of today’s Sinhalese ceramics, even though they could be a thousand years old or more. We found no shards of the crude Neolithic type. It perhaps strikes the reader as strange that I should spend so much time discussing this issue when so many excavations in Europe have established a post-depositional mixing of pottery fragments with earlier stone artefacts. The burial of younger materials mixed up with older ones is certainly a common occurrence in caves. However, the new findings in prehistory of the very influential A. Rutot in Belgium, in not less than a dozen Belg caves, have raised the assertion that ceramics had not been discovered during a later period but already by the Palaeolithic — not just the Magdalenian but indeed the Aurignacian — admittedly in the form of well-developed pots with handles, and this raises the question whether it is time to undertake a revision of ceramics which had previously been considered diagnostically Neolithic. As Rutot says (37, page 523), “After 40 years the facts reveal themselves as demonstrating the verity of the question of Palaeolithic pottery; nothing will prevail against the facts”. In the proto-Vedda case we can remain firm that their mountain dwellings contained no ceramics, and that the ceramic shards and stone splinters mixed together in the caves had been brought together through secondary commingling, and in short that the proto-Veddas lacked ceramics, This lack of stone axes and pottery from the Vedda life-ways prohibits their attribution as Neolithic; in accord with this absence, all other indications would place the proto-Vedda life-ways in the Palaeolithic. As previously noted, the lack of Chellean and Acheulean hand axes, and the lack of Micoquian and Mousterian scrapers and points, would imply assignment of the Ceylon lithics to the Upper Palacolithic. Given also the lack of the laurel leaf points and shouldered points of the Solutrean, and that these types had first been definitively established in Europe with the Aurignacian, this assignment would be restricted to the youngest phase [page 45], the Magdalenian. This impression struck us with the first finds, and became stronger with those that followed, so that already in our preliminary reports (40 and 41) we had announced that the Stone Age of the Veddas had a Magdalenian character, which is so strongly stamped on certain materials as to indicate a Vedda facies. The life-ways of the proto-Toalas discovered by us have also already been compared to the European finds of the successive cultural stages. The lack of every stone axe, whether polished or crudely flaked, the further lack of ceramics (the possession of which cannot be demonstrated by a few shattered pieces, unless undisturbed pottery fragments can be found as well), and other secondary grounds (38, page 23) allowed us to characterize the Toala lithics as Magdalenian with a Neolithic influence. Their mixed appearance, whereby we could recognize a Neolithic influence from a Neolithic people living near the Palaeolithic proto-Toalas, derives from the existence of true, serrated or winged arrowheads. This view preserves the understanding that archery was a Neolithic invention, The peculiar mixing of Palaeolithic and Neolithic attributes prevented us from tracing a direct connection between our proto-Toala lithies and a European cultural stage, which is why we called it the Toalean, a name that I have already ejected as incorrect, for G. de Mortillet wanted to use such names with “an” endings to denote the cultural passages through which all humanity must proceed at one time, whereas the “Toalean” is only 27a local expression, characterized by the admixture of two such cultural passages. It is accordingly eniough to speak of a Mesolithic Toala facies, while our proto-Vedda lithics reflect an Upper Palaeolithic facies. So far it has been possible to characterize such exotic lithics satisfactorily in terms of their European foundations, and one would want to undertake this systematically for other, indeed still living lithic traditions. One could for example sustain the proposition that a hundred years ago, when Tasmanians were still living, a traveller from this island across Australia to New Guinea would have encountered three European cultural stages, the earlier Palacolithic of the Mousterian in Tasmania, the Mesolithic in Australia (albeit at the cusp of developing characteristic stone axes), and in New Guinea the Neolithic (or Robenhausian) as demonstrated through the polished stone axes there. However, one serious objection has been raised against such a framework, namely that a system for dividing up the Palaeolithic finds of Europe, and in particular France, is heavily challenged in terms of its scientific value. On this point we shall not procrastinate in saying that [page 46] the schematic, rigid system of G. de Mortillet is not the correct representation of the facts, and that here, as everywhere in nature, all things influence each other and the divisions are all based on more or less artificial foundations, a point that has already been made often enough (see for example Verneau’s notes over his finds from caves in the Canaries, 47, page 652; Hamy over his finds in the cave at Konakry, 14; further, Hérnes’ experiences in Italy, 15, and those of Knowles in Ireland, 22, etc.). However, we must concen ourselves over the critique that would discard the French system of divisions in its entirety, as this would be a serious mistake for science. We particularly focus on the views of the esteemed anthropologist and explorer H. Klaatsch, who has attained a comprehensive overview on the related questions through his extensive travels. In view of his numerous, moderate opinions, and those of Rutot, we do not share the view of the aggressive prehistorian Obermaier (31, page 5), for whom Klaatsch’s comments from a short study tour of France have no value; in our view there has been no greater contribution to the knowledge of prehistory than the results from studies by a truly great man, In his report from his study tour to France, which he had set himself as a highest objective, on the topic of the reality of the system laid out by Gabriel de Mortillet on the divisions of the Palaeolithic (for this, see 26), amongst other things he said (17, pages 114 and 115): “I have reached the position with clarity that Mortillet’s classification is not feasible and that it was a mistake to identify the types of flint knives as the classification basis for periods. I have recognized that the nomination of'a Mousterian type of flint artefact is an unfortunate step and the resulting construction of a Mousterian period a misapprehension. The construction of the Solutrean period from the knives of laurel leaf form had already appeared to me very problematic. I first reached recognition of the true meaning of these facts from Rutot's collections.” In his highly popular work (18, pages 233 and 310) he stated: “The search for a chronological classification from the varieties of silex forms was in vain. In the Neolithic we again find Solutrean silex types, along with Magdalenian blades; however, more primitive methods of ‘working stone also came to the fore in later periods.” In the report over his Australian tour we read (20, age 667): “To apply the unfortunate French nomenclature to any presence of the Palaeolithic would naturally be the most highly amiss enterprise. In Australia and Tasmania “aeoliths”, albeit of glass, still Cecur, so one has them all together, these technical forms alleged to represent earlier dates, and they show up the efforts to use the form of artefacts for the classification of periods as being in vain.” In the review of a [page 47] French treatise (Zeitshcrift fir Ethnologie, 39 [this should be 17}, 1907, page 765) a direct reference was made to “G. Mortillet’s antiquated system of Palaeolithic periods”. From these remarks one might think that Mortillet’s system was a completely useless, unfortunate business in Klaatsch’s eyes, but in other places we read with wonder that this is not the view of our author and that overall he agrees with it, for elsewhere he writes the following (18, page 216): “Gabriel de Mortillet uttered the right idea, that the technology can offer a direct clue to the passage of the Old Stone Age, just as the progress towards completion of tools altogether serves the division of prehistory into the Old Stone Age, New Stone Age, Bronze Age and Iron Age. It will always be a surviving 28legacy of Gabriel de Mortillet that he was the first to bring a classification to the chaos of the Old Stone ‘Age. Without a doubt the Old Stone Age experienced a developmental trajectory in technology; the Chellean knives belong to a much earlier period than the fine blades of Magdalenian type. The Solutrean finds can be understood as a local efflorescence in stone technology. The concept developed by Mortillct of the Mousterian immediately impinged on a tight circle of ideas for everyone familiar with the Old Stone Age; he knew that by announcing this name, it would implicate a late deluvian layer orsite. Equally, the word Magdalenian linked itself to the unravelling of a quite tight circle of ideas. By placing together these remarks, which would appear to have been advanced by two quite different authors, it is enough to show that our well-travelled prehistorian finds himself swaying helplessly on the judgment of the Mortillet system, that he spoke out on the worthless nature of the system but that, at a deeper level, he himself swum with its current. The glorification of the fact that certain types are the proof of each and every period is, as already emphasized, not new and remains firm; however, to characterize these as the pillars of the Stone Age approaches on overemphasis on certain types, as for example using one-sided retouch on Mousterian stones for the Mousterian, or laurel leaf points, shouldered points and pointed cores for the Solutrean, or the presence of unpolished stone axes (frequently of almond form) for the Chellean, or finer versions of these for the Acheulean, or polished artefacts for the Neolithic, or a lack of any stone axes along with the appearance of a rich bone industry (compared to the absence or scarcity of bone artefacts in earlier times) for the later Palaeolithic, or the appearance of certain bone implements and plastic arts, or the appearance of ceramics as a new phenomenon compared to older cultural materials, etc. These difficulties in the division of prehistoric life-ways should also prepare us for the probability that the carriers of a deeper cultural epoch, in places difficult to access, must have come into contact with immigrant bearers of a higher cultural period, and that the former would have adopted the more advanced tools and so [page 49] assigned a low value to their own tools. Further, this migration of cultural materials occurred perhaps not merely locally but also generally and throughout time, so that a more primitive stage was enriched at length with technically higher artefacts, until after a sufficient time it could be classified in its totality with the higher life-way. Such mixtures of transitional materials are perhaps more common than pure cultural assemblages. When Hémes (15, page 190) joins Klaatsch in saying that the claim for the Mousterian is untenable, along with Obermaier (31, page 36) I can ever so little follow this view, at least after a careful study of it which I have managed to do from time to time. The salience of Mousterian points and scrapers with retouch along one margin is so clearly characteristic of this period, that the stones from the sites of Le Moustier, La Quina, Baumanns Cave (15, page 24), Taubach (15, pages 23 and 48, pages 643 and 644), Wirdkirchli (1, page 347), South Africa (11), and Tasmania are difficult to distinguish from each other. For instance, the pieces from La Quina and Tasmania in the possession of the Basel Museum would be curated together except for the saving fact that the catalogue numbers written on them allow them to be told apart, as the technology used to make them is in such strong agreement (on different, but similarly patinated materials). When Klaatsch further reported that inspection of Rutot’s collections had brought him to a realization of the lack of utility of Mortillet’s system, we should remember that Rutot not only acknowledged Mortillet’s cultural stages and adopted his names, but also approximately doubled their number (on this, see Homes, 15, page 189). ‘We therefore accept it as scientific to look for a link between all exotic and European Stone Age cultures, especially the French research results, and eventually to integrate or to enrich the latter with the results of the finds from places outside of Europe. For that reason we investigated the life-ways of the proto-Toala, and sought a similar connection for the proto-Vedda, and believe we have achieved this in the way described here. Accordingly I may talk of Tasmanian lithics as Mousterian, Australian 29lithics as Mesolithic and Papuan lithics as Neolithic, the last of these not being doubted by anyone; why then drive scepticism about the other stages to such extremes? ‘The Chellean has not yet been found in Ceylon, but is known from several places in India, in the northwest of Madras and northwards as far as Godawery, then in Orissa, the Central Pro , southeast Bengal, Assam and still other places; it is apparent in the rough quartzite material, despite its very crude formation. These stones were first discovered by Bruce Foote, King and Oldham in 1865 (for this, see Medlicott and Blanford, 25, who also cite the literatures). That these finds represent an ‘Asiatic Chellean, and there is also an African version, is not doubted by anyone, and thus this mute witness further reveals the aptness of Mortillet’s Chellean stage. [page 49] Further there is no lack of an indication that the Magdalenian also occurs in India, in particular through flakes and chips of vein quartz, rather similar to those from Ceylon (25, pages 440 ff; 10). In addition India has the mysterious Tardenoisian, with the pygmy flints found by the English in the caves of the Vindhaya mountains (for a new view, and comparisons with European finds, see 12, page 18). Polished axes are completely lacking in India, a circumstance which appears as mysterious as the lack (till now) of Chellean axes in Ceylon. At this point we may report a very interesting observation made by Mr E.E. Green in Ceylon, which we produce here in translation [here translated back into English] from his words published in an inaccessible magazine (13, page 163). “An interesting custom of the Tamil coolies, who work in the plantations of Ceylon, is their production of razor blades from the thick bases of English beer bottles. They need these glass flakes primarily to remove the hair from their heads, or, as we may say, to shave their heads, These thin flakes are struck from right round the bottle base, leaving it looking like a flint core. This Tamil custom perhaps derives from the Stone Age, and before the introduction of glass, stone or mussels from the surroundings must have been used. This custom is mainly limited to the pariahs, the poorest of whom use only glass chips. I have been advised that this custom is not restricted to the mentioned districts, but occurs in all of the south-eastern villages from which the coolies hail.” This observation establishes that prehistoric lithics survive in India amongst the lower castes in the form of glass flakes used for surgical purposes, as we have observed for the Andamanese. Mr Green ‘was so kind as to honour us with some of these glass flakes, and as he himself produced no illustrations, so we here present two of them in a photograph (Fig. 19) to show that they are fully in accord with flakes from flint. [Fig. 19. Glass flakes made by Indian coolies] ‘We also add here that in Ceylon we have observed ceramics from India, as shown by their technical features including the very frequent decoration of fingertip impressions. Further, when the fingertips were pressed in they had not been wrapped in moistened cloth, so the