738 EXPLORATIONS IN THE KURUK-TAGH [Chap. XX
the left bank and followed its winding course for some five miles to the neighbourhood of our camp. In spite of a close look-out, no trace was found of ancient occupation in the shape of pottery, timber debris, or the like. Nor did we, in the course of the next two days' marches, come upon any such extensive ` Tatis ' as one might have expected from a certain reference, in connexion with the Kuruk-daryâ, to ` prosperous villages [having] covered the country for scores of miles ',12
From our camp between L.S. and L.T. I decided on March II th to turn westwards, in order to trace the Kuruk-daryâ over a small portion of its course that had not been surveyed either by Dr. Hedin or by Lal Singh. The latter had closely followed the left bank of the ancient river-bed for over two marches eastwards before turning off to Altmish-bulak, and had found only one more ancient burial-place on the edge of the gravel Sai along which his route led almost all the way. This point (Map No. 29. c. 3) lay fully twenty miles away from our position, and the description given by Lâl Singh of his ` Cemetery No. 3 ' indicated that it was of just the same type as the two we had already searched. A visit to it would have cost us three days and involved considerable fatigue for the camels owing to the increasing warmth of the days. My increasing anxiety about Afrâz-gul and his party was an additional reason for forgoing this eastward extension of our search. His arrival at our proposed rendezvous at Ying-p`an was now several days overdue ; to move farther east entailed a risk of missing him when he passed along the foot of the Yardangbulak-tagh, as arranged in the instructions I had given him. The fires we had kept burning day and night on a high tamarisk-cone near C. ccxliii in the hope of eventually attracting his attention had failed to relieve my anxiety about the party. Yet both smoke in daytime and fire in the dark are visible for great distances in the desert under tolerably clear atmospheric conditions such as had fortunately favoured us during these days.
So on March i i th we set out to the south-west in order to see something of the country beyond the right bank of the river. After crossing ground eroded by the wind into low terraces and trenches all running north to south, we struck the bed again, and cutting across one of its many smaller meanders reached a point about three miles from camp where, at a sharp bend to the north-west, the sand at the bottom, some 15 feet below the bank, felt distinctly moist and there was abundance of living thorny scrub. But on the banks the beds of reeds and other vegetation were all dead. Even these quickly disappeared, as we passed into a maze of small Yardangs, from 4 to 8 feet in height, cut up into short strips and knolls as seen in Fig. 335. As we threaded our way across this curiously reticulated ground, drift-sand in the hollows became steadily more plentiful, and after about three miles of such going we reached the edge of an area completely covered by dunes. They soon rose to 20 or 3o feet in height, forming small Dawâns or ridges such as I had found on a much larger scale in 1906 on my way from the Lou-lan site to the Tarim. Whatever dead wood was to be found in the troughs between them was all bleached and shapeless owing to great age. Evidently the period when the riverine belt had here extended farther to the south lay very far back. From the top of the ridges no indication of other dead river-courses could be sighted. Their existence farther south, however, is suggested by the bearings of the old beds which both Afraz-gul and myself had come upon when traversing the dune-covered desert area to the west and south-west of the Lou-lan site.
The difficulty of crossing these close-set sand ridges with the camels had made us turn again to the north-west.13 There we came upon a small depression where the soil at the bottom felt moist, and close by picked up two small fragments of coarse pottery and a small lump of copper
12 Cf. Huntington, Pulse of Asia, p. 263. requires some correction. It extended farther south beyond
13 The line of our route here, as shown in Map No. 29. B. 3, the point marked by a Yardang symbol.