352 THE LOP-NOR PROBLEM.
Notwithstanding the probable correctness of all these theoretical speculations, we know that the existence of the terminal basin of the Tarim system (disregarding its situation) is extremely precarious, and that the lake shrinks so rapidly that even in the short period of no more than four years the diminution is distinctly perceptible. The causes of these oscillations are not however to be sought for in the wide circle of feeding-areas from which the water is gathered into the Tarim, but they are inherent in • the immense central basin of the desert itself, which the Tarim and its tributaries traverse both lengthwise and transversely, and in the course of which their originally pretty constant volumes, far from being augmented, are on the contrary drawn upon nearly all the way down. Of course I do not mean to deny that the great climatic anomalies, which may obtain throughout the whole of Central Asia, may also make their influence felt upon both the volume and the extent of the Kara-koschun; for it is evident, that an unusually snowy winter and spring must not only give rise to an active thaw in the early summer, followed by an exceptional rise in the volume of the river, but larger quantities than usual of excess or overplus water will contrive to find their way down to the lake. And even though the amount of the winter precipitation is small, there is still always a great deal of snow left from the preceding year in the source regions above the limits of perpetual snow, so that large volumes of water invariably flow down to these rivers, to say nothing of the many thousand springs scattered throughout the mountains, all of which, taken together, add an almost unchanging tribute to the flood that fills the bed of the Tarim. But to this question I shall have to return when I take a general survey of the whole of the Tarim system.
Nor would I for one moment attempt to deny, that the volume of the Karakoschun is appreciably affected by the temperature and wind relations in East • Turkestan itself. An unusually hot summer, accompanied by an active circulation of the atmosphere, increases the evaporation over the whole course of the river. If now the various factors that are operative in the various regional divisons of the basin happen to be simultaneously active, the resultant effects are naturally very evident; though it must of course be extremely seldom that such a remarkable conjuncture occurs. When, as more usually happens, the one contributory factor is operative in one region and another factor in another, the ultimate outcome is, that the resultant effects remain pretty constant; and we do know for certain that the terminal lake has been in existence for at least 2,000 years.
IIIF Color HighRes
