ON THE WAY TO LOP-NOR
SECTION 1.-CHARCHAN AND VASH-SHAHRI
MY aim, when on the i9th December I left the extremity of the Niya river, was to gain as quickly as possible the Lop Desert region, my main goal for the winter's operations and still far away to the north-east. For the first two marches I had to follow my old track of 1901 and 1906 to the Yâr-tungaz river ; to my surprise I noticed that where it crossed high ridges formed by dunes of coarse sand it still showed in places distinct traces of the footprints left behind by my camels. This observation was subsequently confirmed when I revisited the Limes west of Tunhuang.
From our old camp at Helya-beg (Map No. 19. c. 1) I crossed the Yâr-tungaz river, already hard frozen, and in order to shorten the journey to the Endere river struck eastwards across unexplored desert ground. The belts of dunes that I met on the first march beyond the Yâr-tungaz river proved unexpectedly low, none rising to more than about 5o feet in height. Yet they, too, all stretched in the north—south direction of the river, thus conforming to what I had observed elsewhere in the Taklamakân and Lop deserts, viz. that the big sand ridges or ` Dawâns ' always have their axis parallel to the main bearing of the nearest river-bed, whether still carrying water or dry.' Here I noticed that the same rule holds good also as regards the thin lines of tamarisk-cones on our route.
On the march beyond Camp LxvIII the sandy soil with anty scrub soon gave way to wide
stretches of bare clay covered with a hard salt crust, which forc is to steer a more southerly course
for the sake of the camels' feet. Patches with tamarisk bushes ving on flat ground showed that
the area had been subject at one time to occasional inundation. It this must have come from the
Endere river became clear when towards the end of the sect march we encountered a wide
shör-covered depression flanked by lines of high tamarisk-cones i ling from SE. to NW. Beyond it we came upon rows of Toghraks, and there we camped. But sc twelve miles had next morning still to be covered from Camp LXIX before we reached the Ende, iver ; as we did after striking the caravan track towards Charchan. The wide divergence o. at dry western bed from the deep-cut Endere river is of interest as illustrating the great lateral flexion to which the drainage descending over the piedmont gravel glacis of the K`un-lun, here pa ularly broad, may be subject at one period or another.
The observation is of interest as it helps to account for the di: 'ces that separate the main ruins of the Endere Site and the remains of Bilel-konghan, explored on T first and second journeys,2 from the present course of the Endere river. This, where the caraw track crosses it at Enderelangar, runs in a broad ` Yâr' cut into the loess to a depth of over a hi fired feet. Its course here, still almost clear of ice, had a width of about 3o yards ; and carried a ' 'urne of water which could not be less than too cubic feet per second. A still greater abundance c water during the summer