Sec. Hi] ACROSS AN ERODED ANCIENT DELTA 357
slags, were also met with at intervals over most of the march. The conditions under which our desert crossing was made obliged me to keep the line of our route as straight as possible. Search to right or left which would have caused delay was thus practically excluded. That the number of finds would have been far greater if a wider belt of ground could have been searched was proved by the experience gained in 1914, when I followed a different route on my march to theLou-lan Site.4
In the List reproduced at the end of this section it has been found convenient to group and
describe together all finds of small objects made on the wind-eroded desert ground south of the Lou-lan ruins. The description of the stone relics which form the great bulk of them is taken from
notes which Mr. Reginald A. Smith, Assistant in the British and Mediaeval Antiquities Depart-
ment of the British Museum, has been kind enough to furnish. As the same scholar has also made the collection of these stone remains the subject of a separate paper,& I may restrict myself here to
a brief analysis of the main results arrived at. The great majority of the pieces of stone, whether
worked or brought for use, are jasper. Chert, chalcedony, carnelian, and jade are also represented. All of them, it can be safely assumed, come from the south, where the K`un-lun range is known to
abound in these materials.° Of the 140 pieces collected, about half are worked with a varying degree of finish, the remainder being flakes and splinters struck off by man but offering no distinct proof of having been utilized.
From the fact that among the specimens recovered are three undoubted cores (C. 121-122. 002 ; C. 122.002, 006 a ; see Plate xxx), it is certain that at least a portion of the worked stones
must be of local manufacture. By far the most numerous among these are ` knife-blades ', nearly
sixty in fact, ` with single or double ridges showing that they were struck by people who understood the art of detaching regular two-edged flakes '. (See Plate xxx, C. 121. 0028, 0032, 0075 ; C. 122.
oo6, oo8, 0027, 0052 ; C. 122—I23. 009 ; C. 127 128.003.) Such blades' appear in the palaeolithic
period, but seem to have survived into neolithic times. Certain specimens illustrated by Fig. 25 in Mr. R. A. Smith's paper show one edge battered into a broad blunt surface for the forefinger to
rest on in use, a form first attested in the La Madeleine cave period. Undoubtedly neolithic are the well-finished jasper arrow-heads, C. 122. 0023, 0054 (Plate xxx), of which one was found on the way from Camp 121, and the two jade celts, C. 126. 001, L.A. 00145 (Plate xxx), both of which were picked up in the area adjoining the Lou-lan sites.
It is clear that the frequency of these finds, over a somewhat wide tract, definitely proves its having been occupied by men in prehistoric times. But the same physical factor, wind-erosion,
which, on the one hand, allowed us to pick up these relics of the Stone Age with such ease on the
surface, renders it difficult, on the other, to draw from them any definite conclusion as to the chronology of that human. occupation. A number of interesting questions which these finds at
once suggested could not be answered from their own evidence. Did these remains of the Stone Age belong to a single if, perhaps, protracted period, or was it only the erosion of the successive layers originally containing them which had brought here relics of widely separated periods to lie side by side on the present surface ? Was it safe to assume that the difference in level between this surface and the top of the terraces protected by dead trees and drift-sand gave the measure of the extent to which wind-erosion had effected its work of scouring and lowering since the Stone Age ? Or had water continued to make its way here down to a much later period or, perhaps, returned after a long interval of desiccation and consequent denudation ?