decoration emerges quite evenly, as is known to be a feature of Neolithic pottery (Fig. 20). I should name phenomena such as these glass knives and finger impressions [page 50] prehistoric relies; the latter also occurs in Europe till modem times, albeit rarely. (Fig. 20. Modern Indian pot shard with finger impressions] Many things could be said about Indian pottery in the sense indicated here, but this is not the place to pursue that theme further. While the Chellean axes found in the main part of India belong (like the European examples) to the Pleistocene, Noetling (29, page 233) has discovered flint flakes in Burma from the lower Pliocene. The 30stones illustrated by him could be either artefacts or “geofacts”, but, since the existence of the genus ‘Homo by the Pliocene is probable, I add these words from Boule (5, page 70): “It is necessary to treat all the works of Dr. Noetling’s genre with care”. Otherwise there is Noetling’s discovery of hippopotamus shank (30, page 242, with a precise illustration) from the same lower Pliocene layer, which on the facets of its joint end shows marks, at right angles to each other, albeit partly sloping, that Jook like they had been cut with a knife and do not dip deeply down, so that they could be interpreted as created through friction when removing flesh. Noetling argued that no animal could create such even facets with its bite, so it would seem to be an indubitable proof of human working; these facets would then have been created with stone knives, However, we cannot see an answer to the question as to what use these cuts could have served, and also Noetling had forgotten a group of animals stil living in the river systems today, namely the soft-shelled tortoises or trionychids with their jaws that can cut as, sharp as knives. In fact I believe it is clear that the facets found on the cartilage of the end of the bone are the fissures of a Trionyx; [page 51] Noetling’s find is the bone of a dead hippopotamus gnawed by a riverine, soft-shelled tortoise. Ithas already been indicated above that the Andamanese retain the traces of a Stone Age which shows analogies to the Ceylon Stone Age in its hammer stones and lance-like flakes; also found are “stones that appear similar to crude axes and knives”, along with pot shards and a few polished stone axes, which are instruments from the Neolithic. As already remarked, the Stone Age of the Andamanese is due a thorough revision. In Melaka and the Malay Archipelago the exploration for prehistoric finds in caves has been in vain, with the single exception of our discoveries in the Toala caves of Celebes. In Borneo Everett explored 20 caves “without discoveries of special interest”; this accompanying remark however was made: “some quartz flakes were found, and these could be the product of human hands”, a statement that takes on an illuminating meaning given our finds in Ceylon (49, pages 36 and 38). Still, these negative finds from the caves have shown nothing. Polished stone axes are however widely found strewn across these same stretches of land. In New Guinea, at least in its English part, one gains the impression that the prehistoric Neolithic had surpassed its present development, for morsels of grain were found with rubbing clubs strangely decorated across some of their surface, recalling those from the American Neolithic (3, page 1), as well as “pot shards from a higher ceramic technology than that known today in the region” (33, page 602). [The above three quotations have been translated here back into English from the German translation.] In Australia people find stone tools in prodigious amounts, particularly on the summits of hills, which is reminiscent of our finds in Ceylon (44, I, page 362). Slugs of axes polished only along their edge have been recovered from the ground, and so are a prehistoric Australian characteristic, The hypothesis advanced especially by Schétensack and Klaatsch, that the Australians are the inhabitants from an early age, has not been confirmed through prehistory, in that all the stone tools found so far either lie on the surface or just beneath it. Brough Smyth (4, I, page 364) writes on this [translated back into English from the German translation}. “It is remarkable that no stone axe, basalt flake or stone knife from Victoria has been found anywhere except on the surface or a couple of inches beneath it. Even the caves, which have been explored, show only a very recent existence of the race. ‘When only a small part of Victoria has been built up through alluvial deposition, when the land has been occupied by thousands of miners for 20 years, and the Scots have washed through the dirt from countless excavations as deep as bedrock, it is difficult to dismiss the non-discovery of relics lightly, as has been done previously.” Klaatsch (20, page 666) found stone flakes in the vicinity of Windham to a depth of 24 cm, and stone axes in Queensland at a similar depth (Zeitschrift [page 52] fiir Ethnologie, 37, 1905, page 780). The scientific value of the claimed traces of people belonging to early times, made by some others, will not be discussed here, until they have been confirmed through future research (see the axe marks in pieces of stone 30 feet deep, 44, I, page 365, Remarks; also the human footprints at a 31remarkable depth in old sandstone, 6 and 19, page 782; and finally aeoliths in a Pleistocene diprotodon layer, announced by Klaatsch and Rutot, 36, page 43). Otherwise I believe that a Mousterian-like undercurrent, with Acheulean influence, similar to the Mousterian of Tasmania, can be seen in Australian prehistory; in the Murchison District of Western Australia Tasmanian-like stones, albeit no longer in use, have been observed (46, page 199). In Tasmania, as well, the stone tools lie on the surface. As research until now indicates, one must conclude that the original {niederen] forms of people of Ceylon, the Andamans, the Celebes, Australia and Tasmania have not produced any indication of living in the Pleistocene. Only with the Chellean axes of the main body of India, which have been found in Pleistocene layers just as in Europe, do we willingly accept the existence of age-old people on this continent. Future research remains to undertake most of the work. Nonetheless our finds from Ceylon caves have clearly established that, prior to the invasion of a people from sub-continental India, a primitive group of hunters, the ancestors of the present-day Veddas, had occupied the island. Despite all that we have expressed in favour of our view — that the Veddas in their original pure form are the autochthones of Ceylon, and constitute an ancient branch of humanity in their physical properties (small brain, Australian physiognomy, and to some degree their small stature, inter alia), and that they cannot be confused with the Sinhalese — doubts concerning the correctness of our views will not quieten down. In his otherwise truly interesting travel report over the more or less related people in Sumatra, M. Moszkowski (28, page 234) writes the following. “I would again touch on the question whether the Veddas are not a secondary branch. When amongst the races of middling size, such as the Sakais [Senoi] of Sumatra, one suddenly sees a dwarf appear with similarities to the previously mentioned Veddas, one is prompted to think of their degeneration.” We reply to this: in all human varieties, a dwarf bears the physiognomy of the variety which that person represents; so European dwarfs are nothing but miniature Europeans in their physiognomy, Sinhalese dwarfs are miniature Sinhalese, Malays are miniature Malays, and the corresponding remarks apply to giants. A similarity to Veddas can only occur through the dwarfing of Veddas or groups related to them. [page 53] The physiognomy of the Veddas is not a smaller version of the Sinhalese phsyiognomy, as would have to be the case if the Veddas were dwarfed Sinhalese, but truly distinct, of a more Australian character, in contrast to the fine Sinhalese type. Nor is it a case of atavistic dwarfs or an infantile type, which would be revealed through a childlike physiognomy, a point that we shall not follow further here. ‘As for the reasons for the degeneration Moszkowski implicates illness, especially the skin disease of ringworm [kurap], and inbreeding. As for the former, the so-called Cascado of the Portuguese, itis found amongst all groups, large and small, of the Indo-Australian Archipelago; but we found it extremely rarely amongst the Veddas, only once, being the case of the old man illustrated on page 429 in the third volume of this work. On the question as to whether inbreeding could bring about the body-size reduction of a population, or be harmful overall, much has been written here and there without definitive results. In his treatise on this issue, S. Reinach (34) comes to the conclusion that it is not bad as such, but that it admittedly must lead to an accumulation of physical weaknesses and illnesses, as can be seen in an affected family, though it can also lead to a concentration of any excellent characters already present. In particular, on the Sumatran Sakai investigated by Moszkowski, amongst whom he found dwarfs, he directly said about them (28, page 233) that they enjoy “a large surplus of women”, However, their inbreeding would somehow have had to lead to strong, physical reduction. We repeat that the small size of the Veddas is an original character of this autochthonous population, as is the body mass of these wavy-haired people. The people related to them, haphazardly scattered over a 32‘great distance, can be revealed and distinguished as local versions of the old Vedda form, from which point of view we now add Aboriginal Australians to this original group of wavy-haired Homo sapiens, as a secondarily enlarged and further developed local form. Something can be said here on [another of] our original [populations], the woolly haired group. The ‘Vedas do not belong to this “Papuan race” as Moszkowski (ibid.) calls it. Further, I cannot avoid making some phylogenetic observations. At one time (1892) it was still scientifically valid (unlike now) to look for prehistoric, indeed phylogenetically low characters in comparative anthropology. We then believed we had found, on some of our Vedda skulls, characters that approach those on the anthropoids, the so-called serial characters, as also on other parts of the skeleton. We believed we had constructed a phylogenetic succession in the growth of the posterior skull based on the curves of chimpanzee, Vedda and European skulls laid one on top of the other, hand in hand with the posterior ramus of the mandible [page 54], and we wrote on this in our following work (this work, Volume 3, page 369). “We are therefore of the view that the chimpanzee is the closest living anthropoid to humans, without representing [the common ancestor]. We rather believe that the chimpanzee has distinguished itself through a series of its own characters [autapomorphies].” We therefore said that the genus Homo had had an anthropoid point of origin, and that of the living anthropoids the chimpanzee retained most of these characters, but naturally without representing it by any means. This view was portrayed by Klaatsch (18, page 322) as follows: “A considerable time ago the Sarasins tried to establish the Vedda skeleton as the missing link between chimpanzees and humans”; and later: “The attempt to find direct links between certain anthropoids and the lower human races will have its foundations undermined through the correct, dignified procedure, which is to use fossil bone remains instead”. To start with, we still believe that humanity originally had an anthropoid form whose characters are mostly shown by the currently living chimpanzees. The time is absolutely not yet over when the relationship of humans with anthropoids can be fruitfully studied, as claimed by Klaatsch (18, page 159). Indeed, he says elsewhere (18, page 172): “It is correct that the chimpanzee appears most similar to humans of all the anthropoids, and this implies that it has departed comparatively little from the ‘common ancestral stem”. We further read the following (18, page 181). “This very simple comparison naturally does not touch on the close relationship between anthropoids and humans. Along with Branco, we do not rule out that a fruitful union between humans and a humanoid ape could come about. Friedenthal’s experiment on the possibility of mixing blood as evidence for a relationship between the blood lines is full of theoretical importance and is in complete harmony with our ideas founded on anatomy.” These sentences say the same thing that we have said, so why the published rejection? And as for his sentence about how the search for serial characters amongst the varieties of humans “will have its foundations undermined”, he invokes this himself with his observation (18, page 312): “The Australians are the lowest representative of humankind; they approach every condition that had probably characterized the origins of the human line based on our comparative study of primates”. Just as with his objections against the French divisions of the Old Stone Age, so also with his rejection of humanity’s anthropoid relations, does our critic himself adopt the same arguments that he is outwardly contesting. [page 55] After Schwalbe and Klaatsch, in opposition to earlier anthropologists, had established the fossils skulls of Neanderthal type as serial, we also tried to integrate this view with ours on the serial characters of the original Vedda line, and wrote the following (39, page 135). “We therefore depart from Schwalbe’s view on the derivation of recent people, Homo sapiens, from Homo primigenius and 33subsequently Pithecanthropus erectus. To spell it out, we disagree with Schwalbe that the small-bodied wild human forms, which are the lowest today and which influenced the somewhat larger Australians, would be local varieties of recent humans distinguished on size, with their development influenced through some extremely unfortunate circumstances, but instead are the oldest and most original, still living forms of Homo sapiens.” (See also 38, pages 27 and 42, page 15.) This attempt to find a common ground was also soon tumed down by our critic, Klaatsch, who wrote the following (21, page 140). “The connection previously made by the Sarasins taking the Veddas as {a model for] a chimpanzee-like ancestral form has now been weakened by modem views, for the step ladder proposed by Schwalbe for Pithecanthropus to Neanderthals to recent humans shows the connection made by the Sarasins is too close. The primitive status of the alleged Homo primigenius was much overstated.” ‘We read this passage with astonishment, for the same author shortly wrote the following (18, page 299). “The Neanderthal fossil remains from Spy and Krapina ~ as well as of Homo primigenius ~ belong to a very low step and indeed represent the verging of humankind towards its animal ancestors, ‘One compares the skull cap of the Spy individual and the Neanderthals with that of Pithecanthropus and immediately observes the general similarity between their configurations; the latter appears as a physically reduced version of the old deluvial skull.” Darwin said in his Descent of Man (8, page 156): “We therefore depart far, as far as possible, from the belief that humans had originally derived from catarthines, which had originally appeared in the Eocene, for the higher apes had developed by at least the upper Miocene, as shown by the existence of Dryopithecus” [translated back into English from the German translation]. I cannot correctly understand the force of this argument that the anthropoid ancestor of humans would have first come into existence during the Miocene. Darwin’s sentence would not be consistent with the claim ~ if it can be shown correct — of stone tools found dating to the middle Oligocene (36). Such an existence of the genus Homo by the second stage of the Tertiary [Eogen] ~ and according to Rutot, the Oligocene lithics include a great diversity of tools, as only a human could need and invent - raises the question, if it has any sense, [page 56] of looking for equivalent human arts and their phylogenetic relationships with those of the Pleistocene, and speaking of lower and higher varieties. It appears very unlikely that, during the enormous time period of the Tertiary, until today, the phylogenetically lower and higher could have both maintained themselves. Then the anthropology of living people would have no more value than the description of the spontaneous varieties of other animal species, namely, very little. Above all, a definitive clarification of the problem of the antiquity of stone tools clamours for attention, if we are to understand human phylogeny, and this will not be achieved as long as Rutot is correct in naming his latest announcement of Oligocene humans “a serious problem”. Cited Literature (for the “Lithics” section, completed at the end of March 1908) 1. Bachler, E., Die prihistorische Kulturstitte in der Wildkirchlii — Ebanalphdhle, Verhandlungen der schweizerischen naturforschenden Gesellschaft in St. Gallen, 1906, page 347. 