THE WATERS TO THE NORTH OF THE KARA-KOSCHUN. I 3 I
and of live-stock. During that day's march we observed no toghraks, either old or new; we were too far from the old arm of the Tokus-tarim.
Here then, whilst journeying along the north-west shore of the Kara-koschun, we doubled one »polyp-like» extension of the lake after the other, and one marsh after the other, all stretching to the south-west, in accordance with the parallelism which obtains throughout the whole of the Desert of Lop. That denticulated or serrated outline is quite natural, being a result of the relief as fashioned by the wind; the polyp-like arms of the lake are therefore bound to extend towards the southwest. As a matter of fact, I found that this side of the lake presents a totally different outline from that which has crept into our maps since Prschevalskij's journey, although their delineation is not based upon direct observations. Besides this, the Kara-koschun has in many respects changed since Prschevalskij's time.
Fig. 84. OLD BED OF THE TARIM NORTH OF RARA-KOSCHUN.
We followed, then, the last-mentioned division of the lake towards the southwest, having on our left large fields of kamisch and on our right a labyrinth of countless tamarisk-mounds, the bushes being some of them living, some dead, together with kamisch and small dunes. As we afterwards ascertained, this marsh is directly connected with the Kum-tschapghan, it being possible to paddle between the two without a break. There can be no doubt that in recent times the Karakoschun has expanded towards the north-west, an inference warranted also by the number of canals which leave the Tarim between Jurt-tschapghan and Kum-tschapghan, and by the presence of two or three pretty high dunes which rise above the thick kamisch in the middle of the marsh. The dunes are older than the marsh; had the water not come there quite recently, they would have been levelled down , and destroyed. Wild-boar, hares, and hedgehogs were common in that locality.
Then we crossed over a newly formed arm, without any current and with perfectly fresh water, and having isolated, not connected, dunes, 4 to 5 m. high, on its right shore. After that a fresh arm of the lake forced us to make a detour to the north, north-east, west, and south-west, and then we encamped, for it seemed to be a waste of time to try to find our way through these marshes with camels when we had no guide. By the evening we were able to procure guides from the neighbouring Kum-tschapghan. These men told me, that the Kara-koschun was then quite as high as it had been at the autumn high-water season. They seldom remembered a winter so cold and snowy as the last had been. The snow was, they stated, 2 span (about 44 cm.) deep on the ice; and it was to this circumstance that they ascribed the great mortality amongst the fish that spring, and the poor quality of those that were left. The lakes and marshes were said to be everywhere full of dead fish floating on the surface.
At the present time, in the region where the Tarim terminates, there exist only three inhabited places — (I) Kum-tschapghan, with 57 settled inhabitants, be-
IIIF Color HighRes
