m., and a volume of 1.37 cub.m. in the second. Shortly after that we waded across a narrow, shallow offshoot of the lake, and were then surrounded on all sides by wide expanses of water, the mutual connections of which it was difficult to make out. In several places the dunes plunged steeply down into the lakes, and occasionally the tamarisk-mounds projected above the surface like little islands. The second branch had these dimensions—breadth, 10.35 m.; mean depth, 0.483 m.; mean velocity, o,7o14 m.; and volume, 3.51 cub.m. in the second. The breadth of the third branch, which was in the same vicinity, was I 1.9 m. ;the mean depth, 0.593 m.; the mean velocity, 0.4424 m.; and the volume, 3.12 cub.m. in the second. We pitched Camp No. CLXXV on a narrow tongue of land beside the lake from which the arm issued. The kamisch was pretty thick on its northern shore, and beyond the kamisch were some rather low dunes. The arm in question resembled here a well-developed river-bed, having sharp-cut terraced banks on both sides. Thus the Tokus-tarim issued at Schirge-tschapghan from the »dividing» lake, the Jäkänlik-köl, with a volume of 8 cub.m. on 1st April, as compared with a volume of 3.5 cub.m. on the 5th April of the year before. Possibly this difference of date may have had some effect upon the volume, though it is difficult to believe it, owing to the great number of lake-basins which we passed, and which regulate the volume and retard the fall. It may therefore be assumed, that the difference is indicative of an increase between 1 goo and 1901. On I 6th April 1900 the Tokus-tarim had higher up its course a volume of 9.4i cub.m. Had the proportion between the volume at that place and at the crossing-place of 1 got remained the same, the Tokus-tarim would
then have a volume of 2 cub.m., and thus be a considerable stream, quite as big
as the Jarkent-darja in September. Seeing that all this water issued from the Karaköl lakes, it would appear that the eastern lake and river system had experienced a noteworthy increase during the course of the year. Herein we have a fresh proof of the view, that the system of the lower Tarim is as a whole shifting, partly from south to north and partly from west to east; and this moreover is in agreement with the fact that the terminal lake of the system is moving northwards to the old basin of the Lop-nor. The directions assumed by the three branches in question are indicative of the extreme flatness of the country. The first flows towards the north-east, the other two to the south-west. They now discharge their 8 cub.m. into the Kara-koschun, so that this must be added to the mean annual volume of 64 cub.m. which enters it from Jurt-tschapghan. Agreeably with my own observations and the statements of the natives, the Tokus-tarim is growing annually at the expense of the Tschong-tarim. Besides the increase in volume and the addition of a third arm, there is yet another circumstance calculated to confirm this supposition, namely the fact that the northern lakes have also increased enormously since I goo, when they were still but small sheets of water. But while in 1900 they were still perfectly bare and barren of vegetation, in Igo I they were encircled with kamisch; which is here disseminated with incredible rapidity by the currents, not by the wind, for the wind blows in the opposite direction. Thus from Camp. No. CLXXV it would have been possible to paddle with ease to almost any point we chose in the watery labyrinth of the Lop country, nay even half-way across the Desert of Lop in the direction of the ruins of Lôu-lan. I have already drawn attention to the fact that
IIIFカラー高解像度