2. Balfour, H., On the method employed by the natives of NW-Australia in the manufacture of glass spear-heads, Man, 1903, page 65. 3. Barton, F.R., Note on stone pestles from British New Guinea, Man, 1908, page 1. 4. Basedow, H., Anthropological notes made on the South Australian Government North West prospecting expedition, 1903. Transactions of the Royal Society of South Australia, 28, 1904. 5. Boule, M., Uber Noetling, L’Anthropologie, 6, 1895, page 70. 6. Branco, W., Die fraglichen fossilen Menschenspuren in Sandsteine von Wamambool, Victoria, und andere angebliche Spuren der fossilen Menschen in Australien, Zeitschrift fir Ethnologie, 37, 1905, page 162. 3410. ue 12. 13. 14. 15. 16. 17. 18, 19. 20. 21. 23. 24, 25. 26. 27. 28. 29. 30. 31. Camegie, D.W., On a bark-bundle of native objects from Western Australia, The Journal of the Anthropological Institute of Great Britain and Ireland, 28, 1899, page 20. Darwin, Ch., The descent of Man, second edition, London, 1888. Evans, J. The ancient stone implements, weapons and ornaments of Great Britain, second edition, London, 1897. Foot, R. Bruce, Notes on prehistoric finds in India, The Journal of the Anthropological Institute of Great Britain and Ireland, 16, 1887, page 70. Frames, M.E. And T. Rupert Jones, On some stone implements found in a cave of Griqualand- East, Cape Colony, The Journal of the Anthropological Institute, 28, 1899, page 251. Gatty, R.A., Pigmy flint implements from the Landbeds at Scunthorpe in Lincolnshire, Man, 1902, page 18. Green, E.E., Tamil habits of shaving with glass chips, The Taprobanian, 1, 1887, page 163. Hamy, E.T., La grotte du Kakimbou a Rotoma, prés Konakry (Guinée Francaise), L’Anthropologie, 12, 1901, page 380. Hoemes, M., Der diluviale Mensch in Europa, Braunschweig, 1903. Jagor, Uber die Andamanesen oder Mincopies, Verhandlungen der Berliner Gesellschaft fiir Anthropologie, Ethnologie und Urgeschichte, 1877, page 41. Klaatsch, H., Anthropologische und paliolithische Ergebnisse einer Studienreise durch Deutschland, Belgien und Frankreich, Zeitschrift fir Ethnologie, 35, 1903, page 92. [page 57] Klaatsch, H. Entstehung und Entwicklung des Menschengeschlechtes, in Weltall and Menschheit, published by H. Kraemer, 2; undated. Reisebericht aus Australien, Zeitschrift fir Ethnologie, 38, 1906, page 776. Schlussbericht tiber meine Reise nach Australien in den Jahren 1904 — 1907, Zeitschrift fir Ethnologie, 39, 1907, page 635. ~- -- Besprechung der Materialen zur Naturgeschichte von Celebes, 5, second volume, Zeitschrift fiir Ethnologie, 40, 1908, page 139. . Knowles, W.J., Irish flint arrow ~ and spear — heads, The Journal of the Anthropological Institute, 33, 1903, page 44, Man, E.H. List of Andamanese and Nicobarese implements, omaments etc, The Journal of the Anthropological Institute, 7, 1878, page 457. -- -- On the aboriginal inhabitants of the Andaman Islands, The Journal of the Anthropological Institute, 12, 1883, pages 69, 117, 327. Medicott, H.B. And Blanford, W.T., A manual of the geology of India, Culcutta, 1879. Montillet, G. and A. de, Le préhistorique, origine et antiquité de homme, 3° edition, Paris, 1900. ~ -- Musée préhistorique, 2" edition, Paris, 1903. Moszkowski, M., Uber zwei nicht-malayische Stimme von Ost-Sumatra, Zeitschrift fiir Ethnologie, 40, 1908, page 229, Noetling, F., On the discovery of chipped flint-flakes in the Pliocene of Burma, Natural Science, 10, 1897, page 233. Note on a worn femur of Hippopotamus irravadicus, Caus. and Falc., from the lower Pliocene of Burma, Records of the Geological Survey of India, 30, 1897, page 242. Obermaier, H., Die Steingerite des franzésischen Altpaldolithikums, Mitteilungen der prihistorischen Kommission des Kais. Akademie der Wissenschaften in Wien, 2, 1908. . Piette, E. and I. de la Porterie, Fouilles 4 Brassenpouy en 1897, L’Anthropologie, 9, 1898, page 531. . Péch, R., Dritter Bericht tiber meine Reise nach Neu-Guinea, Sitzungsberichte der kaiser!. Akademie der Wissenschaften in Wien, mathem.-naturw. Klasse, 115, 1906, page 601. }. Reinach, S., La prohibition de l'inceste et sus origines, L'Anthropologie, 10, 1899, page 59. Roth, H. Ling, The Aborigines of Tasmania, Halifax, 1899. . Rutot, A., Un grave probléme, une industrie humaine devant de l'époque oligocéne, comparison des outils avec ceux des Tasmaniens actuels, Bulletin de la Société Belge de Géologie, 21, 1907. 3537. ~~ La poterie pendant I'époque troglodytique, Bulletin de la Société préhistorique de France, 4, 1907, page 523. 38, Sarasin, P. and F., Materialien ur Naturgeschichte der Insel Celebes, 5. Versuch einer Anthropologie der Insel Celebes, erster Teil: die Todila-Héhlen von Lamontjong, 1905. The same, zweiter Teil, die Varietiiten des Menschen auf Celebes, written by F.S., 1906. 40. ---- Ancient stone implements in Veddah caves, Ceylon Observer, April, 1907. 41. ~~ Die Steinzeit der Weddas, Globus, 91, 1907, page 255. 42. Sarasin, P., Zur Einfiihrung in das priihistorische Kabinett der Sammlung fiir Vélkerkunde im Basler Museum, Basel, 1906. 43. Schoetensack, O., Die Bedeutung Australiens flir die Heranbildung des Menschen aus enitern niederen Form, Verhandlungen des naturhistorisch-medizinischen Vereins zu Heidelberg, 7, 1901, page 105. 44. Smyth, R. Brough, The Aborigines of Victoria, London, 1878. 45. Spencer, B., and Gillen, E.J., The northern tribes of Central Australia, London, 1904. 46. Tylor, E.B., On the survival of Palaeolithic conditions in Tasmania and Australia, The Journal of the Anthropological Institute, 28, 1899, page 199. 47. Verneau, R., Instruments en pierre des iles Canaries, Bulletins de la Société d'Anthropologie de Paris, (3), 10, 1887, page 652. 48. Verwom, M., Archiolitische und paléolithische Reisestudien in Frankreich und Portugal, Zeitschrift fir Ethnologie, 38, 1906, page 611. 49. Wray, L., The cave dwellers of Perak, The Journal of the Anthropological Institute, 26, 1897, page 36. 39. [page 58] Artefacts from Organic Material. F.S. a) Artefacts from Bone, Plates VII and VIII All the bone that bears the traces of procurement marks from being smashed or broken by humans belongs here. The assemblage is not restricted to the somewhat bland long bones, but also includes all the deer mandible pieces which, in our collection, have an open groove marking. According to Nevill (20, page 189) the Veddas cut open deer bone not only for nourishment but also to obtain ointment for their hair and skin. Some of the smashed bones from the caves are presented in Plate VII; they all show the well-known structure of food remains of Stone Age hunters. Many bones, in particular the thin long bones, had simply been cracked by hand. This procedure is indicated by one of the observations that Mr Allanson Bailey has shared with us (26, page 414), according to which a Vedda from the foothills shared a slain deer with him, and a European who happened to be there stated that the Vedda himself had created the mark [the European saw] on the offered bone. Other bones however had been smashed by stones, as evidenced by the attributes on the smaller and the smallest bone fragments. Smashing bone with a stone hammer has for example been reported for the Andamanese (see Man, 17, page 351). Small bones may also have been bitten, as observed amongst the Andamanese. Many bones show smooth split faces, which obviously had been created with blows from stone knives. A very clear example is the knife marks on the Fig. 216 piece (Plate VII illustrating the backbone of a deer. Also the upper end of a humerus (Fig. 217), from a young sambar, showing where it was detached with blows from a stone knife directly aslant the long axis of the bone, [page 59] We can hardly answer in the affirmative the still disputed question as to whether the bones wwe found, broken in such a way as to leave a point, could have been used as implements, or how they could have been used. The tibia of an axis deer that runs to a point (Fig. 219) could have served as a dagger or something similar; however, it shows no artificial working to assist that purpose and no use 36marks, so we do not believe in its status as an implement. The same applies to the lower end of a femur, broken into the form of a pointed stick, from a young sambar (Fig. 218), and to the awl-shaped bone fragment running to a fine point (Fig. 220). We regard all these superficially useful bone pieces as accidental products. Many bones are bumt, many quite carbonized. The troglodytes therefore possessed fire. For us it therefore follows that they knew the art of generating fire. Admittedly it is openly stated that the Andamanese cannot make fire, but instead keep it burning and carry it with them (Man, 17, page 150). The same point has been observed for Tasmanians, for whom there are detailed descriptions of fire sticks. Thave a further reason for believing that the Veddas must have known the art of starting fires for a long time, and that is the form of their most advanced (albeit primitive) fire implements, which involve a ‘wooden pan equipped with a boring piece rotated with the hand. When one compares our illustrations of a Vedda fire-making tool (26, page 452) with a similar piece volunteered by Lord Avebury as a Tasmanian fire stick (2, Plate 38, Fig. 224), or a second similar artefact from the collection of Bamard Davis (illustrated by Ling Roth, 24, page 83), the similarity is so great that one could easily mix up one with the other. Equally relevant is the fire borer from South Victoria which Brough Smyth (30, 1, page 393) illustrates. The fire borer with two wood sticks spread not only to Australia, but also (as far as one can say) across the entire world (here, compare W. Hough's instructive treatise, 11), from which one may well conclude that it must be an enormously ancient instrument, to be found across such a huge expanse. Further, it is accepted that the Sinhalese, who had already entered India’s Iron Age at the time they invaded Ceylon, would have had a more advanced form of fire borer, so the Veddas could not have adopted their own primitive instrument from the Sinhalese. Such an advanced fire technique can indeed be found amongst the Sinhalese, although today they very rarely use it, because matches and iron strike-a-lights have spread across almost all of Ceylon. In Bibile on the eastern bluff of the central ‘mountains, on our last trip to Ceylon, we asked about Sinhalese fire sticks. We admittedly received no sticks in use, but instead a model created by an old farmer [page 60], which appeared fully typical, and eared that the tradition is still alive, This Sinhalese fire borer (see the accompanying illustration, Fig. 21) consists of a wooden pan, that is, a rounded wooden cudgel of 45 cm length, on which one side has been made into an even flank of approximately 25 cm length with a knife. Within it, three round depressions of 1 cm diameter have been worked. The boring stick, with its conically shaped, lower end passing directly into this pan, is 23/4 cm long and somewhat rounded on top. A coconut shell placed on top is used to press the fore- piece into the pan, and the rotation is achieved with a snare, which makes a sleeve around the boring piece and whose ends are tied at both ends to wooden stubs of 7 cm length, to allow comfortable use [of the instrument]. A soft, white substance that looks like elder pith is used as tinder. This method of making fire, with the wooden bore set in motion by a snare, rather than rotating it using powerful force with the unaided hand, has also spread across the world. The Sinhalese instrument is very similar to those used, for example, by Eskimos (see Hough, 11). (Fig. 21. Sinhalese fire borer. One quarter natural size] According to a remark by B. Smyth (30, page 400), based on Stevenson, the Indian Brahmins still use a fire stick, which resembles the just described Sinhalese instrument, to prepare their cooking fires and for religious ceremonies. To be sure, the bore of the Brahmans is set in motion with a so-called bow. “Bow” is not however the right expression, as it would prompt one to think of a violin bow, whereas in fact there would have to be a snare tied in a sleeve around the boring piece to set it in motion. The 37principle of the “bow”, with the snare tied at both ends to a single long piece of wood, is also exactly the same, with both snare ends fastened to small bits of wood that can be gripped by hand. A progression can be seen in that this bow allows the instrument to be used holding it with only one hand. [page 61] Given the circumstances that the Brahmins use the above-described fire maker for cult Purposes, one may reasonably conclude that it must have a considerable antiquity in India, to which can be added that the Sinhalese, who implement such a type of fire maker, invaded Ceylon from India. In some places this Sinhalese method of preparing fire has also been passed onto the Veddas, for Stevens (34, page CLIII) observed one Vedda group — he does not say which — lay a stone or coconut shell at one end of the vertically positioned, wooden borer, press their forehead against it to hold it firm in the pan, and then sct it in rotation with a snare of bast held at its ends by both hands, However, this is not an old Vedda technique, but one learnt from the Sinhalese. After this digression we retum to our bones. When one investigates which bones show traces of fire and which do not, one soon observes that all of the long bones (for example, those of deer) that are broken into substantial pieces are not burnt. In contrast, the numerous small and miniature fragments of long bone are in a partly or entirely carbonized condition. The foot bones of deer are also burnt almost without exception, Our Plate VIII shows shows carbonized hoof parts of an axis deer and a sambar (Figures 229 and 230), a completely carbonized deer phalanx (Fig. 228), and a cuboid-navicular with burn marks (Fig. 225). Unfortunately, the effects of fire are not shown clearly enough in the illustrations. How are we to explain that the larger broken pieces of long bone have no traces of fire? A custom of the present-day Vedda can perhaps clarify this. When the Vedda have more flesh than they can use at that moment, they cut it into strips which are hung over the fire and then dried out in the sun, till they are as hard as wood and can be stored in this condition (26, page 417). Cutting up the flesh to be dried for storage could account for the pieces of bone addressed here. It would explain the lack of fire traces on the bone. In further support of this hypothesis, there are [page 62] frequently cut-marks from stone knives on these bones, which could have been created through cutting flesh. Such cut-marks from a stone knife can be clearly seen on a lower humerus end from an axis deer (Fig, 222, indicated with an *s"), and on the femur head from the same species (Fig, 221). In addition to these scratches accidentally brought about during detachment of the flesh, knife marks and notches with a more intentional character can be readily detected on some bone fragments. This is shown by the Fig. 208 (Plate VII) fragment, which has a shallow