times S. 25° W., and only occasionally did we need to cross over outliers or wings jutting out from them. We thus had their, steep leeward faces close to us on the right, and on the left their flat windward slopes. The former were deeply in the shade, the latter brilliantly lighted by the sun.
The view from the top of one of these dunes was especially instructive. Before us stretched, flat and desolate, the entire length of the Akato-tagh, with not a scrap of snow upon it. Along its foot was the darker line of the vegetation, nourished by the springs of Temirlik and other places, and even to the unaided eyesight it was quite apparent that we had ascended a pretty good distance since we left them. We could only surmise where Ghas-nur lay to the east-north-east. During the still, warm hours of the day we observed in the west a mirage like those that are common in Tibet, though this one was more sharply outlined than usual. We see not only the lowest slopes of both the Tschimen-tagh and the Akato-tagh, but between them, in the middle of the Tschimen valley, the free-standing mountain of the Tschoka-tagh. In the spaces between there appear to lie the still tranquil waters of a couple of lakes, in which the outlines of the mountains are clearly and distinctly mirrored. By a mere chance these two lakes run together under the Tschoka-tagh, so that this mountain appears to be entirely detached and to be hovering above the horizon. In order that a mirage of this character may be seen, the spectator requires to be at the same level as the landscape that is in this way reflected, for the nearer we approached to the foot of the mountains the fainter grew the mirage, until finally it disappeared altogether. Further conditions are that a warm, still, clear day be followed by a cold night and that the sun shine from the same direction as the landscape »miraged», or at an angle of 90° towards it. In this instance we had the sun in the south and the mirage in the west. I have already made mention of a similar phenomenon at the large salt lake at Camp XXXII, but that moved as the sun moved.
Meanwhile we had crossed the belt of sand and reached the entrance to the glen. Descending a couple of steep erosion terraces on the right-hand side of the glen, we then marched up it along its lowest part. Here the glen of Savughluk-saj descends from the south-south-west, but upon emerging turns to the N. 65 ° E. and continues towards the Ghas-nur. Thus its eroded watercourse keeps close to the granite cliffs on the right-hand side of the valley, the cliffs being in part unincumbered and in part marked by gravel-and-shingle and coarse sand, arranged in terraces. The drift-sand reached down to within a couple of hundred meters above Camp LXXVIII, where it covered the lower slopes on the left-hand side of the glen even at its outlet. Thus both here and outside the glen entrance the drift-sand dunes have piled themselves up exclusively on the left bank of the stream, as indeed one would expect from the configuration of the ground and the fact that the prevailing winds come from the west. The right bank is on the other hand entirely free from sand; in fact there is not much room for dunes to accumulate in.
Narrow though this belt of sand is which we have thus encountered between Ghaslik and Mandarlik, yet in its orientation it reminds us so very forcibly of at least two other sandy localities in the same region, that I feel justified in laying down the following law of geographical homology. In a broad long-axial valley,
IIIFカラー高解像度
