FACTORS TENDING TO DECREASE THE VOLUME OF THE TARIM. 355
that the Tokus-tarim breaks away from the main stream and makes its way independently into the Kara-koschun. It is evident, that a river which divides and subdivides in this way into a number of arms must lose a vast quantity of water, not only in consequence of the greater evaporation surface it thus presents, but also in consequence of the absorption into the ground which must inevitably take place in and through the existence of so many new channels. The alluvial deposits which the river lays down tend more and more to level up the Desert of Lop and make its surface more uniformly horizontal. And parallel with this marches the tendency which the Tarim shows to split up into an increasing number of arms. So long as its basin was less full of sediment than it is now, and consequently the gradients of its fall were more decided, the Tarim made its way to the Lop-nor by the steepest slope, that is the shortest »stream-line», and as a consequence confined itself to a single channel. At the present time it has attained a sort of intermediate phase in its existence; for, while it is divided into a great number of arms, it also follows at the same time a certain dominating direction of fall. The next phase will no doubt be, that it will break up into an even greater number of arms, which will all follow theiF own separate paths into the terminal lake. This development, which would reduce the river to an even yet greater degree, will only hold good so long as its lowermost course occupies its present position. For, if the quantities of alluvium which are deposited every year in the existing Tarim delta should eventually force that river, together with its tributary the Kontsche-darja, to return to the Kuruk-darja and the Lop-nor, it is very probable that it would do as it did before, and confine itself to a single main channel, and as a consequence of so doing would augment considerably in volume. And in the light of the knowledge we now possess, as to the relations of level that obtain in the Desert of Lop, it is not too bold a thing to say, that some time the river must go back to the Kuruk-darja. The country between Tikenlik and Tschigelik-uj is more exposed to the wind-driven drift-sand than the regions to the north and to the south of it. And I have already had occasion to call attention to the immense masses of sand which are carried down into the delta of the Tarim in consequence of the pressure which this river exercises upon the north-east face of the Desert of Tschertschen. It is merely a question of time, but the country hereabouts will become so choked with alluvia that the river will be forced to return to its northern bed. The lowermost limb of the river thus oscillates backwards and forwards like a pendulum, and even though the periodic time of each oscillation does amount to i oo years, yet that, counted by the clock of geologic time, is relatively of no longer duration than one of our seconds. The circumstance suggests a comparison with the pendulum-like oscillations of the lowest part of the Hwang-ho. No direct comparison can be instituted between the two streams, be-
' cause the Chinese river was forcibly constrained to abide in the course which it followed from 1290 to 1852, until in the latter year it burst its bounds and took the more southerly course, though shortly afterwards it was again forced back into its ancient bed farther to the north. In the case of the Tarim Nature is her own regulator, and the population is too small to prevent a catastrophe from happening, or when such does happen, to hinder it from being very disastrous. At the time of my last visit the river exhibited a decided tendency to follow a more easterly
IIIFカラー高解像度