elongated incision of 2 em length made with three parallel knife cuts; two further cuts lie perpendicular to it. On one rib fragment (Fig, 209) there are four parallel cuts somewhat aslant to the direction of the rib, recognizable as deep cuts, and between them the upper surface of the bone is partly splintered; above them, one of the sides has been cut across, leaving a three-sided bony prominence. The cut-marks on a tarsal bone of a wild pig (Fig. 206) are particularly odd and intentional. They show two sets of parallel cut marks. To the left half (as pictured) one can see a deep, long cut aslant the bone’s axis, and two finer cuts with the same orientation further towards the other end. Three other cuts have been made perpendicular to the bone's long axis, two of which are very sharp, close to each other at the approximate centre of the item, and a third near the broken end. Two fine scrapings are also found on the unillustrated side of the fragment. The small muntjak horn in Fig. 215 further shows a deep notch, along with one definitely made at its basal end. The point of the hom is broken off. The bone splinter pictured in Fig. 213 is also deeply incised, with the notch clearly achieved through a sequence of several stepwise cuts; it is possible that this had been used to detach points. A corresponding practical use for the other pieces is however ruled out. So what is the meaning of these unambiguous cuts and notches? Do they reflect a certain use as jewellery? Their parallel arrangement would suggest this, but on the other hand [there should be no mistake] that pieces of 38decorative bone could also have served in some way as instruments or amulets, and a decoration could well have sense specifically for such purposes. Better yet, they could have been mnemonic markers to jog the memory on some awaited event. Sticks with similar notches were for example used in Australia as memory aids for carrying messages; without the verbal explanation from the messengers of what the notches mean, the recipient would be left clueless, as Spencer and Gillen (31, page 141) state for Australian notched sticks. W. Roth (25, Bull. No. 8, pages 9 ff.) volunteers a different view on the nature of Australian memory sticks; he is convinced that the notches carry no information [page 63] of that kind at all. For him the stick is simply a property marker, a type of seal of power, required when for example the stick has to be passed between intermediaries, When someone sends outa request for some wares through exchange, but without sending something of equivalent value with the request, he performs this with a notched stick or, usually, a badge. The seller closely observes the nature of the marks or the badge and supplies the wears. These go back to the sender along with the stick. Now the Payment should follow, so the equivalent is returned to the sender, accompanied by the same stick. The latter recognizes the badge and the deal is settled, Many people send out message shafis in the form of snares with knots, which are similar to the message sticks in Spencer and Gillen’s, but not Roth’s, sense. This occurs with the Kanikars, a tribe of mainland India, as they showed to Jagor as we have already related (26, page 457). From the Celebes We also have the tale narrated by Kruijt (28, page 48) whereby the ancient Toligowi stem fell into hostilities with the Tolage Toraja, because the latter had sent them a snare with too many knots init as the invitation to a sacrificial feast, so that they arrived too late for the feast. In tis case the quantity of knots signified the number of days until the feast, so that by untying a knot every day the correct termination would be reached. Perhaps the notches on our bones could have served a similar purpose. Notches and rows of notches also occur frequently on bones from the European Palaeolithic, concerning which one should consult the work of E, Piette and J. de la Porterie (23) over the grotto of Brassempouy. A noteworthy passage in an anonymous report from a government official dating to the 20" year of the last century [1823], brought to light by Le Mesurier, concems the Vedda message sticks. The passage already quoted by us (1, page 347) runs as follows [translation back into English from the Geman translation]. “The Veddas know no writing, but the various tribes keep up a rough correspondence through small sticks of wood cut into various shapes. Refugees care for them as though they performed the role of a passport when they travel from one tribe to another, andthe refugee's treatment upon arrival hangs upon the recommendation of the talisman.” That this must involve a quite different type of memory stick is clear, as this service could not have been easily invented. Nothing like this is known amongst the Veddas stil living today, less so the just ‘nominated hypothesis of message sticks or (more precisely) message bones, which, given the rapid decline of the remnants of this race, perhaps arouses little wonder, The question now presents itcelf as towhether anything was found in the cave material that could be interpreted in terms of the [above cited] anonymous report. [page 64] The latter indicated that the “passports” were admittedly of wood, but pieces of bone could just as well have served this purpose, Pethaps the heart-shaped plastron of a water tortoise, Nicoria trijuga, illustrated in Fig. 210 could be relevant here. This piece of bone is naturally shaped like a heart, with the tip directed towards the front end of the shell, but it is not likely that it could have obtained this particular form without human modification. The borders are so even that without any doubt they had been finished this way with Cutting work, distinguishing them considerably from the other parts which appear rough and uneven; in addition, three small, fine cut-marks from a knife can be detected at the upper right on the extemal face. This tortoise shell heart could have served equally well as a passport, in the sense of the anonymous report, as a kind of message stick; if however anyone would prefer to describe it as a 39natural amulet, this would not be correct. Churingas, serving as passports, in the sense that they display the bona fides of the bearer, and protect him as long as he carries it, occur in Australia and were correctly distinguished by Spencer and Gillen (Joc. cit.) from the first mentioned type of message stick, whose notches should reinforce the memory of the message. Bone points. There are four such points, presumably spear points, but each artefact shows its distinctive character, so that it is not possible to speak of a uniform type. The most complete is the one in Fig. 200. It is a flattish shard of bone of 29 mm length and 24 mm maximum breadth, and has been worked to a point, obviously with great care, with knife cuts. A piece of the original bone surface can be seen from the middle of the point on the left to its base at the right; all the rest has been cut away, with the knife marks still clearly visible, The other side shows no working. Quite different is the point in Figure 201 (a and b); it is a small bone splinter, of 24 mm length and 11 mm breadth, that has been broken at the base and along one side, and runs to a true point. The lightly swelling outer surface (a) is smoothed. On the badly damaged inner surface (b) one can still observe ‘two smoothed facets running together towards the point; the tip, where they would have met, has unfortunately broken off. The 30 mm long point shown in Fig. 202 is of an irregular form, with its outer surface coarsely smoothed. The unillustrated surface shows clear polishing only towards the tip. How these two points could have been so evenly polished can no longer be determined, but probably with a mussel plane or a boar tooth (see below), The following [bone] point (Fig. 205) is yet again formed and treated differently. It is a bone splinter of 6.0m length and 12 mm breadth. The point itself does not appear [page 65] to have been touched up, although we cannot be certain of this as part of it is broken off. The other end has been cut to shape, obviously to serve as a handle, and the level part on its upper half has been skilfully finished; perhaps it had originally reached the tip. The other side shows no working, only a slight hollowing with groove marks. Spatula instruments. The best example of this type is a dark, partly carbonized piece of bone of 37 mm. Jength and 19 mm breadth (Fig. 204). The outer surface (a) is cut smoothly to shape and polished, and shows manifold knife marks under an oblique light. The inner side (b) partly shows the original hollow form. The forward edge and side edges have been worked into shape by cutting from both faces, but are unfortunately badly damaged. The Fig. 203 piece could be a still unfinished, spatula instrument. It is a smooth bone fragment of 65 mm length and 26 mm breadth. With the exception of an angle springing out below, the sides of the bone had been cut parallel; atthe front it is rounded off with a lightly protruding point in the middle, but not worked there with a knife. Polishing instrument. Here we can assign a piece of bone of some 9 cm length and 1.8 cm breadth (Fig. 214), which has on one of its long faces a skilfully finished depression of about 4 cm length and ¥4 em depth. This appears to have been polished through use, as also does the half-rounded, cut off, end of the bone. The dark sheen of this piece comes from treatment with acid; it had been completely covered with a layer of crust, from which it was freed [by the author] with acid. The disadvantage of this method is that the bone afterwards has a vamished appearance. Quite similar polishing instruments also come from the European Palaeolithic. Handle piece. A unique piece in our collection is a handle, unfortunately broken, from a deer horn (Fig. 207). It is a hollowed-out piece of deer hom of 5 cm length, both ends cut across, particularly so at the thicker end (1 cm in diameter). Knife marks are clear here too. The hollowing most probably served to receive some sort of instrument, perhaps a point of bone or stone, which could then be fastened with resin. This piece had been found at the mere depth of 35 cm in the shelter deposit. 40Bored-through bone. A left, somewhat damaged heel bone of a deer (Fig. 212) shows a deep hole on one side, near the articular end. From here a narrow groove runs to the end with the heel bump, lacking however the cartilage end. This groove is not present on the remainder of the piece at my disposal, which, as with the hole from which it issues, is the reason for recognizing its artificial status. Perhaps we here have a small bone panpipe, which had played such a large role in the European Magdalenian. [page 66] Boar tooth plane. From Man (17, page 402) we learn that the Andamanese used boar teeth as planes and polishers for bows, rudders and the like. In their hands this instrument could be turned to such purposes, and so was highly valued, and when necessary the inner edge was sharpened with whale skin. Stimulated by this report, [inspected our material closely. We have only two fragments of boar teeth, one from the Nilgala cave and the other from the Balawalaboka shelter. The latter, almost 2 cm Jong and blackened through fire, shows apparent use marks. Fine scratches could be observed with a magnifying glass and small breaks along one edge, which could have been produced through planing use. This fragment is not included in the illustrations, but it nonetheless remains probable that the proto-Vedda, as do the Andamanese even today, had used swine canines as tools. Finally, when we glance over this small bevy of instruments, and (as with the stone instruments) try to draw comparisons with the European finds, so we can say with certainty that the technique of working bone, and the form of the stone instruments, testify to a late Palacolithic evaluation of the prehistoric remains from Ceylon, because the late Palaeolithic has quite similar instruments. b) Artefacts from Mollusc Shell, Plates IX and X. Planing shells. Numerous shells of the large and heavy Helix (Acavus) phoenix Pfs. (see below), a proportion of them found lying together in clusters, came to hand from the excavation of the Nilgala cave. The majority of these show a sharply edged hole in the vicinity of their opening on their outer whorl (Plate IX, Figures 243 to 251). Some of these holes are circular or semi-circular, with a diameter of 9.5 to 21 mm (ten pieces in our sample). Thirteen others have an oval opening, with the long axis generally running perpendicular or obliquely towards the opening, rarely parallel with it. The long diameter of these oval holes varies between 17 and 25 mm, and their cross-diameter between 13 and 22 mm. The distance of these holes from the opening ranges in 23 cases between 7 and 15 mm, most commonly between 9 and 11 mm. On several pieces the rim of the opening appears irregularly broken and deteriorated. In eight examples there was also a second hole near the opening, at a distance of 18 to 34 mm from the first hole, located closer to the upper margin of the outer whorl (Fig. 250). Three examples have the second hole opposite the first hole on this outer whorl (Fig. 249). Additionally, one has the second hole on the border between the last and second-last whorl, at a distance of 15 mm from the first hole. Finally, two [page 67] cases reveal three holes, with the third hole within the second whorl (Fig. 251). Considering also the circular or oval form of these holes and their sharp edge (if not broken), we can clearly ascertain that we are not dealing with accidental damage, but with some intentional purpose, even if we cannot quite determine the true significance. We had developed the view that the purpose of the holes may have been to extract the snail from its shell, with a small stick, to eat it, but then the regularity of the holes would be an unexplained luxury, for the same purpose could have been achieved through a simpler manner. Then we were fortunate to have a meeting with two gentlemen, Prof. K. von den Steinen and Director W. Foy, in our house. When considering this part of our Vedda collection, they called out “You are dealing here with planing snails”. Von den Steinen followed this with “I know them from Brasil” and Foy with “and I, from Australia”. So in fact the holes were instrumental. 