148 THE KHOTAN OASIS : ITS GEOGRAPHY AND PEOPLE [Chap. VI
were we to accept this apparent territorial vicinity as an adequate explanation of the racial connexion above indicated. No inhabited part of Tibet lies nearer to the oases than two hundred miles by the map, and the intervening belt of high mountains and equally barren plateaus, reaching a width of over seven hundred miles in the direction of Central Tibet, is too formidable a natural barrier to have ever permitted continuous intercourse between the populations on either side. Hence the idea of any gradual intermingling of the two peoples must be dismissed as highly improbable on physical grounds.
It seems almost equally difficult to account for so large an admixture of Tibetan blood by any ethnic movements that could possibly have taken place within the historical period. We are sufficiently well informed by the Chinese records about the main political events which affected Eastern Turkestan and Khotan in particular, since the beginning of our era, to know that the Tibetans did not begin to play a part in them until the last third of the seventh century 21. These early Tibetan invasions were soon checked by a reassertion of Chinese power, and though we hear of subsequent irruptions, it was only some hundred and twenty years later that the Chinese hold over the Tarim Basin was finally brought to an end by the Tibetans 22. As regards Khotan itself, documentary evidence supplied by finds from DandanUiliq proves that Chinese influence was not upset by the Tibetans until after 790 A. D. 23 Considering that the previous invasions were scarcely more than mere raids, and these, it seems, mostly directed against the extreme east of the ` Four Garrisons ' from the side of Tsaidam, they cannot reasonably be credited with having left so deep an imprint on the racial character of the people of Khotan. The period of Tibetan predominance in Eastern Turkestan, which lasted from the end of the eighth century to circ. 86o A. D. 24, appears also far too short for this. After that date the creation of the powerful Uigur kingdom in the east of the Tarim Basin finally closed what must have always been the natural avenue of Tibetan invasion, the routes from Tsaidam.
If we weigh properly the geographical and historical facts here briefly indicated, we are necessarily led to consider the possibility of the ` Tibetan ' infusion in the present population of Khotan and Keriya reaching back to a far earlier epoch. And it appears to me that there are indications which distinctly support this solution. The most important among these is the old local tradition of Khotan, to be discussed in detail in the next chapter, which ascribes the creation of the kingdom to joint occupation of its territory by two colonies originating from India and China, respectively 25. The genuineness and relative antiquity of this Khotanese tradition is attested by its being recorded in a substantially accordant form both in Hsüantsang's Hsi yü-chi and in certain Tibetan texts 26.
We need not consider here any of the quasi-historical details which this tradition pretends to record, nor examine to what extent it may help to explain peculiar features in the archaeology and linguistic remains of ancient Khotan. Leaving these points for subsequent consideration, we must note only the essential fact that indigenous tradition, dating back to a period long before Tibetan influence could have made itself felt politically, recognized as the earliest settlers of Khotan two entirely distinct elements, one which was believed to have come from the side of China and the other from the extreme north-west of India. That the tradition, as far as it relates to an immigration from the latter side, possessed some historical