of trees that form the village boundary on the W, marks its position at a distance. Itis at about this place that an arm separates from the rest of the river and takes a more easterly direction, at first pressing against the outrunner of the Qara Teke and later flowing across the plain at an appreciable distance from the other arms of the river. A white ribbon of ice, marking its course, was visible for several hours from the high northern bank. Here the river diverges more and more from the Qara Teke mountains that appear to retreat further and further. The distance to the mountains in the N seems to diminish slightly. On the height of the Yamansu bashenda Akanbez mazar the northern bank rises and drops from a height of about 6 fathoms. It keeps at this height, which makes it quite inaccessible, especially as there are no crevices, for about half-an-hour. Removed from its frame of mountains in the S, the river spreads out here and is about i i /3 miles broad.
The large village of Yamansu with about 200 houses begins at the mazar, and is inhabited partly by Kirghiz, but mostly by Sarts. The fields are poor at first and the buildings isolated, but the latter become more numerous, though they never form any large group of houses. On the southern bank you see the villages of Atbashi, Chailung and Sugaetlik, forming almost a connected mass, if viewed from the northern bank. The bed of the river that narrows to 2/3 of a mile in the western part of Yamansu, grows wider again and is quite 1 113 miles wide in the eastern half of the village. The ledge of the bank which had kept at a height of about 4 fathoms, descends gradually. Only a modest arm of the river flows along it. The main mass flows in 2 or 3 arms along the flat southern bank. From the eastern part of the village of Kök-jigda the water channels have become so much shallower that the small horses of the population ford them even at high water.
February 25th. We continued our journey to-day further down the river in lovely spring weather.
Onylik village. From the end of the village of Yamansu miserable stony ground begins, called Qara Jas across the course of innumerable large and small rainwater channels, clefts, ravines etc. As soon as you leave the road you encounter obstacles at every step. In order to reach a spot only a few fathoms off, you are often obliged to make a détour of 200-30o fathoms. Riding up and down these clefts, all running in a direction diametrically opposed to ours, was exhausting for the animals and trying to our patience. The river flows here in 3 channels, connected here and there by small runnels. The river and the steep ledge of the bank turn almost due east here, but by degrees they return to their former direction and in about 30 minutes the bank goes once more in a definitely NNE direction. Here the river flows for about 2/3 of a mile in a single arm, but then again divides into two main and subsidiary arms. At a distance of 2 or 3 miles a band of ice was again visible parallel to our road, probably indicating the arm of the river that took such a pronounced E direction yesterday from the village of Kök-jigda. We were now approximately on a level with Uch Turfan, easily recognisable by the small mountain next to the town. The chains of mountains in the N and S, especially the latter, seemed to diverge very much from the course of the river, and the valley or plain on either side of it grew broader and broader. In the S its breadth must be about 5 miles and in the N it is at least as far to the mountains. The bed of the river had grown narrower and was not much more than 1/2 mile wide,
) 152(
IIIFカラー高解像度
