Sec. iii] TO ANCIENT GRAVEYARDS BY THE KURUK-DARYA 737
a surface crust of gravel permeated with shôr and extremely hard, the six graves examined contained only bare bones, in some cases not even complete skeletons. No coffins were found in any of them. From one grave in the centre I removed the skull of an adult, L.T. 03. In the same grave was also found a knob-like piece of perished wood, 6 inches long and 4 inches in diameter, which perhaps once belonged to a coarsely carved wooden figure like L.T. or. This latter figure had been picked up by Lai Singh almost on the surface when he passed here, and brought to me when he joined me at the Lou-lan station. It is in poor preservation and shows by its splintered surface in front that it had lain exposed for some time to the fierce summer heat of the desert. Comparison with the crude female figure carved in wood ' which an excellent photograph in Professor Huntington's Pulse of Asia shows set up ` at the head of a half-opened ancient grave of poplar posts on the edge of the zone of gravel, near the Kuruk Dariya ',10 makes it appear highly probable that it is the remains of that identical figure. This is suggested in particular by the loss of the front of the head in both figures and the curious convex representation of the lumbar portion. Professor Huntington's description of his journey from the Lou-lan site to Tikenlik 11 makes no reference to this burial-place. But as he followed the foot of the glacis above the ` Dry River ', which offered the easiest line of progress, the graves of L.T. were necessarily close to his line of march. It is very probable that a little digging at one of them brought to light this wooden figure, and that it was left behind at the spot, to be picked up eight years later by Lâl Singh. Its likeness to the figure L.Q. ii. of (Pl. xv) found by Afraz-gul at a burial-place in the vicinity of L.F. is striking. There can be little doubt that these graves, like those of the neighbouring site L.S., belong to approximately the same period as that indicated by the finds at L.F., and that the human remains contained in them were those of indigenous Lou-lan people. The absence here of coffins may, perhaps, be taken as pointing to this being the resting-place of humbler folk among the herdsmen who once frequented this now dead riverine belt.
In all the graves examined at L.T. the bodies showed the same condition of complete decay. This led me to stop the opening of the remaining graves and to use what was left of the day for a survey of the riverine ground to the south-east. Descending to it from the gravel Sai we first crossed a flat stretch of bare clay eroded into small Yardangs. Then after passing over a depression where the cracked surface of the clay indicated occasional flooding, perhaps by rain-water descending from the Sai, we reached the ancient river-bed plainly marked by thick rows of dead Toghraks on either bank. The banks rose about 25 feet above the bottom of the river-bed, here 93 yards wide at the point where we measured it. Its bottom along the deepest portion was covered with a crust of mud showing cracks manifestly recent, and digging down here for only four feet we struck water. That it was utterly salt was no cause for surprise. But it conclusively proved that water even now at times reaches this portion of the river-bed, if only on occasion of exceptional rain in the hills northward.
The belt of Toghrak jungle occupied banks of sandy soil rising about ro to 12 feet above the bare flat ground on either side. Judging from the thick beds of dead leaves which could be laid bare by a little digging, the age when this jungle flourished could not be very remote. Many of the trunks were small, as if there had not been time for the trees to reach their full size before the temporary return of water, which had permitted of their growth, had ceased again. Then we passed on the right bank over eroded ground, with low tamarisk-cones here and there, and then struck the river-bed once more at a sharp bend to the south. From high island-like terraces, protected by masses of big dead poplars, a wide view was obtained to the south. But we could see no sign of any other bed lying in front of the high dunes visible in the distance. Then we crossed back to
io See Pulse of Asia, plate facing p. 262. 11 Cf. Pulse of Asia, pp. 262 sq.