Sec. i] BY THE EASTERN COAST OF THE DRIED-UP SEA 317
halt at Camp cvii (Fig. 182) was rewarded by an interesting discovery, which otherwise would have escaped us.
Tokhta Akhûn on searching the much-eroded clay terraces to the north-east of our camp 5 came upon fragments of pottery on the steep slope of one of them. On proceeding with him to the spot I found a number of potsherds, C. cvii. 02-6, 08-9, together with part of a large iron buckle, C. cvii. 07 (Pl. XXIII), and charred fragments of wood lying among big boulders of clay just outside a shallow rock cavity. The latter might have once been larger and the boulders have fallen away from its projecting roof. In any case the cavity, recalling those often found at sacred sites in India where Hsüan-tsang speaks of ` rock dwellings ', had evidently served as a shelter for wayfarers. Among the pottery fragments, all of greyish and probably hand-made ware, there is one, C. cvii. o8 (Pl. XXIII), showing a raised moulding with a modified dog-tooth ornament which I had often noticed among potsherds of the Lou-lan Site (L.A.). The conclusion seems therefore justified that these modest remains go back to the period when Lou-lan was inhabited, and were left by travellers on the ancient route leading to it. Mention may here be also made of the small stones, C. cvi. 01-3, some of them, perhaps, worked, which were picked up on sandy soil on the way from Camp cvi. They had also probably been brought there by traffic, though of earlier date.
On the morning of March 6th, we started early in the face of a bitterly cold north-east wind, which blew all day. The atmosphere had considerably cleared, and we took our course towards a large detached Mesa which we sighted to the east. From its summit I hoped to secure a bearing on some feature on the southern side of the bay which would guide us to Kum-kuduk. The reed-beds thinned out as we proceeded, and tamarisks, too, grew rare. But on the gravelly soil that strange hard-trodden track, with the footprints of wild camels beside it, which we had met with again and again since reaching the northern shore of the bay, showed up here still more clearly. About two miles farther we came upon human footprints running across it ; on following them up to the foot of the plateau, we soon made sure that Lai Singh's party had camped there, as it turned out subsequently, on the night of the 4th. I was thus relieved of anxiety about my indefatigable surveying companion.
Half a mile beyond we reached the Mesa, which rose to a height of about a hundred feet and was coated with shôr to a line about fifteen feet above the level of the surrounding ground. From its summit I could descry a line of high Mesas rising in the direction of S. 150° E. above the quivering white haze which lay over the arm of the sea-bed. From their bearing I identified them with the large group of Mesas passed in 1907 to the west of Kum-kuduk, and, as the result showed, rightly. The appearance of white cliffs far away to the east-south-east, which I took as belonging to the plateau of Yantak-kuduk (Map No. 35. A. 4), and of other terraces to the south-west, those near Achchik-bulak, confirmed the location. But considering the great distance which still separated us from the southern side of the bay, it was clear that the visibility of the last two features could be due solely to refraction.
The line S.150° E. which we now struck from the foot of the Mesa took us first over soft salt-impregnated clayey soil for about a mile and a half. Then followed a strip of hard salt crust with small channels of briny water showing here and there within narrow fissures. Fortunately these did not prove difficult to get round. Patches of soft brown shôr were also encountered up to the seventh mile from camp. Thence for close on five miles we continued our march over a surface of hard corrugated salt. As, however, the twisted edges of the contracted salt-cakes did not rise