296 THE LOP-NOR PROBLEM.
For my own part I do not believe that the presence of this lake on the map is due to a duplication of the kind suggested, but that it, like the other lakes in the locality, actually existed, and had existed from time immemorial, just as well as it existed in the beginning of the i 8th century, and exists to-day in the beginning of the loth. In common with the entire region south of Lop-nor, it is placed too far to the north; to find its true position it must be transposed, like the Khas-nur, to the southeast, though to barely half the distance that this latter lake has to go. That would bring it into the position where the Kara-buran now is. It is indeed true, that this lake has shrunk immensely since i 877, and that not even the proximity of the Tarim, and its passage through it, are able to maintain its vitality. But we have also seen, that the Tarim is on the point of separating itself altogether from the lake by embanking itself within fresh ramparts or banks. The Kara-buran would disappear entirely were it not for the Tschertschen-darja. In i 899, relying upon Roborovskij's statement, that there was an old bed of the Tschertschen-darja in the desert about 65 versts to the north of the existing bed, I formed a very different opinion with regard to this stream. But after having myself sought in vain for the old riverbed in question, and having therefore no reason to believe in its existence, I must entirely change my former opinion. I believe, then, that the Tschertschen-darja, setting aside the usual slight deviations within very narrow limits, has on the whole always had the same position that it now occupies. It has always issued into a lake situated in the vicinity of the Kara-buran, which possibly has also from time to time shifted its position, as the Lop-nor has done, though to a far less extent. If I am right in my supposition, this part of the map dates from a time antecedent to the year 1740, for it does not show the Tarim issuing into the lake. This must therefore have been of very small extent, and would only contain water after the spring flood came down. The presence of the conventional sign for a mountain placed on its eastern shore means nothing, for similar marks are put in parts of the desert where not the smallest chip of a rock can be discovered. Four watercourses, corresponding to the deltaic arms of the Tschertschen-darja, enter the lake from the west and the south. The man who first sketched this lake can never have penetrated farther to the south-west, and consequently had no idea that those four streams were the arms of a large river delta. Hence between the Lop-nor region and Kerija the map shows an extensive patch of white. On the other hand, Öndörtü and Dschiseken cannot be anything else but the head-streams of the Tschertschen-darja, and, remarkably enough, they are mapped in the right positions. Hence, while the map shows us the beginning and termination of the Tschertschen-darja, all the middle part of the river is wanting. Accordingly the following statement of Kosloff is only in part correct: »It is clear that the Kara-buran, equally with the Tschertschen-darja, which empties into it, and which is about 600 versts long, was not known to the compilers of the Chinese map; nor did they know anything about the oases of Tschertschen and Nija either.»
The resemblance between the terminal lake of the Tschertschen-darja and the Khas-nur is only accidental. The latter lake also receives from the south-west three short affluents; and at the present day several small torrents, fed by springs, enter this lake from the south-west, as I shall describe in vol. III.
IIIFカラー高解像度
