ent on irrigation. Had not conditions of extreme aridity already prevailed in ancient times, it would be impossible to account for the survival in almost perfect preservation of a multitude of objects, very perishable by nature, in places so exposed as mere refuse heaps outside houses. Exactly corresponding observations are furnished by what archaeological explorations at other ancient sites of the Tarim basin have taught us. The climatic condition$ of the periods immediately preceding abandonment must have been practically as arid as they have been since and are now.
This uniform and important fact has a direct bearing upon an important and much-discussed geographical question usually spoken of as that of `desiccation'. It is too large to be more than touched upon here. If the climatic conditions were sixteen centuries ago quite as dry as they are now, how is it to be explained that cultivation at those two ancient sites and at others also has since their abandonment become wholly impossible?
I believe that the explanation as far as the Tarim basin is concerned is supplied by the diminished volume of the rivers upon which cultivation is wholly dependent. The most likely cause of this diminution may be sought in the shrinkage of the glaciers on the high ranges which mainly feed those rivers. The shrinkage itself can well be accounted for by assuming, as has been suggested by Sir Sidney Bur-
rard and Professor Von Ficker, that those glaciers comprise great reserves of ice which have been left behind by the last
glacial period and have since been undergoing slow but more or less continuous reduction through milder climatic conditions. This process of using up what might be called `fossil ice' would suffice to explain shrinkage in the sources
IIIF Color HighRes
