146 THE KHOTAN OASIS : ITS GEOGRAPHY AND PEOPLE [Chap. VI
spoken by the contemporary inhabitants 13. Since this fact fully accords with the ethnic relationship which the Chinese records just quoted indicate between the Khotanese and the Galchas of the same and the immediately preceding periods, it is clear that the prevalence of Homo Alpinus in the anthropological composition of the present Khotan people must be attributed to direct inheritance from the pre-Muhammadan population.
It is different with the admixture of Turki blood. That this admixture has taken place only since Khotan was converted to Islam is clearly shown by historical and philological evidence. We know that until the close of the tenth century, when that conversion took place, Khotan under its Buddhist rulers remained independent of the Karluk Turk dominion established in the north-west of the Tarim Basin. During the prolonged struggle which preceded the introduction of Islam, an appreciable influx of Turki people into Khotan is highly improbable. The same observation applies to the ninth century, when Khotan with a great portion of Eastern Turkestan was under Tibetan control, while the total absence of Turki words in the Brahmi documents of Dandan-Uiliq, already mentioned, excludes the possibility of the population of Khotan having received any Turki element down to the close of the eighth century.
The admixture of Turki blood, which must thus be ascribed to the period since the conquest of Khotan by Satok Boghra Khan's family, is shown by Mr. Joyce's analysis to have been relatively small 14. This may appear surprising, in view of the universal adoption of the Turki language in Khotan, as throughout the oases of the Tarim Basin, but it is in reality easily accounted for. In this region, as in other Central-Asian territories where subsistence is possible solely by the laborious cultivation of irrigated lands, or else by industries and commerce, the Turks, nomads by origin and habits, appeared primarily only as soldiers. By their superior military qualities and organization they were able even in small numbers to place their chiefs in undisputed sway over the far more civilized but peace-loving people of the ancient oases. They likewise succeeded, by a peculiar faculty for ethnic attraction, often illustrated in the case of Turkish conquests, in making subject populations rapidly adopt their own language and willingly accept their political predominance. On their own part these Turks could not escape gradual amalgamation with the people of the oases whom they ruled and protected, and in the end they became wholly absorbed in them 15. But their numbers were far too small to affect fundamentally the racial character of the population.
The latter remark applies probably with even greater force to the Kara-Khitai, Moghuls and Kalmaks or Oirat (Eleuths), all tribes of Mongolo-Turki race, who during subsequent periods exercised political predominance in Eastern Turkestan. Retaining their nomadic habits longer than the Turks, these nations had their main seats of power to the north of the Tien-shan, in the country known after them as Moghulistan or Zungaria. In view of what we know of the manner in which their temporary power was exercised in the settled portions of Turkestan, they could but slightly have strengthened the Turki element in so distant an oasis as Khotan 16.
13 In this connexion the survival of certain Iranian words in the Turki now spoken in the Khotan region deserves attention. These words (e. g. sag ` dog') are unknown to the people in other parts of the Tarim Basin ; comp. Grenard, Mission D. de Rhins, iii. p. 64.
14 See J. Anihrop. Insl., xxxiii. p. 323.
'6 The political effects of the Turki conquest of the Tarim Basin have been judiciously analysed and described by M. Grenard, Mission D. de Rhins, ii. pp. 7I sqq. It must, however, be noted that the influence which the varying local
conditions presented by the different territories exercised upon the progress and extent of this ` Turkization ' has scarcely been sufficiently realized.
16 Mr. N. Elias has called attention to the very small size of the armies' with which, according to Mirza Haidar's accounts, the various Moghul Khans warred against each other in Eastern Turkestan or conducted their raids into neighbouring territories. The chiefs of these tiny forces were, indeed, Moghuls of various clans ; but ' the tribal following which each chief could muster was a mere handful',