the usual Tibetan character; south of it rises an alpine region, with lofty cliffs, steep and difficult to climb. At first these do not reach up to the snow-line, which runs there at an altitude of 5200 m. But with every ten or a dozen versts that you advance towards the south this alpine country assumes an increasingly wilder appearance, and ere long you catch a glimpse of the snow-capped mountain-peaks of Gatij-dschu. At the same time the chains acquire a meridional direction, and both vegetable and animal life increases. The rocks consist almost exclusively of schist. Out of every ravine issues a small swift torrent, which joins the Di-tschu, and in summer these carry a good deal of water. Higher up the river's course the relief is gentler, and the country merges gradually into the Tibetan plateau. The climate in this region is extraordinarily harsh, as it is throughout the whole of Tibet. According to the statements of the natives snow falls in winter to a considerable depth and the cold is intense. Even in spring it is cold and stormy ; in the summer it rains or snows every day, and even the autumn is disagreeable.»
As I shall in a later chapter attempt to summarize our existing knowledge about the orographical structure of the middle Kwen-lun, it will not be without interest to recall the opinion with regard to it which Prschevalskij held twenty years ago. I will therefore append here what he says about it; it will also serve as a specimen of the way in which he describes a complicated mountain-system. »The famous Kwen-lun, the 'backbone of Asia' as Richthofen calls it, was, before my last journey, perfectly unknown over a distance of I2°, counting from Naidschin-gol to the neighbourhood of the oasis of Kerija in East Turkestan. We now succeeded however in penetrating along this unknown part of the oldest mountain-range in Asia, and gleaning some light as to the topographical arrangement of the principal massive of the system. The main chain makes a curve, the two extremities of which lie on almost the same latitude, namely 36° N., whereas the tangent of the curve lies on 38° N. The western boundary of the Kwen-lun may be put at 87° E. from Greenwich, where the Altin-tagh breaks away from the main chain. On the other hand the middle Kwen-lun stretches eastwards, according to Baron Richthofen's investigations, as far as toe, after which it widens out to a broad system of parallel ranges. The most imposing of these, that is the Kwen-lun proper, serves, like the whole of the western part, as an immense border-wall to the highlands of Tibet, and in the east stretches to the desert and the salt marsh of Tsajdam. Farther east it is cleft by the upper course of the Yellow River, and finally dies away in the interior of eastern China. That this range really does constitute the Kwen-lun proper is evident from the fact, that it is the only one which is uninterruptedly connected with the main chains both east and west, the entire system thus stretching over more than 40° of longitude, whereas all the other parallel ranges terminate near the meridian of Lan-tscheo. My own explorations later on in this journey resulted in my discovering an immense snowy range which breaks away from the western Kwen-lun near the transverse glen of the Kerija river, so that this cannot be regarded as the main range of the system ; although it is reputed to extend for a whole month's travel in a south-easterly direction, and is perhaps connected with the Tan-la or with mountains north of the Tengri-nor, and thus ends before the true main range does, in that it leans upon the meridional chains
IIIF Color HighRes
