Sec. iJ TO THE TARIM AND THE DELTA OF THE CHARCHAN-DARYA 451
necessarily reduced the chances of finding relics even of nomadic existence or passing traffic. Yet
that this riverine area was visited during the historical period of Lou-lan occupation, probably as
jungle grazing, was proved by some of the finds made there. Thus, close to our Camp 126 a well-
preserved Wu-chu coin was picked up, and some nine miles beyond it the fragment of a Chinese
bronze mirror, decorated in relievo, C. 126. 002. This was the last object in metal found, excepting
the fragment of a bronze spear-point, C. 128. ooi, which turned up unexpectedly some eight miles
beyond Camp 128, long after the last remains of ancient tree-growth had been left behind, about
four miles from Camp 127. It may have been brought there by some hunter who had strayed into
the desert after game ; for that ground must have been wholly devoid of vegetation even in the
earliest historical period which at present is accessible to us.
In prehistoric times, however, human occupation had evidently extended beyond the limits of Prehistoric the area containing the dried-up river-beds which were still traceable ; for stone implements of the remains. same primitive type as those found on our march to the Lou-lan Site, and fragments of coarse
pottery, probably neolithic, cropped up at rare intervals, not merely in the deltaic area but also on
the desolate ground passed between Camps 127 and 130. Yet, judging from the exceeding scantiness
and complete decay of such wood débris as we could find there, this ground must already have
been an absolute waste in the early historical period. Specimens of these Stone Age remains found
between Camps 125 and 129 have been described in the list above.'a A well-preserved celt, C. 126.
001, and the `blades' in chalcedony and jasper, C. 127-128. 002, 003, are reproduced in Plate XXX.
As soon as the last riverine belt of dead Toghraks was passed, about four miles from Camp 127, North-south big ridges of piled-up dunes, or Dawans, were encountered in more and more frequent succession, Dawans. direction of and bare patches of eroded ground grew rare, even in the broad sandy valleys between them. But,
these huge accumulations of drift-sand left little chance of discovering relics of prehistoric occupation, yet they, too, had their quasi-antiquarian interest. They stretched invariably from north to south, and previous experience, gained from the rivers which lose themselves in the Taklamakan, had taught me to recognize the significance of this regular bearing. Such ` Dawâns' are always found running parallel to river-courses where these penetrate into areas of drift-sand. Under the action of physical causes, which need not be set forth in detail here, these sand ridges conform in their bearing to the direction of the barrier which the river presents to the movement of dunes, and to which they indirectly owe their origin.2 It was easy to see that, far away as we still were from the Tarim, it was its course, running roughly from north to south in this section between Tikkenlik and the Charchan River junction, which determined the bearing of the Dawans we encountered in steadily growing heights.
The observation is worthy of record here because it is relevant to the question which certain views set forth by Dr. Hedin have raised, whether the present course of the Tarim south of Tikkenlik is an old one or dates only from the time when the Kuruk-darya ceased to carry water. This is not the place to discuss the question at length. But I may point out that the north-south bearing of the Dawans, which remained constant over the forty-six miles, in a straight line, of our
la Cf. above, pp. 367 sq.
2 For such riverine ` Dawans', always crossed at right angles where my route led from one river-course towards another in the Taklamakan, cf. Ancient Khotan, i. pp. 309, 418 sqq., 444, 453, 483 ; also above, p. 241, and below, p. 455. My map of the Khotan region accompanying Ancient Khotan graphically illustrates these observations.
Though the force which has built up these ' Dawans' of