SURVEYING THE DESERT. 241
a channel that linked together two or more different lake-basins. As this trench ran from north-west to south-east, it forced us also to travel in the latter direction. It wound backwards and forwards like an ordinary river-bed, and had all the appearance of one; for example, its greatest depths occurred at one of the sharpest bends, where we crossed over it, the vertical distance from the edge of the erosion terrace to the bottom amounting to 8 m. That it was as deep as this I convinced myself by using the levelling-staff. Nor need this great depth occasion any surprise, for in the existing Tarim system there are places where we obtained depths of i 2 to 14 m. But it is wonderful to find a river-bed, or at all events a short fragment of a river-bed, in a part of the desert where one would least expect it, where one would . in fact only expect to discover traces of former lakes. A couple of toghrak trunks and some pieces of tamarisks were the only traces of ancient vegetation remaining on its banks. Its bottom consisted of soft, powdery material, sometimes dust evenly distributed, sometimes sand in the shape of a couplé of dunes which had formed under the lee terraces. As the bed lies at right angles to the prevailing wind, one would expect it to be already filled up, instead of which it is amazingly deep. But at the second place where we crossed it, the depth did not exceed 3 to 4 m. The bottom was plentifully strewn with mollusc-shells. Camp No. CLXI was likewise pitched in a hollow, possibly a continuation of the river-bed I have just been speaking of, though its outline was irregular. It is through some such channel as this that the water possibly once flowed from the northern basin to the southern basin, or vice versa. Were the desert lakes described above to be laid dry, several similar beds would come to light, namely those through which the water flows now from the south to the north. In which direction it formerly flowed through the bed we are considering it would be impossible to say. It is also conceivable, that this bed may have formed the immediate continuation of the Tarim at a time when that river discharged into a lake situated in the middle of the desert.
Fig. 150. OLD RIVER-BED IN THE DESERT.
On the 13th March we covered a distance of 13,007 m., in the course of which the surface again rose 2.763 m., so that upon reaching Camp CLXII we were only o.1 m. higher than at our point of departure. We had therefore now reached the point at which we were about to leave the northern depression of the Lop-nor behind us. Here too the features of the country changed very materially in appearance. The jardangs grew fewer and fewer, and at the same time the expanses of level dust between them grew broader and broader. The jardangs also altered both in form and consistency. In the northern part of the desert they consisted of hard and compact yellow clay, and were in general more energetically modelled. Here however they were built up of a greyish friable clay, the .edges of which crumbled away
IIIFカラー高解像度
