328 THE LOP-NOR PROBLEM.
shown in the sections, the region north of the Lop-nor is forty times more broken and irregular, each interval of ioo m. between staff and telescope crossing over a number of jardangs and wind-eroded gullies, which are not shown at all in the section (see fig. I 48 above). Even on the greater section (Pl. 36) all these irregularities do not find a place, and even if the horizontal scale of the section were to be prolonged twenty times, the jardangs would still lie quite close together.
The dissimilarity between the northern and the southern slopes of the swelling becomes therefore all the more pronounced, the former being very much flatter than the latter. The entire surface there consists of schor, once covered by a salt lake. The northern slope belonged at a later date to the basin of the Lop-nor, the southern at a still later date to the basin of the Kara-koschun. Possibly the greater arching of the latter may be due to the more recent desiccation and the accompanying process of expansion, while the slightness of effect which the wind produces upon a schor surface as compared with a clay surface may have helped to make the Lop-nor slope flatter than the Kara-koschun slope.
The greatest inclination along the whole line occurs between station No. 79 and station No. 81, namely 5.938 m. in a distance of only 400 m. Hence along this stretch the difference of elevation is two to three times greater than that between the terminal points of the whole surveyed line.
I have made my Plate (37) also embrace the Kara-koschun, although its basin, in contrast to that of the Lop-nor, is absolutely unexplored in this meridian. The representation I have given therefore in the transverse section of the lake does not claim to be anything more than approximately correct. I have assumed its breadth to be 25 km., while I have taken the conditions of depth as being about the same as they are in the quarter where the river-arms A—F flow out of the lake. Hence my section shows a maximum depth of 4.4 m. But I hasten to say, it is extremely unlikely that such a considerable depth as this occurs anywhere in this eastern part of the Kara-koschun. We did sound that depth, it is true, in the arm C, but not only does C lie a good deal farther to the west, in a district better protected against the drift-sand, but the sounding was taken in a spot that had been excavated by an eddy. It is very likely that the deepest hollows in this part of the Kara-koschun are not deeper than the deepest hollows of the Lop-nor, i. e. 1.689, 1.509, and I.699 m. below the level of the Kara-koschun at stations No. 83, No. 98, and No. 129. It is also conceivable that a considerably greater part of the breadth (25 km.) than what I have shown in the section consists of dry land. And, last but not least, it is very probable that the breadth of the Kara-koschun on this meridian is not greater than that of the Lop-nor, or I I to 12 kilometers, although I have taken 25 km. as the average breadth of the Kara-koschun.
I have also entered on the section several other measurements for comparison. At 81/2 cm. (on the Plate) from the southern shore of the Kara-koschun we already reach the 200 m. line above the lake; hence on the section it ought properly to lie 2 m. above the line which marks the surface of the Kara-koschun. This shows plainly the extraordinary difference in the slope north and south of the lake. The deepest sounding taken in the Kara-koschun, 5.150 m., is rather more than the highest point on the surveyed line, 4.791 m. The highest point on the shore of the Kara-
IIIF Color HighRes
