pay 30 tongas (the usual charge being from 16 to 20), and soon after one o'clock I was on my way towards the desert. This route was well known to European travellers, and it seemed unlikely that any uncertainty existed as to the correct geographical position
- of the villages along the road. But in the evening of the first day after my departure from Yarkand, time hung heavy on my hands, and, as two of the chronometers were going very steadily, I set about taking astronomical observations to check the position. The result of my work differed from that of previous travellers as indicated on the maps, and I thought it well to continue the observations, checking the longitude at each halting place.
At Karghalik, after some little difficulty and delay, three camels were obtained from a neighbouring village, and these joined me a few days later at the oasis of Guma. From this fertile spot a portion of the caravan accompanied me one day's march to the oasis called Kara Targaz, where we found many large trees in full foliage. Not far beyond this, we entered on the famous Takla Makan Desert, the appearance of which surprised me. The landscape was quite green with a profusion of jilgan, a species of tamarisk, and trees, which seemed as if bursting into leaf after a long drought. The large sand-dunes, some nearly 20 feet high, were covered with jilgan and invariably surrounded with a ditch, narrow and shallow but very sharply defined. Among these sand-dunes I and my companions, Dalbir Rai, Raju, Islam Akun, two local guides, and the camel man with the camels, slowly wended our way. So slowly did we move, and in such a winding course, among the dunes and ridges, that we were at no great distance from Kara Targaz when we found it necessary to bivouac for the night. We used up the water in two of our portable tanks, two of the camels having drunk freely from a
IIIF Color HighRes
