404 THE TSCHERTSCHEN DESERT.
The well of Julghunluk-köl-kuduk was choked up with dust and drift-sand, but nevertheless is, it was evident, a well-known and frequently visited camping-ground, for by the side of it there was actually a wooden trough for animals to drink out of. The well itself is sunk in the deepest part of the bed, in a spot where formerly an eddy in the stream scooped out the bottom to a depth of 4 or 5 m. below the edge of the scarped bank. But even at that depth the bed was perfectly dry.
Beyond that point the young forest grows conspicuously thick, and then for a time we left the river-bed on our right to avoid a tiresome bend which it makes to the east. Once again we encountered patches of level mud, either barren or producing a scanty vegetation, marking the sites of former marginal lakes. At intervals there are small dunes, seldom more than 4 m. high. We frequently observed marks of old shore-lines. Tamarisk-mounds are common. At the point where we again struck the river, there is a path coming from the south-east, from Tschigelik-uj and Tokum; this again showed recent traces of travellers' presence. I was told that the path in question leads through a desert with tschaval-kum, or »scattered dunes». Here the forest is fresh and vigorous, the trees being of medium age, and frequently it is dense and even magnificent. This points to the presence of groundwater near the surface, so that dig were we might, we should be able to get fresh water. We were now over against Schirge-tschapghan on the Tarim, according to the bek's calculation a short day's journey from it. •
After that young forest alternates with mature forest; in places it is as vigorous as any forest there is beside the middle Tarim, and yet it is asserted, that it is 25 to 3o years since water flowed along that bed. For two or tree years indeed pools had survived in the deeper parts, but these too disappeared many years ago. The well in the district of Kutschmet had fallen in. The fact of the old names remaining proves that that road is still frequently used; otherwise it would be forgotten, like the names beside the Kuruk-darja in the Desert of Lop, where nobody was able to give me a single name, although there were once as many names there as beside any other river of East Turkestan. When the natives definitely cease to visit the Ettek-tarim, then its names too will in like manner disappear. Its bed is extraordinarily easy to trace, being as sharply defined and accentuated as the older parts of the Tschertschen-darja; which indeed it greatly resembles, the only difference being that the vegetation beside the Ettek-tarim is still living, and in fact shows no signs of incipient decay. Although there are dense forests on both banks, the riverbed itself is bare and empty; sometimes we might have fancied ourselves threading an avenue or wooded defile. ' The deeper channel in which the current actually ran is still distinct, meandering backwards and forwards from bank to bank. These characteristics continued until we reached the district of Taschtan Kullu, which is reported to lie on a level with the Jäkän-boldschemal on the great Tarim. Thence we turned off towards the north-west, still keeping to the river-bottom; this brought us quite close to the high sand on the west, with dunes i 5 to 20 m. high. At the same time the sand on the east grew unexpectedly formidable, and ahead of us, through the now predominant underwoods, we actually saw glimmering, a small mountain of sand, desolate, isolated, barren, rising to an altitude of at least 5o m. At its foot we encamped in a tract covered with kamisch; this we had hitherto
IIIF Color HighRes
