370 THE LOP-NOR PROBLEM.
Bajin-gol flows towards the Hollusun-nor. If this stream only carried a sufficient volume of water, it would probably continue in the same direction as far as the neighbourhood of the Tadschinur-nor, and the Naidschin-gol and Tschulak-akkan (with the Utu-muren) would then become southern affluents, similar to the Chotandarja and the Kerija-darja. Something similar to this has no doubt existed at some period in the past, and probably the river system of Tsajdam possessed at that time, even as the Tarim system does now, a terminal lake, which possibly was just as restless and inconstant as the Lop-nor. On the other hand in the unique hydrographical system of the former we have, in the splitting up of the main stream into a number of smaller streams mutually independent of one another, evidence of a more advanced stage of development, and may therefore assume that this condition is typical of what will happen in the Tarim system at some epoch still very distant. The fate of the Naidschin-gol has already overtaken the Kerija-darja, though the latter, contrary to the former, has created no terminal lake, the reason being the distinct fall which the surface of East Turkestan has towards the north, so that the stream continues to flow on until it dies away in the sand.
The interior of the basin of Tsajdam is however far less known than the interior of East Turkestan; for whereas we are able to draw hypsometrical curves for the latter, we do not possess sufficient observed data to enable us to do this for the former, even approximately. And even in the case of East Turkestan there are restricted areas in which we fumble with a good deal of uncertainty, as, for instance, the desert of the eastern Takla-makan. Where it not that we have reason to suppose, that the Kerija-darja follows a chain of bajir-depressions running north-northeast, and that the Tschertschen-darja coincides with a wind-eroded hollow, we should be tempted to look for a natural swelling of the surface, however slight, between these two streams.
One difference between the basin of the Tarim and every other basin I have visited in the interior of Asia is that in the latter the absolutely deepest part generally admits of being easily pointed out, whereas in the former that is not the case. If we take, for instance, the trough-like valley of the Baghrasch-köl or any self-contained basin you please in Tibet, we should be able from the edge of the saucer-shaped depression to draw radical gradient lines which would cut the hypsometrical curves at right angles, and would all meet in one point, namely the absolutely lowest point. In the Tarim basin the absolutely lowest point we ascertained to be the 14.0 m. sounding I obtained at Markat, but at the same time we have also ascertained that this value is entirely ephemeral, being dependent partly upon the filling up of the basin with sand and sediment, partly upon wind-erosion in some other region. It is a necessary implicate of the very nature of the case, that, if the terminal lake is ambulatory, the absolutely deepest point must be ambulatory also, and the conception of gradient lines in the Desert of Lop is so uncertain and so transient that it would be a waste of labour even to attempt to draw them.
Of lakes that are disappearing, or have a'lready disappeared, there exists no lack in Central Asia. On the other hand desiccated depressions are extraordinarily common, especially in Tibet, where, later on in this present work, we shall have an opportunity to become acquainted with several of them. Obrutscheff visited two of
IIIFカラー高解像度
