a few, it is true, but they were too close to the shore to afford any idea of the general bathymetrical relations of the lake, or of the relief of the lake-bottom. This failure was not however of any real consequence, for other subsequent trips, which were favoured with better fortune, proved that these sister-lakes, which hang on the Tarim like leaves along a stalk, are all exactly like one another, so that what is true of one is true of all. After mapping two or three of them with minute exactitude, and becoming familiar with their general relief, one may, without risk of serious error, construct bathymetrical curves of the others, provided the draughtsman is already familiar with their contours.
On I 8th May 1900 Gölme-käti, the Lake of the Lost Fishing-net, was cut off from the Tarim by a dam put down in its connecting canal, and situated 133 m. from the river. From that point to its issue into the lake, or 620 m., the canal contained sufficient water to allow of our paddling the canoes. The storm compelled us to keep to the eastern shore, where we obtained some shelter under the steep dune-walls. Along the line we in that way followed the depth increased as we proceeded from north to south, and the greatest depth was obtained quite close to shore at the extreme south end of the lake. Under the west shore the depth was less. Along both shores were a few small clumps of poplars, rather young, as well as some of the usual tamarisks and kamisch. The strip of flat land bordering the lake on the west was considerably broader than that on the east side. There too some of the tamarisks were actually growing in the lake, as though the lake were slowly advancing westwards, for these bushes do not of course take root except on dry land. It might be supposed perhaps that they grew up at a time when the lake-level was still lower than it was on i8th May; but, on the other hand, it is to be remembered, that at this date the lake was actually at a low level, for it was then cut off from the river. I do not of course pretend that these tamarisks growing in the water are a proof that the lake is travelling westwards. On the other hand, it is quite natural that the greatest depths should be found under the eastern shore, for in reality the lake is nothing else but a dune valley, or more correctly a valley between two dune-accumulations. The dunes which overhang its eastern side plunge steeply down to the maximum depths of the lake, and thence the bottom shelves gradually upwards towards the shore-line, and so farther, in like manner, up towards the crest of the dunes which shut in the lake on the west. There cannot exist a shadow of doubt but that the dune »waves» are advancing, though at an extremely slow rate, towards the west. Hence it may reasonably be assumed as likely, that the lakes are also subject to a movement in the same direction. The dunes will not of course allow themselves to be for one moment arrested by stationary water; they have no choice, they must continue moving westwards. Every storm that comes — and these storms blow with unexampled violence, — scoops the sand up from the windward side of the dunes and deposits it on their summits, whence it trickles over to the leeward side, so that in this way the eastern shore of the lake is gradually being pushed westwards. Similarly on the west side of the lake, the wind scoops up the loose-lying sand, beginning at the very edge of the water, and lifts it to the top of the next wave in the desert ocean, which thus wave by wave marches irresistibly westwards. In process of time these lakes are bound to disappear. We have seen that
IIIF Color HighRes
