524 GENERAL HYDROGRAPHY OF THE TARIM SYSTEM.
The difficulty of drawing anything like a reliable boundary is greatest in the east. We do indeed know, and in vol. III we shall learn more fully, that a number of glens, some of which sometimes carry streams of water, cut their way through the Astin-tagh northwards to the lowlands, and the Kum-bulak is by no means the last of them; but it is difficult to draw the dividing-line between those that flow down to the Kara-koschun and those that go to the Chara-nor. And yet this is, after all, a matter of little moment, for the eastern extremity of the peripheral zone is for our immediate purpose une quantilé négligeable, in that it makes no contribution whatever to the maintenance of the terminal lake. I have not entered on the map the corresponding part of the Kuruk-tagh, that is to say the eastern extremity of the northern peripheral zone, for the simple reason that our knowledge of its geography is so very defective, it is not possible to lay down any sort of hydrographical boundaries at all. Let us assume however that the part which drains into the Lop depression, or speaking generally into the Tarim basin, is equally as large as the corresponding area on the south, or say 30,000 sq. km. This region again is a factor that may be disregarded entirely, for it does not send down a single stream from the Kuruk-tagh to the Kontsche-darja: the brook of the Suget-bulak, that of Kurbantschik, and that of the Budschentu-bulak all alike fail to reach the river.
Yet even though, as I have already pointed out, the figures in the table given above are not trustworthy in detail, still it is fair to assume that the sum of all the areas put together is not very far from the true area. The total area therefore of the region which drains into the Tarim basin is 446,000 sq. km. But that part of its basin from which the Tarim and its daughter streams do not receive a single drop of water possesses almost exactly the same area, namely 471,000 sq. km. Now not only this proportion between the contributory and the non-contributory parts of the basin, but also the morphological and hydrographical arrangement of the system as a whole, may be regarded as being in the highest degree unusual and peculiar. With regard to the symmetry of this conformation, there is scarce another hydrographical basin in the world that can compare with it. In Tsajdam, for instance, it would be hard to detect any trace of symmetry whatever. Some of the small basins of Northern Tibet are indeed fairly symmetrical, but their arrangement is very simple. Small torrents gather from north and south off the mountain-ranges and from east and west off low latitudinal thresholds or transverse ridges, and meet to form a salt lake — a type of basin that is in fact very common in Tibet, and which in point of orographical formation is repeated to a monotonous extent. In shape all these basins are elliptical, and in arrangement concentric throughout. In the Tarim basin the arrangement is however excentric, its deepest depression being situated a long way in the east; in fact it is so far to the east that it almost falls outside the boundaries of the regular ellipse. On the other hand the hydrographical arrangement is, as I have already said, exceptionally symmetrical. If the deepest part of the basin lay beyond the long axis of the ellipse, the symmetry would be ideal; but, situated as it is at the southern foot of the Tien-schan, it has occasioned a dislocation towards the north of the main stream of the system, though that in no wise disguises the striking lines of the symmetry, which are of course occasioned by the positions of the encircling mountains. If we compare the northern peri-
IIIFカラー高解像度
