west; but from the route we were following I was unable to determine whether the principal stream was destined to enter some self-contained lake on the plateau or whether it belongs to the system that pierces the northern range and feeds the Kum-köl. This present valley bore but little resemblance to the broad, open latitudinal valley south of the Arka-tagh that I traversed in 1896. In this latter no sooner did I pass one lake than another hove in sight; here on the contrary there were no lakes at all. By travelling from north to south as I was doing now I obtained a much clearer conception of the structure of the Arka-tagh than I obtained in i 896. The system turned out to consist, at any rate along this meridian, of an entire series of parallel ranges, and by no means of a single compact chain that sends out ramifications, especially towards the north.
The next thing was to cross over the mountain-range which borders the latitudinal valley on the south. To that end we forded the principal river, which did not carry more than one cub.m. in the second, the current flowing along a broad, shallow bed amongst reddish yellow mud; its altitude was 4879 m. So far this stream had flowed towards the west, but upon reaching the base of a large pyramidal dune it turned towards the south-west, and soon became lost to sight. We left it therefore on the right or west, as also a couple of crescentic pools of bright, intensely blue water. These, which had collected at the eastern foot of other crescentic dunes (fig. 65), contrasted in a striking way with the yellow muddy water of the stream. Next we traversed a peculiar crater-like sand-ridge, curving round some small pools. Similar formations were for a space quite common towards the south. The dunes in this new sandy area often stand quite separate one from another. The ground upon which they have been built up is hard and level, sometimes moist and with some gravel. After that we forded a larger brook, with about 3 cub.m. of reddish brown water. Turning off to the south-west, it eventually joins the former one. Higher up in the latitudinal valley this united stream appeared to become lost amongst the dunes, at any rate for short distances; but here, where we were, they again emerged. The latter, the larger stream, was heavily charged with glacial mud; consequently some of the water it contained could not have been filtered through the sand. On its banks, especially its right bank, there are several pretty big dunes, though the dunes of this drift-sand area are very much lower than those of Kumköl. This area, like that of Kum-köl, occupies only the southern half of the valley. How far it extends up the valley and how far down it I could not ascertain; at all events it continued westwards for as far as we were able to see. On the south it begins on the lowest slopes of the glaciated mass; but the space between the lowest moraine-faces and the first of the dunes is filled with a pretty broad belt of hills and knolls,
IIIFカラー高解像度
