356 THE LOP-NOR PROBLEM.
course; for the westernmost branch, the main stream of the Tarim, had become filled with mud and precipitated sand to such a degree that its channel was contracted to the merest ribbon, and in fact for long stretches was altogether abandoned. On the other hand the Laschin-darja and the eastern Ilek had very greatly increased in volume. This is no doubt a purely local oscillation, with a brief periodic time, but it involves the beginnings of a greater restlessness and instability. The Chinese did, it is true, talk about trying to put down dams and piles, and so forcing the river to keep within its old course, because its threatened desertion would be injurious to certain villages and communities, and fatal to their fields, all alike dependent upon its remaining true to its original course; but their labour would be wholly in vain, and would be rendered nugatory by Nature's own arrangements.
The reason of the river having deserted its ancient bed, the Kuruk-darja, and made a new path for itself south-eastwards through the Desert of Lop is of course the same as that already alluded to, namely the excessive accumulation of sediment.
The third factor which at the present day tends to weaken the Tarim is the creation of marginal lakes. I allude only to the lakes which occur below Karaul, for above that great bend the river is relatively more constant to its bed, and its marginal lakes are less changeable in both number and area, if not indeed in situation as well.
At the time of Prschevalskij's visits the bed of the Ilek and the basin of the eastern lakes did indeed exist, but they happened at that moment to be dry; accordingly we do not find any indication of them on Prschevalskij's maps, and the information afforded by the natives is in full agreement with these facts. At that period all the water of the Tarim and the Kontsche-darja was collected at Arghan. When the water again returned to the eastern line — the Ilek, Avullu-köl, etc., — the greatly enhanced evaporation thereby occasioned robbed the river of a large part of its former influx. In proportion as the Kara-köl lakes became filled, so did the Karakoschun dwindle.
But an even more powerful influence is exercised upon the volume of the Kara-koschun by the 35 marginal lakes that are strung along the right bank of the Tarim below Karaul. I have already dealt with them in detail in volume I, p. 300, and may therefore dismiss them here somewhat summarily. We calculated that, if all these lake-basins were empty, and if the Tarim carried a volume of 66 cub. m. in the second, as it did at Jangi-köl on i 6th May 1900, it would take a flood of that magnitude a period of 468 days and nights to fill the lakes in question. We also calculated that, if in the course of a year these lakes lost on an average a layer of water i m. in depth through evaporation and similar causes, the aggregate loss would represent a volume of 564 million cubic meters, and that, to make good this loss, as well as to preserve the lakes at the level at which we found them, it would be necessary for the entire volume of the river, amounting to 66 cub. m. in the second, to empty itself into the lakes in question without any diminution for a period of 99 days. In round figures we estimated, that the loss which these lakes inflict annually upon the Tarim amounts to one-fourth of its volume, or in other words that the creation of these lakes diminishes the mean volume of the Kara-koschun by one-fourth. The course which the Tarim follows in the vicinity of these lakes
IIIF Color HighRes
