Sec. iv] POPULATION OF KHOTAN : CHARACTERISTICS AND ORIGIN 149
foundation has been demonstrated in a striking manner by the numerous documents in an early Indian script and language which I discovered at the Niya Site. As will be seen later, the character of these documents renders it certain that the administration of the territory was during the third century of our era carried on in an Indian language, and that this language was familiar to a considerable portion of the population 27.
Similar evidence cannot be adduced at present for that early immigration from the side Alleged of China which the same tradition alleges. But there are certain facts and statements which f oml China point to some substratum of truth in that direction also. In the first place, we have the interesting and much-discussed passage in the notice of Khotan given by the Annals of the Northern Wei (386-532 A.D.), which tells us that ` the people of all territories west of Kao-ch`ang (the present Turfan) had deep-lying eyes and prominent noses, and that the inhabitants of this territory (i.e. of Khotan) were the only ones who did not present a very strange appearance but had rather a Chinese look ' 28. ` Deep-lying eyes and prominent noses' are certainly the features which would strike a Chinese observer most in physiognomies of the so-called Aryan' type, and we may safely conclude from their mention that this was the type then already prevailing in Eastern Turkestan.
The appropriateness of this description renders the exception made in regard to the people of Khotan all the more significant. But can we accept with equal confidence the statement made about the Chinese look of the latter ? Would not the presence of those features in a less pronounced fashion, have sufficed to suggest to Chinese eyes something more familiar in Khotanese looks, especially when the local tradition of an early immigration from the East was remembered ? These are questions to which in the present state of our knowledge we must hesitate to give even conjectural answers. Yet if we assign to the observation of the Wei annalists this qualified bearing, and at the same time assume that the tradition related to an element in its population which had come from the side of China but was in reality rather Tibetan than Chinese, an explanation is furnished which would equally satisfy tradition and the evidence of anthropological facts.
There can be little doubt that even the least Mongolian type of Tibetan, Mr. Rockhill's Chinese
` Drupa type ' above described, with its broad thick nose and high cheek-bones, must approach Tibetans. nearer to the Chinese than to any ` Aryan type. Chinese tradition regards the Tibetans as descendants of certain tribes who were exiled from China to the Koko-Nor region in a prehistoric period 29 ; in the legend of Khotan, too, the immigrants are represented as being led by a prince exiled from China. We need not attach any weight to this point of similarity. Yet' it might help to explain how an immigrant element akin to the present Tibetans, which had reached Khotan through Tsaidam at a very early period, could in later tradition figure as having been of Chinese origin. That the presence of such an element would have been far more noticeable in the racial appearance of the Khotan population at the time of the Wei dynasty than at present does not need special explanation.
The same ethnic element would also account for an important fact not yet noticed here, Linguistic the appearance in our oldest Khotan records, the Kharosthi documents from Niya, of numerous traces
PP ~ y Tibetanof
words, mostly titles or terms, which are certainly neither Indian nor Iranian nor Turki, but connexion. often suggest a Tibetan origin 30. The very fact that some of these words, though technical
n See below, chap. xi. sec. vi.
28 Comp. Rémusat, Ville de Khotan, p. zo ; Franke, Zur Kenniniss der Türkvölker, p. 19; for earlier speculations on the passage comp. e. g. Ritter, Asien, v. p. 362 ; Richthofen,