steep mountain-side as easily as they do up the dune-accumulations of the Desert of Tschertschen. But this is not the case; for these two mountain-arms, which are in any case not very much higher than the dune-masses of the Desert of Tschertschen, place an insurmountable barrier in the way of the invading drift-sand — a wave-breaker as it were to stop their advance. The small dunes which have drifted within them have entered direct by the rocky gateway. This goes to show, that the wind is not able to lift the drift-sand to any very great height above the surface of the earth. It is well known to all who have had personal experience of a hard tempest in a drift-sand region, that by far the greatest amount of drift-sand travels immediately above the surface of the earth, the individual grains being rolled along by the wind. It is only the lighter, smaller particles that are lifted one or two meters above the ground. The impenetrable haze which comes on during a desert storm is caused by the fine dust, which in part is already intermingled with the sand, in part is caused by the friction of the sand-grains one upon the other. Yet even this drift-dust does not ascend to any great height. In very violent gales in the Takla-makan I have observed the blue sky shining in the zenith, while in a horizontal direction I was unable to see farther than a score of meters or so.
The watercourse shoots away from the mountain foot at a right angle. The dunes on both sides of it are built up for the most part in pyramidal shape. We therefore selected the bed of the watercourse as the most convenient route by which to make our way through them. Its bottom consists in part of hard, greyish yellow clay, in part of sand, and has an imperceptible fall towards the north. But in places, even in the bed of the watercourse, the original soil is to some extent sanded over, though the coating is extremely thin. Here, from the convex elbows of the bed, there project flat tailing tongues of sand, corresponding to the alluvial peninsulas in a river, and not seldom they stretch right across the bed. On the opposite side, answering to the deeply penetrating concave sweeps of a river, the dunes rise as steeply as they can to their full height. Nothing but running water could keep such a path open through the sand-field; that is the reason why we have here all the characteristics of an ordinary river-bed. It looks as if the watercourse avoids the biggest of the dune-accumulations, and gets out of their way, for though it does, it is true, penetrate at every bend into the high sand, it at once swings out of it again, and crosses over to the opposite side, where it again curves. Hence it is quite natural, that the river, through its erosive energy, operating at the foot of the sand, should preserve a steep face to the dunes. But a similar undermining process does not attack the convex bank: there the sand is flatter.
The profile of the bed of the watercourse is generally V-shaped. Consequently its bottom is extremely narrow, so narrow in fact that there are not many places in
IIIFカラー高解像度
