462 THE TARIM DELTA.
the ground. Every now and again, though not really very often, we observed older toghraks, still quite vigorous. Next comes a chain of long narrow lakelets, sending out ramifications in every direction, but all alike embedded amongst the sand. It is evident that the large dunes which occupy the narrow strip of ground between these lakes and the' Tarim, without however quite touching the left bank of the river, travelled into their existing positions at a time when the eastern waterway did not yet • exist. Their advance has been cut short quite recently by the formation of this waterway, because they rise to the same altitude on both its banks. The reason they do not reach the bank of the Tarim is to be sought in the assumption I have already posited, and had confirmed, namely that at some former epoch the whole of the Tarim system discharged through this eastern waterway, and that it continued to do so a sufficient length of time to allow of the dunes, which were then on the west side of it, advancing a long way from it in that same direction; just as we have already ascertained, that there exists at the present day on the west side of the Tarim a zone that is free from sand.
The kamisch and sedge continued to increase as we advanced northwards. The fact of fishing-nets being set at intervals was a proof that there were human beings somewhere not far away. Next followed a lake of medium size, across which we paddled steadily towards the north-north-east for a pretty long stretch. Finally we pulled up beside five kamisch huts, which turned out to be the same fishing-village in which I spent a day during a storm in 1896. But how changed the aspect of the scene! Had I not already determined the position of the place, and remembered the natives themselves, and the situation of their huts, it would have been perfectly impossible to identify it. The name of the village is Märdäktik; that of the lake, Sadak-köl. This last, which four years previously had been pretty large, was now in great part overgrown with reeds, and filled up with sand and mud. The sheets of open water that remained were not large, and the greatest depths south of the village amounted to 2.25, 2i30, and 3.75 m. Thus the transformation of lake into river had here taken a distinct step forward.
One of the older men of the village told me, that in the autumn of 1899 a natural river-arm had been formed eastwards, coming from the district of Kumtscheke, and that its water had reached his village about a month later. Immediately east of the village there is a patch of high sand, from the top of which I understand it was possible to see this new waterway, which entered the lake east of the patch of sand in question. I at once visited the place, making my way along the reed-grown base of the dunes, and found the statement quite correct: the new stream did enter the lake by two small arms, and they had a pretty lively current. North-east and east-north-east there was a lake which extended as far to the north as the dunes would permit us to see. The only vegetation in the Sadak-köl is a belt of reeds along the southern shore. Here and there are living poplars, some young, some of moderate age; from which it is fair to infer, not only that there has been water here before, but also that the ground-water has remained sufficiently near to the surface to prevent the trees from dying. The lake sends off polyp-like arms in every direction amongst the dunes; these last appear to have been to some extent levelled down since the country was first overflowed. From the southern end
IIIF Color HighRes
