Appendix A 277
the direct action of water being substituted for the under-cutting, which must, for extreme elevations, be almost the only possible method of attack. One sees even now throughout the mountains we traversed this process constantly at work, producing thousands of landslides -- a favourite sport of the spirits that inhabit mountains geologically young—and particularly those that are high enough to have their upper portions covered all the year round, or for a long season. They change watercourses and caravan routes, give birth to short-lived lakes which burst their sudden bounds and maddeningly confuse the exposed strata.
Snow appears on the Kuen Lun range at an altitude of about 14,500 feet, and at about i 6, 5 oo on the mountains rising from the plateau, the level spaces of which, in September and October, are substantially clear. The light snowfalls, such as we experienced, were quickly evaporated during the warmer hours of the day by the fierce winds blowing quite regularly from the south-west, and constituting one of the serious hardships of travel.
A great part of the whole plateau is therefore exposed to direct erosion. The effect near the edges of the plain is marked. In the interior, the water being of small volume, the streams meander hither and thither, current scarcely observable, and form shallow lakes which hold for a time that small part of the flow which has not been drunken by the thirsty sands. Such tremendous action of the past, of which we found the witness above described, has not, to my knowledge, been discussed by other travellers. I am therefore the less confident in theorising about it--yet suggest that, in spite of a probably constant precipitation of snow and rain, the summer-volumes of these plateau streams may have been vastly larger when the area above eighteen thousand feet was much greater, thus permitting an accumulation of winter
IIIF Color HighRes
