| | | | | | valley which, because of its incessant winds and storms, I called subsequently the Valley of the Winds. Our first stage was one of 35 versts without water across a perfectly barren plain covered with sand, loess, and fine gravel, the rise being 400 m. We started in the afternoon, rested during the night, and continued again early the following morning as far as the locality in which the Sajsan-sajtu, * in consequence of the greater steepness of the surface, disappears underground, though it emerges again in the form of springs at Tschong-jar and the salt marsh of Ghas. These springs give rise to rivulets, which, after they unite, run out into the southwest corner of the lake of Ghas. The river Sajsan-sajtu comes down from the glaciers on the southern face of Mount Kreml, separates the Moscow Chain from the Columbus and Tsajdam Chains, and then wheels east into the Valley of the Winds, shortly after which it disappears into the ground. After flowing 20 versts underground it again comes to light in the form of a multitude of springs, which eventually unite and give rise to a fair-sized river; though farther down this once more disappears underground. We measured the breadth of the stream on the ice, and found it between 40 and 5o m. ; the ice was 2 I%2 feet deep, the water under it only I I/2 feet deep. The valley through which the river flowed had a breadth of one or two versts. Its surface consisted of loess and sand, brought thither by the winds. The grazing was in general good. On each side of the valley reaching to the foot of the mountains on the north as well as to the foot of those on the south is a barren stretch with a tolerably steep slope. We now changed from a westward to a south-westward course, and then turned south where the Sajsan-sajtu breaks through the Tibetan border mountains.** Into it we then plunged, our intention being, first to visit the Tibetan plateau and then to return to the Valley of the Winds, which we now left on our right hand extending a long way towards the west. Here the main range of the middle Kwen-lun is a good deal lower and its outlines less wild and inaccessible. The glen of the Sajsan-sajtu, which was very easy even for camels, serves as the boundary between the Tsajdam (= Tschimen-tagh and Piaslik-tagh) and Moscow (= Atschik-kol) Chains. These border-ranges are almost entirely barren, and in great part covered with loess deposits. The eastern half of the Valley of the Winds, in which we were now travelling, is bounded on the north by the Tschamentagh (= Ilve-tschimen and Akato-tagh), about which I had heard speak whilst still in the neighbourhood of Lop-nor. This range has an east-west strike, is I oo versts long, and is connected at the one end with the Altin-tagh and at the other with the Nameless Range that lies north of Ghas. It attains a considerable altitude, rising in three places above the limits of perpetual snow. We were told that on its northern face a river originates, which, after breaking through the Altin-tagh, empties into the Lop-nor. This is probably the Tscharklik-darja. This mountain-range, like the adjacent valley, is absolutely barren. There is an entire absence of springs and brooks. The mountain slopes are very steep and in the upper regions are strewn with drift-sand. The transverse glens too are steep, confined, and unfertile. South of the Valley of the Winds and of north-western Tsajdam is the stretch of mount- |
IIIF Color HighRes
