to the east, and soon found the track marked by the footprints of the small advance party which I had sent ahead two days previously under the guidance of Kasim. He had orders to dig wells at all places suitable for camps, and after reaching the ruins of Dandan-Uiliq to push on to the Keriya Darya, whence Ram Singh was to join me.
Marching in the drift sand was slow work, though the dunes were low, rising only to 6 to 10 feet in the area crossed during the first two days. The feet of men and animals sank deep at every step into the fine sand, and the progress of the heavily laden camels was reduced to about 1i miles per hour. In view of the want of sufficient fodder and water, it was essential to save them . all over-exertion ; hence I soon found that the direct distance covered by a day's march could rarely exceed 9 to 10 miles. The tamarisk and ` Kumush ' scrub which was plentiful at first, grew rare in the course of our second march, while the wild poplars or
` Togliraks ' disappeared altogether as living trees. Luckily
amidst the bare dunes there rose at intervals small conical hillocks thickly covered with tamarisk scrub, the decayed roots of which supplied excellent fuel. Close to these hillocks there were usually to be found hollows scooped out of the loess soil, evidently by the erosive action of wind. These hollows, which reach down to at least 10 to 15 feet below the level of the little valleys separating the neighbouring sand dunes, offer of course the nearest approach to the sub-soil water. It was accordingly invariably in these depressions that Kasim's advance party had dug their wells, which we also chose for our camping places. The water, which was reached after digging to an average depth of 5 to 7 feet, was very bitter at the first two camps and scarcely fit for human consumption. But as we moved further away from the Khotan River it became comparatively sweet. I have no doubt that geology would furnish a satisfactory explanation for this observation, which was well known to my guides as generally applicable to these parts of the Taklamakan and has been noticed already by Dr. Hedin. The supply of water furnished by these wells was decidedly scanty for
IIIF Color HighRes