41Von den Steinen (32, page 375) reported the following from Borord, on the southem Lourenco, southeast of Cuyabé. “Their plane was a mussel, 10 cm long, in which sharp holes had been knocked with an Oaussiinuss.” The illustration, which we reproduce here (Fig. 22), shows four holes on the last whorl of the mussel, clearly made jagged through use. The same author found the same mussels in use amongst the Indians of the Xingu headwaters not only as knives but also as scrapers, planes and polishers. An Anadonta “served as a plane, to polish the grip of a stone axe or an oar With fine scrapes, but with a hole fixed in the middle. The people bite off the outermost shell (the cuticle) with their teeth and knock the hole in with a pointed Brazil nut”. (Fig. 22 has the caption: Shell plane from the Borord, after K. von der Steinen (32).] ‘Von den Steinen further sent us pictures, from his work on the Marquesas Islands, showing a large Cypraea shell with holes at both ends, which is “a breadfruit plane or scraper. With the toothed aperture in the hand, a person pushes one hole in and the pulp emerges from the other hole”. This is an interesting specialization of the same idea, We now proceed to Australia. We think Mr Foy for sending us some rare literature, The first example is a report by H. Basedow (4) on the coastal populations of northem Australia (the Northern Territory of South Australia). According to this author (page 49) the shells of the large helicoid, Xanthomelon pomum (Pfi.), were used in the preparation of weapons and instruments. For this purpose, a hole was carefully punctured in the uppermost whorl, [page 68] with its rim made as sharp and abrupt as possible. For planing and polishing, the shell was gripped as shown in the accompanying Fig. 23a (taken from Basedow); the aperture rim that lies deeper in the downward planing direction served as the planing edge. The shell would be held differently (Fig. 236) for sharpening the point of a spear or other weapon; the weapon’s point was locked between the shell and the thumb, with the thumb pressing the point against the hole in the shell, which was then (in all cases) drawn towards the body. (Fig. 23 has the caption: Use of shell planes in Australia, after H. Basedow (4).] The Xanthomelon-type shells appear very similar in size and form to our Ceylon Helix {examples}; the plane openings correspond to each other, not only in being close to the mouth, as is the case with most of our planing shells, but also in backing [directly] away from it. Walter E. Roth (25, Bull. Nr. 7, page 21) describes the use of a snail shell for planing that is somewhat different, but in principle a matching use, from the Tully River district and neighbouring islands, also from North Queensland. A Xanthomelon shell is used here too, Xanthomelon pachystyla (Pfr.), indeed in the following way. An empty shell is stoppered with grass and then rubbed over a sharp stone edge until a long opening is made behind the aperture; the rubbing process is facilitated with a little water, and soon the rubbed part is so thin that it can be broken with the nail of a thumb. At this stage the manufactured hole is not used as a planing hole, as was the case with the snails described by Basedow, but instead the entire whorl of the shell is detached with carefully produced breaks, so that only the rim of the aperture (with the shell all around it broken off) and the base of the [outer] shell remain, This piece is then held with the right hand, with the tips of the thumb and middle finger juxtaposed in the mouth, and the base turned upward; the sharp edge behind the aperture, exposed through the polishing, functions as the planing blade. This scraped-off little plane, held between thumb and aperture and repeatedly drawn up and down, is used for planing Castanospernum and Cycas nuts. Another instrument described by Roth is used quite similarly to our proto-Vedda planes, with the sole difference being that the shell of a mussel and not a snail is employed. According to Roth (page 21) the Aborigines [Schwarzen] told Mr T. Petrie that, during Brisbane’s early days, they took a Donaxschale, made a hole in the middle, and held it with the concave side inwards against the palm, Then, with the upper edge of the manufactured hole pressed against the spear point, they scraped the latter, shaving by shaving (Which fell through the hole in the shell), until the weapon point was ready. The accompanying [page 69] diagram [Figure 24] shows this form of use. Compare here the above-communicated 42notification (page 67) from von den Steinen over mussel planes in South America. [Fig. 24 has the caption: Use of a mussel plane in Australia, after W.E. Roth (25).] Given these nominated examples, there can be no doubt that our Helix shells from the Nilgala cave, furnished with similar holes, were also used as planes. Probably they were used for polishing the shafts of spears and other wooden implements. Our experiments with ash have shown that its bark could be easily removed and the timber polished. How the holes in the shells were made can naturally not be answered with certainty; however, the very regular form of the holes shows that they could not have simply been knocked out with a stone, but more likely would have been cut out with a stone knife. They could also have been made through rubbing the surface away and then breaking the thinned wall of the shell with a fingernail, or otherwise finished partly with an implement. It is an open question whether the Veddas today still use such a shell plane, for it has not come up during observations and so nothing is known on the issue. We have not found any mussel shells bored right through, as are used in America and Australia; obviously the Helix plane satisfied the desired purpose. The case of our snail planes is a further, interesting instance of the fact that observations on the primitive material culture of still-living people are unreplaceable for the elucidation of prehistory. Without these expeditions to observe such living peoples, itis hard to imagine how the idea would have occurred to anyone of the use of the Helix shells with holes, in the Vedda caves, as planes. It now appears very likely that many of the bored-through snail and mussel shells, which occur frequently amongst the prehistoric finds of all lands, and which have been interpreted as pendants, could have had practical significance as planes. A thorough review based on this point might readily lead to interesting results, ‘Mussel implements. Remains of two types of mussels were found in the Nilgala cave, which I have identified as Unio thwaitesii Lea and [Unio] corrugatus Milll. (see below). Only one undamaged shell of the former, larger and usually thinner-shelled type has been found. Most of them have one end broken off along a semi-circular margin, leaving jags sticking out (Plate X, Figures 253 and 254), which without a doubt could have served as spatulas. That this form was intended can be seen from the knife cuts on the two pieces, particularly clearly on the unfinished piece in Fig. 252. Of the three finished pieces, one (Fig. 253) has its obviously broken-off lower margin completely and regularly blunted, while the border of its [page 70] other margin appears thin and sharp. It cannot be decided whether this effect on a fairly large part of the shell had been caused through use, or intentionally (to make the shell more solid). In the illustration the blunting of the border of the shell is unfortunately not well displayed. Many smaller shells could have served as planes or scrapers. Indeed they may have straight segments with use wear running all the way between the free edges, as with the example illustrated in Fig. 255. Otherwise they were employed as broken pieces, with a part of the thicker clasping margin used as a handle, while the free edge sustained damage through all kinds of use. The other Unio type, corrugatus, is much smaller and thick-shelled, and it also was used as a scraper, in fact so often that most shells no longer retain an intact, oval margin. We now regret that we did not pay more attention to the mussels during our cave excavation. As we thought we were simply dealing with food material, we collected only a small number of pieces on which we could determine their artificial shaping. Fig. 258 displays one of the few intact shells we have, although even here itis likely that one part of the edge, at the lower right, is broken off. Another (Fig. 260) shows a cut running straight along its lower margin. The Fig. 259 shell, which is a particularly large example, is quite different in its form; it has been rounded off and shows thick borders around it, obviously created through use, including a few places where it appears to have been dragged. As we had earlier reported (26, p. 431), there is an account from Nevill (19, p. 33) conceming how the ‘Veddas in olden times fashioned arrowheads out of Unio shell, before they came in possession of iron arrowheads. This account would appear even more plausible given that Nevill had been shown no signs 4Bof stone implements. We found no points of mussel shell in the caves, although we had paid close attention to this possibility, based on Nevill’s account. Further, why should this non-durable material have been valued for points when stone could have been fashioned for the same purpose? However, it is always possible that the tradition communicated by Nevill is grounded in the use of Unio shells as serapers and grabers, and that Nevill misinterpreted it as applying to arrowheads. Whether the Vedas still use mussel shells as scrapers is unknown, but it seems quite unlikely because iron arrowheads perform the same function much better. Once we had seen a mussel shell in the possession of a Vedda, in the district of Devilani; after he had burt it, the shell served him as a [portable supply] to be chewed for the lime that he needed. ‘Mussel shells are used as smothers, polishers and planes amongst many living people. Man (17, page 376) says, regarding the Andamanese, they always had Cyrena shells with them, which were useful for many things, such as working the wooden parts of their [page 71] arrows, sharpening their bamboo and reed knives, taking the place of the inner edge of boar canines (when these should be used as planes) and of spoons, or as knives to cut roof coverings and to separate meat joints, or as scrapers to remove the pulp from the fibre to be wound into snares, and further to make decorative cuts on weapons, implements and so forth. The mussel is therefore a truly universal instrument for the Andamanese, Similarly, the Papuans of southern New Guinea use Cyrena shells to sharpen their bamboo knives, according to D”Albertis (8, p. 221). Very diverse is the use of mussel shells amongst Australians, According to Brough Smyth (30, I, p. 349), the Aborigines of Victoria use mussels to scrape and Prepare coats. He nominates the mussel as a worthy instrament that would be commonly used for other Purposes too, but when a well-formed and favourite mussel became blunted through use, people would sharpen it with a stone, According to W. Roth (25, p. 21) fragments of strong shell would be used as scrapers, including Donax, Cyrena, or even Myrtilus if its edge was strong enough for the purpose. Mussel shells also served as knives, according to him. Cuts with a Cyrena shell (p. 22) can allow bark to be peeled off, and to crack the distal end of a throwing stick to insert one’s fingers. At Cape Gratton, certain nuts were cut into slices with a Perna. On the Pennefather River, Roth even saw Tellina shells gripped by the hand to make cuts in the body, before glass came into use for that purpose. Basedow (4, p. 50) saw the strong edge of Cyrena used, among the northern Australian tribes that he studied, to give stone points and knives their last, perfected shape through the means of tiny flakes. Mussel shell was also used amongst them to cut wood, through a combination of force exerted against the wood and semi-circular movements with the right hand, which held onto the shell. To show how widely spread across the world is the use of mussels as implements, let us finally look again at our reference to von den Steinen's report (32, p. 200) on the Xingu Indians of Brazil who use river mussels to cut, scrape, plane and polish a wide variety of materials. The manioc root would be scraped with one Anodonta piece, wood would be finely shaved with another, and so on. To show that this is not the only agricultural tribe using mussel knives, we have an observation by Jagor on the indigenous people of Luzon who, lacking sickles, use the sharp edges of freshwater mussels (probably Anodonta purpurea Val.) to cut off rice ears; a similar practice occurs on the Marianas (see E. von Martens, 18, page 31). With that I close this overview which could easily be stretched further. [page 72] c) Artefact from Wood There can be no doubt that, among the proto-Veddas, as the world over, the role of implements of wood would have been important, indeed probably more significant than that of stone, bone, snail shells and mussel shells. But the latter alone have survived, while wood has perished except under very favourable circumstances,The sole surviving artefact of wood in our assemblage is a small forked piece (Plate VII, Fig. 211) of 25 mm length, strongly weathered, which has been fritted by termites and broken at one end. The intact end shows what is without any doubt an artificial incision with a rounded base, apparently to take a snare, How could this little piece have been used? A plausible solution for this riddle comes from a Sinhalese device for sling-stones. The pocket which takes the sling-stone is fastened with two snares, and to keep this setup stable, a curved end of a small piece of wood is used, which takes both of the sling strings tensed apart in this way (see Fig, 25). Most probably the small piece of wood from the Nilgala cave was the gable piece for such a sling bow. That would raise the question which is by no ‘means easy to answer: did the Stone Age people possess slings, and in the same form as the sling bow? There is no shortage of stones which could have served as sling-stones amongst our finds (see the section on lithics), and a simple sling could have been very easily made, but it would still be unlikely that a complicated apparatus such as a sling bow was known by the proto-Veddas. Here we point out that the gable piece is the only piece of wood in our assemblage, thanks to its bark which conserved this time-wom piece of wood despite its disintegration. A better idea would be that this wooden gable fragment could be of modem, Sinhalese derivation and just happened to infiltrate the Stone Age layer from the upper layer. (Fig. 25 has the caption: Sinhalese sling bow for stones.] These days the Sinhalese sling bow is used in the lowlands to hunt birds and as a child’s plaything. We have also found it amongst some Vedda groups, at Angesiedelten in the vicinity of Kaloday and at Mudugala near Mahoaya, but we prefer to think that it had (along with other cultural possessions) been borrowed from the Sinhalese and does not represent an original Vedda instrument. [page 73] Postscript to our Work on the Toala Caves of Lamoncong, Celebes (27) Stimulated by the results obtained from the review of the literature, and implements of snail and mussel shell yielded up by the Nilgala cave, I have given the mollusc material from our Toala caves in the Celebes another examination. Previously we had understood the rubble of mollusc shell as nothing more than food remains (27, page 49); a closer inspection however quickly revealed that this is not correct. We had collected numerous shell fragments of Cyrena suborbicularis Phil., and since a proportion of them revealed traces of fire, we had not doubted their culinary significance. Closer scrutiny however soon indicates an apparently more important use as scrapers or graters. Fig. | shows a Cyrena shell with a tough operculum and a semi-circular break with numerous traces of use; it had apparently been worked by placing the thumb in the recess underneath the shell’s mouth. However, the Cyrena shells were used not only this way but also to produce particular implements (Figures 2 to 5); these have a quadratic or elongated rectangular form, whereby a piece of the thick operculum always functioned as the hand grip, and the edge opposite was the working edge. The working edge on the scraper in Fig. 5 had been sharpened with retouching flakes, as we can also observe on the other illustrations. Along with the Cyrena shell tools that had been produced to a desired form, there are also the more usual shell fragments that had been directly utilized, as for example the carbonized piece in Fig. 7 with its retouched, broken edge. We had only brought back a few pieces of Barissa violacea Lam. shell, but these also include implements. As with the above-described Cyrena scrapers, the mouth rim functioned as the grip, and the opposite edge oriented cross-wise functioned as the working edge (Fig. 6). The most common fragments of mollusc shell in the Toala caves belong to a land snail, a large, thick- shelled species of Nanina which we have denoted as Nanina toalarum (page 50), and which appears to have been an acceptable dietary item. Close examination shows that here too the shells from the cave appear to have been brought into a quite different purpose, namely as scrapers or planes. These are unusually delicate and ingenious tools (Figures 8 to 38). The spindle (columella) of the shell functioned as the hand grip and [page 74] a piece of the last whorl, varying in its form and size, as the scraper. 