Passing over Kosloff's long description of the migration of the Kontsche-darja towards the south-west, I will merely cite here the following passage: »According to the natives, some 9 to 15 years ago a breach occurred almost exactly midway in the Kuntschekisch-tarim, where its banks were low, and possessed of but little power of resistance; the water poured out and went back to the Ilek, filled the sandy, saline basin of the Tschivilik, revivified the lake-like, dried-up valley of the Ilek, which the natives began to call the Avullu-kul, Kara-kul, Sogot (Arka-kul), and so on.» But in supposing that these names originated at that late date Kosloff is making a mistake. As early as 1722 we find the names Kara-kul, Sadak-tu, and Kara-khodscho, no doubt the names of Turkish tribes dwelling at Lop-nor, but since that time, or even then, they were also the names of lakes. In 1876-77 Prschevalskij divided the people into Kara-kultsi and Kara-kurtschintsi, and adds the, explanatory note, »not far from this village (Ak-tarma), on the opposite side of the Tarim, lies Lake Kara-kul, which has given its name to the inhabitants of the Tarim valley.»
Kosloff has followed Pjevtsoff in confounding the eastern waterway with the inner waterway which belongs to the immediate neighbourhood of the Tarim. His lake Sogot, which is essentially apocryphic, i. e. a pure marginal lake (see vol. I fig. 171), is at any rate not identical with the Arka-köl. He says that »the shores of the Sogot are inhabited all the year round ;» but it is only sometimes in spring that the shores of Arka-köl are visited by fishermen and collectors of wild-ducks' eggs, but it is quite an exception if they remain longer than one night. My map of this region will show that Kosloff's Sogot is neither the actual Sogot, which has been dry for several years, nor the Arka-köl, which lies somewhat farther east, but is identical with the Tosghak-tschantschdi. He states that from the Sogot in question there issues an arm Ilek, which empties itself into the Tarim 20 versts below Ajrilghan; perhaps he means by this the Almontschuk-kok-ala. The real Ilek, on the other hand, which issues from the Arka-köl, joins the Tarim at Schirge-tschapghan, almost 5o versts below Ajrilghan. Owing to the misleading information which Kosloff has given, the lower part of the eastern waterway on the map of the Russian General Staff (sheet Hami, 1899) is so confused that it is like a species of bastard between my lake route and the marginal lakes on the east side of the Tarim. And yet I had exhibited to the Imperial Russian Geographical Society a large map of the entire region, and in my book which came out at London in 1898, and at St. Petersburg in 1899, I gave a description of my journey along the lakes in question. But on the Russian map the lowest part of the Ilek, which joins the Tarim at Schirge-tschapghan, is not plotted at all, nor even so much as indicated by name. The compilers of the map have preferred the hearsay evidence of Pjevtsoff and Kosloff, given to them by the natives (for neither of them was ever there himself) to that of the map which I made on the spot.
When describing the Kuruk-darja, I alluded to Kosloff's view with regard to the migration of the Kontsche-darja. According to him it was the Kontsche-darja only which flowed in the now dry bed, whereas the Tarim continued in the same bed that it now follows down to Lop-nor (i e. Kara-koschun). »In the course of time the river inclined more and more towards the south,» until eventually it formed
IIIFカラー高解像度
