as that, its vitality upon reaching that point would be so small that it would have been utterly impossible for it to have scooped out a bed to such an energetic depth as is exhibited by the channel which we see in the desert to-day, and still less would it have been able to originate a lake to which the Chinese applied the name of the Great Salt Lake. Here again, then, we have a proof of the contention I have argued for above, e. g. p. 300, namely that the Kuruk-darja formerly carried the entire flood of the Tarim and that the Kontsche-darja was then, just as it is now, nothing more than a tributary of the Tarim. If now Kosloff's idea, that the Kontsche-darja alone formerly flowed through the bed of the Kuruk-darja, and then turned south through the desert to the Kara-koschun (Pl. 35), be compared with what I have just said above, the absurdity of his theory, which rests upon no sort of proof or support whatsoever, becomes even more patent than it was before. By that route the distance between Korla and the Kara-koschun is almost 500 km., and nobody will be so bold as to maintain that the Kontsche-darja, even in times of exceptionally high flood, ever possessed the power of preserving its vitality so far down as that, especially when its flood was drained away at Lôu-lan for irrigation purposes, a fate it now escapes.
By dwelling upon this matter I have not only been able to offer an illustration of the way in which the rivers of East Turkestan decrease in volume from the moment they issue upon the lowlands, but I have also been given another opportunity to prove the untenableness of Kosloff's assumption, that formerly the Tarim and the Kontsche-darja flowed by different routes to the Kara-koschun.
Resuming where I broke off. — The upper courses of the northern tributaries of the Tarim — the Ak-su-darja, the Tauschkan-darja, and the Kontsche-darja — belong to a part of Central Asia with which I am not familiar, and consequently I can only pass them over without saying anything about them. With regard to the streams that rise on the southern border of the basin — the Jarkent-darja, Chotandarja, Kerija-darja, the rivers of Kirk-saj, and the Tschertschen-darja — their upper courses belong to the Tibetan highlands, and I shall have an opportunity to discuss them when dealing with that part of Tibet in the fourth volume of this work. That region might conveniently be termed the »northern peripheral zone of Tibet», being a belt of country that belongs intimately to Tibet, as well as intimately to East Turkestan. To the former it belongs a priori, as forming a part of the stupendous mountain swelling, from which it is merely discriminated in so far as it supports the border-ranges of the Kwen-lun and drains into the Great Central Asian basin of the Tarim; and it is in virtue of this same property that it belongs just as intimately to the basin of the Tarim. To the north of this border zone such precipitation as does fall is excessively small, and under no circumstances does it make even the smallest contribution to the rivers of the Tarim system. If therefore the rainfall and snowfall of this border zone did not gather concentrically towards the central basin, this last would be a complete and continuous desert, without a single oasis. South of this peripheral zone stretches the self-contained basins of the Tibetan highlands, a veritable mosaic of basins, great and small, in countless number, most of them with a salt lake in the deepest part of its depression. The only portion of the pheripheral zone which is not backed by this self-contained drainage basin is that which lies in the
IIIFカラー高解像度