45With some of these implements the spindle is merely a narrow and short piece of shell, and when the working edge has not been damaged through use, it shows a perpendicularly oriented, direct or convex long axis (Figures 8 to 16). This edge had apparently been shaped with a stone knife and functioned as the scraping edge. In other cases the piece of shell is larger and more fan-shaped, with its end either semi-circular or branching directly across (examples in Figures 17 to 29). The scraping edge is very often damaged and irregularly broken through use. Finally there is a sickle-shaped implement, whose scraping edge is not oriented across the long axis of the spindle but more or less aslant to it. These sickle scrapers (Figures 30 to 38) are either quite thin or of more or less considerable breadth. Our best pieces are shown at their natural size in the accompanying text plate. In addition there are farther examples, some of which had not been finished, and some of which were strongly damaged through use. Our entire collection amounts to 66 examples, which show that this was a highly prized implement amongst the proto-Toala, and suggests that if we had recognized their significance during our excavation of the caves, more would have readily come to hand. As a handle, the spindle permits a particularly firm and practical grip to these small implements. I know of no other examples of these spindle handles, either in ethnography or from prehistory. [Text plate showing Lamoncong shell artefacts appears between pages 74 and 75.] [page 75] The Animal and Plant Remains from the Caves. F.S. a) Animal Remains from the Caves, Plates VIII to X Almost all of the teeth and bone which can be identified came from the Nilgala cave. The Katragam grotto yielded only a few splinters, Koladoy some deer remains, and Balawalaboka even less. Except where otherwise indicated, the following description will refer to the assemblage from the Nilgala cave. Its deposit held a quite rich yield of animal remains, most of which show the burning traces and fragmentation of cooking refuse. However, the quantity is much less than that of the stone implements and flakes, so the latter have the greatest value for demonstrating a lengthy occupation of the site. One might also well conclude that the greater percentage (by fat) of the food remains had been affected by dogs, which make their presence known as will be shown further below, and then further degraded by Jackals and other scavengers at night during times when humans were away. Termites also would have had a strong destructive effect. Finally, very many of the bones would have been thrown outside the cave by the humans themselves, and would have rolled down the slope outside. Accordingly, a larger faunal assemblage would have been recovered if our excavation had proceeded outwards from the interior of this grotto. An initial answer to the question of the age of the Ceylon stone artefacts can be obtained from ascertaining the animal world with which they are associated. Accordingly the remains were identified with great care, and in difficult cases (as earlier done with the analysis of our Celebes finds) I consulted with our Basel palacontologist, Dr H.G. Stehlin, who was happy to help. Our Vedda work (26, page 412) presents a list, compiled as carefully as possible, of the faunal food resources of the Vedda today, which will be useful for comparing with the cooking refuse of the Stone Age inhabitants. The following types of animals revealed themselves in the cave material. [page 76] Mammals 1, Semnopithecus priamus Blyth We found at least three mandible fragments of this monkey (one is illustrated in Plate IX, Fig. 233), as well as a loose canine from the upper or lower jaw, a cranial fragment with the border of the orbit, tibia, femoral head, radius and some further bones. 46The identification of the genus is quite secure, but there would be some doubt as to whether the species is priamus Blyth or cephalopterus Zimm., because these two forms differ particularly through their externally visible features. A female monkey that we shot down at Tanamalwila, southwest of Nilgala, showed itself to belong to priamus and not to cephalopterus, from its coloration and the strong development of the hair above its eyebrows. According to the literature (see Blanford, 5, pages 32 and 35; Sir J. Emerson Tennent, 37, pages 7 and 10; Kelaart, 12, pages 2 and 3), the cephalopterus species lives on the southern and western lowlands, and the priamus species on the northern and eastem lowlands and their forested slopes, which thus makes it probable that the remains from the cave can be ascribed to priamus. According to our list of edible animals all forms of Semnopithecus (Sinhalese, “Wandura”) are eaten by the Veddas with relish (26, p. 412), in some places as the most preferred food. Indeed our Tamil coolies happily mixed pieces of monkey flesh into their rice. 2. Macacus pileatus (Shaw) A Jeft mandibular ramus (Fig. 234) has a build and dental characteristics that show that it does not belong to the Semnopithecus genus but to Macacus, which in Ceylon occurs as the species pileatus Shaw, which our bone compares well with. According to Kelaart (12, page 8) this monkey is common in all parts of the island, and according to Tennent (37, page 5) in the south and west. In the Nilgala district itis less common than the previously mentioned species. Our list of edible animals (p. 414) states “We observe here that according to our knowledge only the Wandura, and not the Macacus pileatus Shaw (Sinhalese, rilawa), was seen as a wild animal suitable for consumption”. The recovered mandible piece makes it probable that the macaque had not been disdained, or less so in the past. Admittedly there are no bum marks as there are on some of the remains of the previously mentioned species. 3. Canis familiaris L. (or Canis aureus L.) One mandible corpus fragment (Plate IX, Fig. 235) containing two incisors, the canine with its tip lost, the first premolar, and the socket for the second premolar, certainly belongs to a canid, but the question whether it is a pariah dog or a jackal cannot be answered absolutely clearly. [page 77] Dr Strehlin and I think that the relatively thick and short symphysis indicates a dog, as this area would be more elongated on a jackal, but to avoid a risky assessment I sent the mandible to our foremost expert on canines, Prof, Th. Studer in Bern, requesting his opinion, which came back as follows. “The fragment unfortunately belongs to a part of the skull that has few diagnostic characteristics; it is from a young animal, around eight months old, so the canine had not yet fully erupted, the three premolars had not yet all come up, and the first premolar was still abutting the canine. The second [premolar] is broken thickly at its root, which is still present in the socket. I have compared the specimen only with Ceylon jackals. With these the canine is much slimmer, less broad sagitally and less flattened interiorly; this specimen shows a posterior edge that is much less developed on a jackal; also the anterior part of the mandible is much thicker and plumper than with jackals. On the other hand the most anterior premolar has strong cusp development that best compares with jackals; these are weaker and lower on the dingo. The height of the corpus at the location of the anterior premolar, and the sagittal diameter of the canine crown, Compare best with those of a pariah from Sumatra; dingos have larger dimensions. Overall my diagnosis is a dog, not a jackal. This is all I can say based on a comparison with numerous dingos and pariahs and two jackal skulls from Ceylon; unfortunately this small fragment does not include a definitive part.” 47If we assign this fragmentary jaw to a dog on the basis of [Studer’s] kind service, then it would be clear that a dog jaw should not be included amongst the game animals, as the dog would certainly not have served as food. Of course jackals would hardly have been eaten either, but they are quite often killed when found rummaging around human habitation locations. Whether the tooth marks found on several of the bones belong to a jackal or a domesticated animal would naturally be difficult to decide, Previously we had concluded (26, page 450) that the Veddas had acquired the dog through the Sinhalese, because the Vedda dog appears no different from the pariah dog of the Sinhalese and Tamils, and also because the Veddas occasionally trade for young dogs from Sinhalese towns, when their own breeding efforts are insufficient. Also relevant here is the case of isolated “Wild Veda” families living alone on rocky mountains without owning any dogs, as for example the small horde on the Danigala which we had visited in March 1907. Further examples of tribes that had acquired the dog very late are not lacking. Before the European oécupation the Andamanese had no dogs, and first used them for hunting in 1865; now they value them greatly (see Man, 17, page 341). Similarly the Tasmanians first acquired the dog, which they [page 78] soon learned to use in hunting, from Europeans (see H. Ling Roth’s compilation, 24, page 111). This find of a presumed dog jaw in the Nilgala cave re-opens the question of a Vedda dog, and should be examined carefully in future research on the caves. 4, Melursus ursinus (Shaw)? ‘A small fragment from the canine of a large carnivore probably comes from a sloth bear, less probably from a panther. Veddas do not eat sloth bears (26, page 414), nor can I find evidence that they were eaten in India. Conspicuously, the major food items of sloth bears are vegetables, honey and insects. 5. Hystrix leucura Sykes Ten loose teeth were found of this Indian porcupine, three of them illustrated in Fig. 236. According to Kelaart (12, page 70) this animal is found in almost all parts of the island, and according to Tennent G7, page 46) its flesh is a delicacy in Ceylon, being similar to a piglet in consistency, colour and taste. Porcupines also figure in our list of Vedda dietary items. 6. Sciurus macrarus Pennant Two left mandible rami (Fig. 238) are our only traces of this animal, This well-known type of squirrel, which our jaws accord with in every respect, is no longer very common in the Nilgala district. The flesh of all Ceylon’s types of squirrels is eaten by the Veddas (26, page 413). 7. Lepus nigricollis F, Cay A distal humerus end (Fig. 237) shows the presence of this species. The appropriateness of the hare needs no discussion; it is also in our Vedda list of food items (page 413). The overwhelming majority of our animal remains belong to deer species and especially the axis deer, followed in second place by the large sambar, and then the muntjak as the rarest. 8. Cervus axis Erxl. This species is represented by numerous remains of jaw fragments, many isolated teeth, several antler pieces, and above all broken and burnt bones. Several bones of the axis deer are illustrated on Plate 48VII, including tibia (Fig. 219), femoral head (Fig. 221), [page 79] distal humerus end (Fig. 222), phalanges and hoof bones (Figs 228 and 229), the small hom of a young animal (Fig. 226) and the hom tip of an older one (Fig. 227), as well as the largest of the extant mandibular tooth rows (Fig. 223); its dental canal [Markkanal] is broken off, as is the case with all of the mandible remains. Only some isolated teeth from the upper jaw remained to be found. Judging from the teeth in the Nilgala cave, we have remains of at least five adult examples and at least three with their deciduous dentition. One tooth, one carbonized metacarpal end, and one broken tibia also came from the Kaloday grotto. ‘The axis deer, which used to be particularly common across the whole island, has suffered a stark reduction in numbers through persecution by village Sinhalese and by Arabs, but is still common in the wild. 9. Cervus unicolor Bechst. = Aristotelis Cuv. On Plate VIII we have a distal femur end (Fig. 218), also the proximal humerus end of a young animal (Fig. 217), hoof bone (Fig. 230) and cuboid-navicular (Fig. 225). ‘The remains of this large deer are less common than those of the axis deer. The available teeth represent only an adult animal and one with deciduous teeth, but the bones are fairly numerous, Axis and sambar deer are today the most important animals for the Veddas, after monkeys and boars. 10. Cervulus muntjac (Zimm.) A solitary first incisor and the notched antler piece in Fig. 215 (Plate VII) can certainly be ascribed to this species, as well as some other bone fragments. The muntjak is less common than the other two deer species named previously. In our dietary list (page 413) we could only speculate that it is very probably hunted. 11. Buffalus bubalus (L.) ‘The water-buffalo reveals its presence amongst the cave’s faunal refuse through a broken first lower molar (Fig. 224, Plate VIII) and a carbonized bone. The latter is the left malleolus (Fig. 231), which is the bone protruding from the distal end of the fibula, where it articulates with the tibia proximally and the calcaneus distally. Medially it rests against the talus and on this surface shows a furrow (sulcus malleolaris) ranning from behind above to in front below. The malleolus of an adult, tame water-buffalo from the Basel collection is positioned next to it (Fig. 232) to show that the latter is a good match, albeit smaller. [page 80] Bone from the cave Malleolus of a tame adult Length...............39 mm 35% Height on the lateral side 25% “* 23% The bone from the cave therefore belongs to a large example of the animal. This allows a negative answer to the question as to whether this bone could perhaps not be a water buffalo but instead a domestic steed. Further, our large European domestic steed shows certain differences in its articulating facets, and in the form of the sulcus, from the malleolus shown here, whereas the tame water-buffalo agrees with it on all essential points. Also the permanent lower molar shows quite large dimensions, 49namely a breadth of 15.5 mm, compared to 14 mm for the water-buffalo in our Basel collection (the length of the tooth unfortunately cannot be measured as a result of degradation). In general, the agreement is so great that, notwithstanding the difficulty of distinguishing between water-buffalo and Bos molars, the tooth can be comfortably assigned to a water-buffalo. In'addition to the domestic steed, one other type of bovid comes into consideration for the identification of the specimen from the cave. This is namely the view expressed by several authors that perhaps the gaur (Ham. Smith) used to be present on Ceylon. The basis for this is that is that some names for landmarks carry the word Gaura (Gawara), and that in 1681 Robert Knox (13, page 21) described a bovid in the menagerie of the king of Kandy as “Gauvera”, which tempts one to think of a gaur, Already in 1852 Kelaart (12, page 88) expressed the opinion that it should be possible to find the gaur in its wild state, although this admittedly had not occurred and may never occur. Nevill (22, page 5) opined that the Sinhalese king had introduced the gaur in olden times but that it was later extirpated or died out. Others however have with reason challenged the idea of this deer’s dying out in a land so rich in forest as Ceylon, which even today supports herds of elephants (Sanderson, cited by Blanford, 5, page 485). On the other hand Nevill, in the same article, blunted the point of an etymological meaning in showing that the word “Gawara” (Galgowara in the Nilgala) did not mean a bovid to Veddas but instead a sambar deer, as we also found in several visits to the Nilgala Vedas (26, page 573). From the point of view of animal geography, there would be no objection to the presence of gaurs in Ceylon, for they inhabit all of the large forested areas of sub-continental India from the foot of the Himalayas to the vicinity of Cape Comorin (Jerdon, after Stemdale, 33, page 481), and if the gaur is the only one of India's large animals that is lacking from Ceylon, there would have to be a particular reason for it. That however is not the case, On the contrary, Ceylon lacks a large number of sub- continental Indian mammals, which till today [page 81] are spread to the southern spit of India or used to be so in the past. Leaving aside the small species, I mention