latter the percentage of salt is relatively very small. The floor of the northern salt lake consists of an impermeable »saucer» of crystallised salt, so that no water can be lost there except by evaporation; but the bottom of the southern salt lake consists of soft sedimentary matter, in part probably of sandy earth, and is permeable. In no other way can we account for this lake not being bigger than it de facto is, seeing that it is the recipient of such an enormous inflow, and that this water is collected over such an extensive area and no doubt holds in solution considerable quantities of salts. But since these salts, which all eventually find their way into the terminal salt lake, do not avail to fill it up, the only alternative is to suppose, as I have done above, that merely a portion of its water is lost by evaporation, and that the rest finds its way out again by a subterranean emissary, and that the percentage of salt which remains in the lake is the residue left over by these two sources of loss.
The differences of level between the three lakes are remarkably small: that between the two freshwater lakes is a shade greater than that between the lower freshwater lake and the salt lake; in the latter case, if one may judge from the extremely sluggish current in the sound, the difference can only amount to a few centimeters. The greatest influx of water is brought by the western river, which rises amongst mountains of considerable altitude, where consequently the rainfall is fairly heavy. But so far as we were able to see, neither this river nor that which enters the eastern freshwater lake has its origin in glaciated mountains. Hence both alike are fed by the natural precipitation, which gathers into them in part directly, in part indirectly through springs. In the more northerly pair of lakes, we have found however, that a considerable quantity of the water which goes to swell the stream running south into the freshwater lake is derived from the melting of glacier arms. It is hardly conceivable that in such a short distance there should be any noteworthy difference in the amount of precipitation, even though in the latter part of the summer, as I afterwards learned, and observed even thus early, the precipitation does increase in eastern and central Tibet from north to south, in such wise that, while on the Astin-tagh it is minimal, in the country immediately north of the Tengri-nor it is very abundant. The difference between the volume of 43 cub.m. contributed to the northern salt lake and the volume of 56 cub.m. which enters the southern salt lake may be purely accidental, and does not necessarily prove that the southern self-contained drainage-basin possesses a more copious precipitation than the northern. The fact is, the rivers that feed the southern lake-complex have their sources at lower absolute altitudes than those of the northern lake-complex, these latter coinciding with the altitude of the snow-line and the faces of the glacier-arms. When therefore the winds that are charged with atmospheric moisture pass over this part of Tibet, their contained moisture condenses upon the border-ranges of the southern basin in the form of rain or snow, which quickly melts; at all events it finds its way directly and by the nearest courses into the lake-complex, while at the same time vast quantities of moisture are arrested and stored up on the great glaciated mass and in its fir regions, and consequently only reach the northern freshwater lake gradually and in even distribution, spread over a long period of time. The fact of the southern salt lake receiving just then more
IIIFカラー高解像度
