Sec. iv] POPULATION OF KHOTAN : CHARACTERISTICS AND ORIGIN 145
and dark eyes, becomes more marked in the people of Keriya, who otherwise ` seem to show, as might be expected from their neighbouring position, a very strong similarity with the Khotanese'.
Mr. Joyce thus sums up the principal facts deduced with regard to the people of Khotan Principal
and Keriya. Both are, in the main, of so-called " Aryan " stock, the chief factor beingonsracialons
Lapouge's Homo Alpinus. There is, however, in each case an admixture of Turki blood and character. a further admixture of Tibetan. The latter appears to be stronger at Keriya than at Khotan, and at the same time here Mongolian influence begins to make itself felt.'
` The Pamir valleys, as far as Asia is concerned, seem to be the locality where Homo Alpinus appears in his greatest purity. In the Galcha he appears with a slight Turki and
Iranian admixture. In the Khotanese the Iranian is replaced by a Tibetan element, and further east, among the inhabitants of Keriya, true Mongolian traits are just beginning to appear.'
These last-named slight traces of Mongolian blood, present only in a diluted form among the inhabitants of Keriya, are attributed by Mr. Joyce to a more recent admixture and need
not be specially considered here. The other racial constituents traced by the above analysis are, however, important ; and the question arises whether we can account for them without assuming a radical change in the population since the pre-Muhammadan period. This, I believe, we are able to do with the help of the indications furnished by our available historical and philological materials.
Let us take in the first place the chief racial element which enters into the composition of Affinity the people of the oasis, that of Lapouge's Homo Alpinus. We have seen that the typical Khotanese representatives of this element in Asia are the Galcha tribes of the Pamir Valleys who speak andGalchas.
Eastern Iranian dialects, and to whom belong also the closely allied Wakhis and Sarikolis. The close approach in outward appearance between the latter people and the present Khotanese
has already been noted. Now it is a fact of no small significance to find the Tang Annals
distinctly asserting of the inhabitants of Ho-p`an-Co or Sarikol that ` their external appearance and language are the same as those of the people of Yü-t`ien (Khotan)'7. I had occasion in
a previous chapter to disctiss this important statement 8. I also pointed out that it receives confirmation from an exactly corresponding observation made by Sung )(ün and Hui-shêng concerning the population of the territory of Karghalik, which forms the natural link between Sarikol and Khotan 9, and which still contains, in the Pakhpo inhabitants of its mountains, an ethnic element unmistakably allied to the Galchas 19.
To these historical attestations of an ethnic connexion between Khotan and the eastern- Linguistic most territories still possessing a Galcha population, we can add a weighty piece of philological evidence.
evidence. As will be seen in a subsequent chapter", the ancient site of Dandan-Uiliq has furnished, both to me and to ` treasure-seeking' natives who previously ` explored ' it, a con-
siderable number of documents written in Brahmi characters and in all probability belonging
to the eighth century of our era. The language of these documents has been proved by Dr. Hoernle, their first decipherer, to be an Indo-Iranian dialect having its nearest congeners
in the Galcha dialects of the Pamir region 12. There is every reason to believe from the character of these documents that the language in which they are written was the one actually