the conspicuous examples of the tiger (Felis tigris L.), striped hyena (Hyaena striata Zimm.), Indian wolf (Canis pallipes Sykes), Indian wild dog (Cuon dukhunensis Sykes), Indian fox (Vulpes bengalensis Shaw), honey bear (Mellivora indica Kerr), rhinoceros (Rhinoceros unicornis L.), the Madras sub-fossil (Blanford, 5, page 473), the Nilgiti wild goat (Hemitragus hylocrius Ogilby), and the Indian antelope (Antilope cervicapra L.). The number of large Indian mammals not found in Ceylon is so great that the reason for it must be the separation of Ceylon from the mainland and not extirpation by humans. So the absence of the gaur from Ceylon is no more unusual than it is for these other animal species, Even though we have considered the possibility that our remains could be gaur, we may abandon that thought for the reasons explained here, and accept that we are dealing with part of the skeleton of a water-buffalo. Perhaps these water-buffalo remains can have light shed on them from another view. For the Ceylon water-buffalo there is specifically the question, which is by no means solved, as to whether it is a naturally wild form or a feral domesticate. The wild and domesticated water-buffalos are very difficult to tell apart, most of the difference relying on the tendency to a larger size amongst the former. The domestic animals regularly and happily join wild herds if they become feral (Stemdale, 33, page 491), and the wild animals mix in with the domestic herds and pester the tenders (Tennent, 37, page 55). The burnt bones from the caves and the broken teeth in any case could hardly have belonged to domesticated animals, since Indian culture, on religious grounds, bans its followers from eating bovids. The question therefore boils down to the question of whether we could be dealing with a water-buffalo, from more recent times, which had become feral and suitable for hunting. The domestic water-buffalo ‘was without any doubt brought from India by Sinhalese, whose immigration during the first half of the first millennium before Christ is set down, but the escape of this domestic animal into the wild could have happened much later. Such an antiquity for our remains could unfortunately not be determined with absolute certainty, because, as explained in my introductory remarks, considerable disturbance has occurred through burrowing animals and termite activity. Hence we could not rule out infiltration of 50these materials from an overlying layer into the Stone Age layer, as discussed for the pottery shards. The water-buffalo malleolus and tooth certainly do not appear recent, but instead look visually just like the other bones found with the stone implements. Without taking a strict line, we are therefore of the view that these fragments belong to the olden times before the Sinhalese, and accordingly speak for an origin of the wild water-buffalo in Ceylon (page 82]. However, this is a matter that calls out for renewed research, because the question of whether the wild water-buffalo had a home in Ceylon cannot be resolved without a larger number of water-buffalo bones in clearly undisturbed deposits. 12. Sus eristatus Wagn. [This species] is encountered through a bone fragment, some foot bones and a broken shoulder blade from the Nilgala shelter, as well as a canine fragment from the Balawalaboka cave. A fourth left metatarsal with notches is illustrated (Fig. 206, Plate VII). AAs the boar is common everywhere, the sparing nature of the recovered remains must stand out. Many ‘Vedda groups today, including the Nilgala group, value boar flesh particularly highly (26, page 413). 13. Manis pentadactyla L. A few right metatarsal bones (Fig. 239, Plate IX) represent this species, which is present on our list of dietary items. The flesh of this species is consumed as an aphrodisiac in sub-continental India (Sterndale, 33, page 521). On the basis of a report from L. Fraser, Flower and Lydekker (7, page 207), write that the flesh of Manis rricuspis Rafin. is thoroughly appetizing and hence was treasured by the natives of Femando Po. Accordingly the Indian-Ceylon form would have served well as a prey animal. As for birds, we recovered no definite remains. Reptiles 1, Emyda vittata Peters (ceylonensis Gray) Broken fragments from the shell on the back and the plastron were comparatively frequent; some of the latter are illustrated in Fig. 263, Plate X, as examples, This soft-shelled tortoise is very widespread in Ceylon and lives not only in the largest reservoirs but also in small marshes. It is mistakenly omitted from our list of food items. However, we can vouch for the fact that itis eaten til today by the Veddas from our experiences in Nilgala. [page 83] 2. Nicoria trijuga (Schweigg.) var. thermalis (Less.) All that we found of this tortoise species is the heart-shaped plastron (Plate VII, Fig. 210) described above, where we mentioned its artificial cutting to shape. It is also useful as a food item, as previously indicated (26, page 413). 3. Varanus bengalensis (Daud.) Several vertebrae belong to the Varanus genus. Of the two species present in Ceylon, bengalensis (Daud.) and salvator (Laur.), to the best of our knowledge only the former is eaten; its flesh is equally prized by the Veddas, Sinhalese and Tamils (26, page 414). Tennent (37, page 271) says the delicate flesh is turned into a curry, and Kelaart (12, page 147) praises a soup prepared from tenderized 51Varanus bengalensis and compares it to soup from a hare, Further, our vertebrae show some slight differences from those found on the skeleton of Varanus salvator, and so I ascribe the cave remains without question to Varanus bengalensis. We did not obtain any snake remains from the caves, whereas carbonized python vertebrae were abundant in the Toala shelters of Celebes (27, page 49). In addition the Vedda today disdain snakes, according to Bailey (3, page 288). Even stranger is the absence of fish bones, firstly because our cave lies quite close to the Galoya which is a large river that yields abundant fish, and secondly because the Vedda today shoot fish with arrows, as we had the opportunity to see on this same river (26, page 443), or else capture them through the use of plant toxins that stun them. Our negative finding must be due to chance. Molluses 1. Helix (Acavus) phoenix Pfr., Plate IX, Figures 243 — 251 This imposing snail, which is Ceylon’s only species belonging to the Acavus group, is very variable in its form and size, with the result that more highly and lowly wound examples appear next to each other, as do larger and smaller specimens. Their occurrence was truly rich in the cave floor, and in many cases they cropped up in nests. All the fragments are bleached white, some of them encrusted with earth. With some of them a reddish violet coloration was still evident at their mouth; with the others, they only exceptionally showed any reddish coloured spots, Most examples show the planing holes described above (page 66). The snails were apparently not eaten, because fragments of snail shell were rarely found. These would have been present [were the snails eaten] because the troglodytes lacked pottery to boil food, so the animal could have been obtained only through [page 84] breaking the snail or roasting it in fire, thereby shattering the shell. 2. Cyelophorus (Litostylus) ceylanicus (Pfr) var., Plate X, Fig. 256a and 256b ‘Numerous examples occurred in the cave floor, some intact, some broken. It is not probable that these snails were eaten but not completely impossible. If their occurrence in the cave is not a natural event, then an observation from Nevill can explain their presence in the cave floor. According to this author (20, page 91) the Veddas chew lime together with various types of tree bark, and the lime is obtained through burning Cyclophorus involvulus snails, which belong to a species closely related to ours. This use would have already been familiar to the proto-Vedda if the Cyelophorus snails had been collected for that purpose. The usual Cyclophorus ceylanicus is, in its form and architecture, nonetheless considerably smaller, ‘more finely built and lighter than our specimens from the cave, which distinguish themselves by their size, heaviness, and their opening with a more strongly developed, externally bulging lip. As for the living snails’ coloration, the only traces on our bleached snails are the dark band beneath the keel and occasionally brownish flames on the upper spirals, Dimensions: _ largest example, height 31, maximum diameter 42.5 [mm], Smallest example height 28, maximum diameter 36.5 [mm] Comparative data from Kobelt (14, page 98) for the living ceylanicus are a height of 24 and maximum diameter of 37 [mm], while two specimens from the Basel Museum have the dimensions of 21 and 22 for height, and 31 and 31 [mm] for maximum diameter. These are distinctly smaller values. 52‘Mr Kobelt was kind enough to send me an example from the Méllendorff Collection provenanced to Ambegamuwa (the mountain district of Central Ceylon), whose dimensions [in mm] are 27.5 (height) and 37.5 (diameter), which correspond to our smaller specimens. In addition, the architecture and construction of the aperture rim match, even if the outer lip of most of our examples appears considerably more bulging. Kolbert would prefer not to recognize our examples from the cave as [too] separate, on which point I agree with him; Kolbert will shortly be describing them, named under their own variety. 3. Paludomus (Tanalia) loricata Reeve, Fig. 257 Only a single example of this omately sculptured water snail was found. As for the habitat of this species, Layard (15, page 91) says they live in rapid, tumbling streams where they adhere to the rocks; they inhabit most of the waterways of the southern provinces of Ceylon, where these conditions are found, whereas they are absent from the calm streams of the north. Amongst these rapid, tumbling streams [page 85] belongs the Galoya, which as mentioned flows past our Vedda hole at a not far remove, and the snail would certainly have come from its waters. The form is very variable; our specimen corresponds well with the example illustrated by Hanley and Theobold (10, Plate 121, Fig. 2) except for being somewhat smaller; it also matches Figures 1 and 3 in Brot’s (6) Plate 3. Our snail has a hole of 5 mm diameter to one side of its last abdominal coil. It is therefore possible that it had been converted into a plane, as described for the large Helix phoenix Pfr. (see page 66 above). 4, Paludomus (Tanalia) neritoides Reeve, Figures 261 and 262 The cave floor held a large quantity of examples of this species, very variable in form and size, some slimmer, some more bulging, some with shorter spirals and some with longer spirals. Some items correspond exactly to the illustrated P. dilatata Rve in Hanley and Theobold (10, Plate 125, Figures 5 and 6). However, Brot, and certainly with reason, sinks dilatata into Paludomus neritoides Reeve in view of the full array of highly variable forms, so I have chosen this species name. Some specimens are coloured grey as though but, so it is therefore possible that they had been eaten. According to v. Martens (18, page 28) the freshwater snails, Melanien and Neritinen, are still eaten in the Philippines and Maluku, whereas in East Asia (rather strangely) it is land snails that are used as human food (page 27), Perhaps the Paludomus snails had also been burnt for use as lime. In any case these water snails must have been brought by humans into the cave, 5. Unio (Lamellidens) thwaitesii Lea, Plate 10, Figures 252 — 254 For the use of this mussel as a scraper and spatula see page 69 above. The cave floor holds numerous large and small shell fragments but not a single, completely undamaged example. In form, stoutness and tooth development these remains correspond best to Unio thwaitesii Lea (Lea, 16, Plate 37, Fig, 125), somewhat better than to U. layardii Lea (ibid, Plate 36, Fig. 122 = /amellatus Lea of Simpson, 29, page 856). On some specimens one can see traces of salmon-coloured mother-of-pearl, which Lea (16, page 246) identifies as one of the features distinguishing Unio thwaitesii from U. layardii (page 243). Both forms are however very close, and Hanley and Theobold (10, page 19) declare that they are unable to determine the precise boundary between them. In addition Unio consobrinus Lea (Lea, 16, Plate 45, Fig. 152) from China, India and Ceylon is a quite similar [page 86] form, as is U. marginalis, all being examples of Simpson’s Lamellidens genus (29, page 854), One may well accept that the mussels served both for technological purposes and as food. 536. Unio (Parreysia) corrugatus Miill, Plate 10, Figures 258 - 260 ‘As with the previous mollusc, a rich number of intact and broken shells from this variable species, equipped with strong teeth, were found in the cave floor. Many show traces of having been used as a scraper (see page 69). According to Simpson (29, page 842), Unio corrugatus is a synonym for Unio tennentii Hanley, which is illustrated in Hanley and Theobald (10, Plate 45, Figures 7 to 9) with the dubious find spot of “Ceylon”. As far as can be determined from these hardly informative illustrations, they accord well with ours in form, but then many of our shells are larger than the ones depicted there. The mollusc types noted here were compared for a provisional identification with the materials held in the Berlin Museum, Mr Dr J Thiele has been helpful and very kindly to us in this endeavour. b) Edible Plant Remains Without further ado, it can be assumed that plant foods did not have a smaller role in the diet of the proto-Veddas than they have amongst their descendants. In our Vedda work we compiled a truly considerable but certainly incomplete list of the plants which serve partly as food and partly for enjoyment. Fruits, roots, leaves, bark and fungus are used in similar quantities. Itis clear that these perishable materials need not be expected in the cave floor. Nonetheless we had the ‘g00d fortune to find such remains. At one precisely recorded place, at a depth of 80 cm, we came upon a large fossilized supply of calcined or encrusted wood. These were very flat pieces of 4 to 12 cm ength, and 3 to 5 om breadth, but generally 4 cm thick or perhaps a little more. The ends of these pieces showed very straight cut surfaces, apparently made with stone knives. Some of these pieces of fossil wood are illustrated in Plate X, Figures 264 to 267. When viewed from face on, one face appears smooth and fine-grained, the other rough and corrugated. With a long and intensive treatment with acid, they would completely dissolve. Thin sections viewed under the microscope immediately reveal their nature as plant, and indeed as pieces of bark, with the rough surface being the outer bark, and the smooth surface the inner bark. [page 87] As a pastime, the Vedda chew the bark of various trees and bushes for its aromatic or astringent qualities; the inner bark of the wild mango tree is eaten for nourishment. It is hard to say as to which these fossil remains belong. In our collection of Vedda food and pleasure plants there are pieces of tree bark, cut to shape, which accord well with our fossil pieces. These have the remark: “Vedda food, Sinhalese Dawata, the Veddas chew its bark”. This Dawara in our list (page 402) is Carallia integerrima D.C. (Rhizophoraceae), whose bark is full of tannin, astringent and tasty. This could well be the correct identification, even though the outer bark of the fossil pieces is somewhat thicker and more corrugated than the Carailia bark in our collection. In our plate, the recent piece (Fig. 238) is illustrated next to the fossil pieces (Figs 264 to 267), to show their similar cut to shape (with a stone knife for the fossil pieces but a steel axe for the recent piece). Whether the fossil bark belongs to this Carallia or some other tree, this find tells us that the proto-Vedda must have chewed bark. For the preservation of these pieces in the cave floor we can thank the occasional seepage of calcium-tich water, as also shown on the main bones with a thick encrustation, We found no traces of cultivated plants, no indications of coconut in the cave, whose absence from the Toala shelters of the Celebes — apart from the overlying layer — was also noted (27, page 51). The cultivation of coconuts was evidently brought to Ceylon by a group with a higher culture, as was also the case in the Celebes. 54Summary of the Animal Remains in the Cave We place here a collected list of the represented types of animals: Semnopithecus priamus Blyth Sus cristatus Wagn. Macacus pileatus (Shaw) Manis pentadactyla L. Canis familiaris L. (aureus L.2) Emyda vittata Peters Melursus ursinus (Shaw)? Nicoria trijuga thermalis (Less.) Hystrix leucura Sykes Varanus bengalensis (Daud.) Sciurus macrurus Pennant Helix (Acavus) phoenix Pir. Lepus nigricollis F. Cuv. Cyclophorus (Litostylus) ceylanicus (Pfr.) var. Cervis axis Erxl, Paludomus (Tanalia) loricata Rve Cervis unicolor Bechst. Paludomus (Tanalia) neritoides Rve Cervulus mutjac (Zimm.) Unio (Lamellidens) thwaitesii Lea Buffalus bubatus (L.) Unio (Parreysia) corrugatus Lea [page 88] The most important result to emerge from the faunal list is, without a doubt, that we have a completely modem fauna. All of the represented mammal and reptile forms still inhabit the vicinity of Nilgala (and presumably so for the molluscs, but we do not know for sure), which means that the remains of hunting undertaken today would bequeath the same picture as the cave finds. This indicates arelatively young age for the artefacts found lying with the animal bones. This is an unusually strong support, and to our minds a proof, that the bearers of the Ceylon Palaeolithic were in fact the direct ancestors of the Vedda living today. Had we encountered a layer with older, extinct fauna with the Stone Age industry in the caves, and accordingly pushed back the age estimate, then we would not have had the courage to declare the proto-Vedda status of the inhabitants. But since all of the animal forms present a recent character, so it would be highly unnatural to look for another people as the ancestors of the Veddas who even today live in the same forests with the same animal ecology. ‘The results from the Toala caves of Lamoncong were quite similar. Here also the stone artefacts were accompanied by a recent animal ecology lacking any extinct forms. However, in contrast to the Nilgala situation, the faunal collection from Lamoncong illustrated a few differences, in that the babirusa and anoa, which occurred abundantly in the cave material, had disappeared from the vicinity of Lamoncong, the former evidently from the entire peninsula, and the anoa only from the immediate surroundings - whilst still being an important animal in the higher mountain forests (see 27, pages 52 ff.) In contrast, exactly the same animals are found in the Nilgala district whose remains were found in the cave. It would not however be correct to infer a younger age for the Stone Age of Ceylon, because the change in the animal status of Lamoncong can be ascribed solely and entirely to the relatively modem influence of the Malay culture bearers, who have cleared the forest across a large part of South Sulawesi, whereas the wild districts of Ceylon have never been affected in such an intensive way. We now compare our animal glossary with the list of edible foods of the currently living Veddas (26, page 412). We can see that some forms are lacking almost certainly as the fortuitous result of not having come into our hands. Examples include (1) fruit bats, especially Preropus medius Temm., which overall is the most common fruit bat and with flesh very suitable for consumption, and (2) the mouse deer, Tragalus meminna [page 89] Exxl., which is very much hunted for its very prized, fully tasty flesh, On the other hand the absence of elephants could hardly be fortuitous, in view of their size. Admittedly the Veddas of today do not hunt elephants, but it seems that, as we found out at various places, in earlier times they had long obtained ivory to trade to the king of Kandy as tribute, so elephants had 55been hunted for barter but not for their flesh. The ivory was used for making arrows with very long shafts and an extraordinarily long and broad iron blade. The meagre stone weapons would obviously not have been adequate for slaying elephants, for even though the exchange of ivory with a higher culture would not have been of much worth to Stone Age hunters, nonetheless the flesh of younger animals (despite its toughness) would have been put to good use in the cuisine of primitive people, as with the mammoth of the European Palaeolithic. In conclusion, everything from the cave debris suggests a picture of pure foragers, as the Veddas were until a short time ago. The probable presence of a dog is the only indication of domesticated animals, but itis not certain, and there are no signs of any possession of domesticated plants. [page 90] The Human Remains from the Nilgala Cave. F.S. Plate IX, Figures 240 — 242 Unfortunately the yield of human remains from the Nilgala cave is overall very sparse, and the other sites had none at all. This is particularly bothersome as we had been hoping to be able to reconstruct a physical picture of the Palaeolithic cave dwellers. The recovered fragments in no way reach that, objective; the most we can say is that our view, that the Stone Age people were the ancestors of the current Vedda, is not contradicted. There are grounds for why this negative result could have been anticipated, in view of the burial practices of the Veddas. In our Vedda work we write on this topic as follows (26, page 492): “Itis a very strange fact that originally the wild Veddas simply left a corpse to lie where the person happened to die, without worrying themselves about it any further. The only further thing they did was to cover it with twigs or leaves, and now and then, but not always, lay a heavy stone on the breast of the corpse. The place where it lay was abandoned and visited as little as possible until the corpse had fully decomposed. During the rainy season a death naturally occurred mostly in a cave, because this was the time of the year the Veddas used caves as living space, and in these case the cave would be deserted, so the decomposition of the abandoned skeleton would then have happened in the cave.” The last remark needs an addition. Theoretically it is quite correct, but practically the event would have played itself out differently. It is decidedly clear that a corpse left freely lying in a cave must have been ruthlessly set upon by all of the animal scavengers which are definitely not lacking in Ceylon. Only in the most exceptional cases could anything have been preserved other than a few sparse, broken pieces of a skeleton, most of all teeth, {page 91] Vedda skulls and bones, allegedly found in caves, play a peculiar role in the literature. We ourselves have never seen them in the many caves we have searched, and the above reflection makes it very probable that in most or all cases these skulls did not come from caves but from re-opened graves ~ burial is now generally practised everywhere. So when a concerned collector alleges “found in a cave” out of shame or discomfort, this would better read “taken from a re-opened grave”. If we therefore accept, as we may readily do, that the Stone Age cave dwellers had not shown their deceased any more honour than their descendants have done until recently, but had simply left the corpse of their relative to lie where the death happened and avoided the place for a while, then the sparse recovery of human remains from the cave floors can be comprehended. Four different individuals are represented amongst the human remains in the Nilgala cave, of which two have yielded only a single tooth or small pieces of jaw. 1. A bone fragment from the left upper jaw of a child, with milk teeth, is depicted in both anterior and posterior views in Plate IX (Figures 241 a and b). It is 25 mm long, ca. 15 mm high, and bears a 56deciduous molar, a canine and both incisors, of which however the second is broken off at the root. This alveolar fragment shows a distinct slant of the teeth (prodonty), which our illustrations do not exhibit sufficiently clearly. Some broken fragments of a thin, child’s vault belong with reasonable certainty with this small jaw. The largest such fragment, a frontal fragment, is ca. 7 cm long and 4 wide; another small fragment shows a simple sagittal suture running for a distance of 3 om. The form of these fragments indicates intentional smashing by humans; the latter appears to have been achieved through a falling stone, because one small piece shows a mosaic of cracks and is pressed flat instead of bulging out. The latter condition would have prevailed if the skull had still been in pristine condition. 2. A grown individual with very wor teeth is represented by several bones, especially through some lower left teeth held together by an encrustation, being the first and second incisors and the canine. Additional fragments that belong here are the broken crown of an upper first incisor, and finally the right upper second incisor and canine bound together. 3. The remains of a further mature individual represent a chronologically younger person than the previous one, in that a crust covering is almost absent. Four vault fragments of 4-6 cm length and 3-4 cm breadth were found. Their thickness (up to 9 mm) documents a powerful skull construction, male without a doubt, as [page 92] only our strongest Vedda skulls show. The lower jaw, which has exactly the colour and preservation status of the cranial fragments, is represented by a small piece from the left side, 31 mm high and ca. 25 mm wide, in which the roots (without their crowns) of a canine, incisor and first premolar are inserted. The inner and outer sides are illustrated in Figures 242 a and b. Upon inspection, the alveolar part shows a strong prodonty. A broken molar with a strongly wom crown probably also belongs to this person, 4. An isolated incisor of a mature person has an almost unworn crown and so cannot be assigned to the other skeletal remains of adults. Therefore it must have belonged to a fourth corpse. Some fragments from long bones were also found, and with these it cannot be said whether they belonged to the three individuals represented by their teeth or else are the remains of additional skeletons. A small piece of a femur was so heavily encrusted that it could only be freed through acid treatment, In addition, Fig. 240 illustrates the articulating head of the first phalanx of a large right toe, which also had been coated with crust, but (fortunately) an easily removed coat. There are also some strongly degraded, almost worthless bone fragments. The four individuals whose remains are described here presumably do not exhaust the number of people who had died in the Nilgala cave. It would be far more acceptable that many more people had found their last resting place here, but that their remains have not survived. The interesting result from these finds appears to us to be that these Stone Age troglodytes apparently employed the same method of treating the dead, namely the simple abandonment of the corpse, as did their later descendants. Had the corpses been buried in graves, then much more of them would have been preserved than the sparse fragments we found. This allows us to draw the further conclusion that they had the same primitive attitudes to life and death [as the recent Veddas]. Of no less importance is the point that cannibalism was apparently unknown to the Stone Age people, for in that case we would have found burnt human remains along with the bones of the prey animals, which is not the case. This observation additionally connects the Stone Age denizens of Ceylon to the Veddas of today, who also have known nothing of cannibalism as far back as their history reaches. [page 93] Cited Literature (for the sections on the artefacts of faunal material, the animal and plant remains from the caves, and the human remains from the Nilgala cave) 3710, 11 12. 13. 14. 15. 16. 7. 18. 19. 20. 21. 22. 23. 24, 25, 26. 27. 28. 29. 30. 31. 32. 33. Anonymous 1823, The Veddas of Ceylon, official report, published by C. J. R. Le Mesurier, Journal of the Ceylon Branch of the Royal Asiatic Society, 9, 1883. Avebury, Lord, Prehistoric Times, sixth edition, Williams and Norgate, 1900. Bailey, J., An account of the wild tribes of the Veddas of Ceylon etc., Transactions of the Ethnological Society of London, 2, new series, London, 1863, Basedow, H., Anthropological notes on the western coastal tribes of the Northem Territory of South Australia, Transactions of the Royal Society of South Australia, 31, 1907. Blanford, W.T., The Fauna of British India, Mammalia, London, 1888 ~ 1891. Brot, A., Die Gattung Paludomus, Systematisches Conchylien-Cabinet von Martini und Chemnitz, 293. Lief., Erster Band, Heft 89, Niimberg, 1880. Flower, W.H. and Lydekker, R., An Introduction to the Study of Mammals living and extinct, London, 1891. Franks, A.W., On Signor S. M, D’Alberti’s travels in New Guinea, The Journal of the Anthropological Institute of Great Britain and Ireland, 6, 1877. Giinther Alb., The Reptiles of British India, London, 1864. Hanley, S. and Theobald, W., Conchologia Indica, London, 1876, Hough, W., Fire-making apparatus in the United States National Museum, Report of the National Museum 1887 — 1888, Washington, 1890, Kelaart, E.F., Prodromus Fauna Zeylanicae, Ceylon, 1852. Knox, Rob., An Historical Relation of the Island Ceylon in the East-Indies, London, 1681. Kobelt, W., Cyclophoridae, das Tierreich, 16. Lieferung, Berlin, 1902. Layard, Edg., Observations on the genus Paludomus of Swainson etc., Proceedings of the Zoological Society of London, 22, 1854, Lea, J., Descriptions of exotic Unionidae, Journal of the Academy of Natural Sciences of Philadelphia, Ser. 2, 4, 1860. Man, E.H., On the aboriginal inhabitants of the Andaman Islands, The Journal of the Anthropological Institute of Great Britain and Ireland, 12, 1883. Martens, E. von, Uber verschiedene Vewendungen von Conchylien, Zeitschrift fir Ethnologie, 4, 1872, Berlin. ‘Nevill, H., Arrow heads and spoons of Unio shell, The Taprobanian, 1, Bombay, 1887. ~ Veddas of Ceylon, The Taprobanian, 1, Bombay, 1887. ~ Virchow on the Veddas of Ceylon, The Taprobanian, 3, Bombay, 1888. ~ The bison or gaur in Ceylon, The Taprobanian, 3, Bombay, 1888. Piette, E. and I. de la Porterie, Etudes d’Ethnographie Préhistorique, V, Fouilles a Brassenpouy, 1897, L ‘Anthropologie, 9, 1898, page 531. Roth, H. Ling, The Aborigines of Tasmania, Halifax, 1899. Roth, Walter E., North Queensland Ethnography: Bulletin nr. 7, Aug. 1904, Domestic Implements, Arts, and Manufactures, Brisbane (Home Secretary’s Department), 1904, Bull. Nr. 8. Nov. 1905, Notes on Government, Morals, and Crime, ibid., 1906. Sarasin, P. and F., Die Weddas von Ceylon und die sie umgebenden Vélkerschafien etc., Ergebnisse naturwissenschaftlicher Forschungen auf Ceylon, Ill, mit Atlas, Wiesbaden, 1892- 1893. ~- Versuch einer Anthropologie der Insel Celebes, erster Teil: die Todla-Héhlen von Lamontjong, Materialien zur Naturgeschichte der Insel Celebes, 5, 1, 1905. - Zweiter Teil, die Varietiiten des Menschen auf Celebes (F-S.), ibid. 5, 11, 1906. Simpson, Ch. T., Synopsis of the Naiades, or pearly fresh-water mussels, Proceedings of the United States National Museum, 22, Washington, 1900. Smyth, R. Brough, The Aborigines of Victoria, 2 vol., London, 1878. Spencer, B., and Gillen, E.J., The Native Tribes of Central Australia, London, 1899. Steinen von den K., Unter den Naturvilkem Zentral-Brasiliens, 2, Auflage, Volksausgabe, Berlin, 1897. Sterndale, R.A., Natural History of the Mammalia of India and Ceylon, Calcutta, 1884. 3834. Stevens, C.S.V., Amongst the Veddas, Royal Asiatic Society, Ceylon Branch, Proceedings 1886, Colombo, 1888. 35. Studer, Th., Beitrige zur Geschichte unserer Hunderassen, Naturwissenschafiliche Wochenschrifi, 12, 1897. 36. Tennent, Sir J. Emerson, Ceylon, An Account of the Island, etc., vol. 4, ed., London, 1860. 37. ---- Sketches of the Natural History of Ceylon, London, 1861. [The monograph concludes with Plates I to X] 59