Appendix C CHINESE TURKESTAN AND THE PAMIRS i361
confirmation by more measurements taken in the latter country. In the Pamirs is a series of tribes, who, though chiefly of Iranian stock, begin to exhibit slight traces of Indo-Afghan blood. In at least one tribe, the Kafir, these traces are considerably more than slight. The Chitrali also seem to stand in closer relationship to an Indo-Afghan people (but a rather specialized Indo-Afghan people) than the other Pamir tribes. Some admixture has taken place between the Turki and Desert folk.* In the case of Ak-su the Turki element predominates. In the cases of Niya and Keriya, who should be classed as rather aberrant members of the Desert group, it seems probable that their departure from the norm is due to Turki admixture. Faizabad appears to be a mixture of all three groups, Pamir, Turki, and Desert ; and this is what might be expected. The root-stock of the population would thus be Iranian, though it has been exposed to Turki influences since Indo-Scythian times, and has thus become somewhat modified. In the East, Chinese influence begins to make itself felt, but only over a very restricted area. Hami, Turfan, and Korla have been thus affected, and it is not unlikely that the Desert people have had some corresponding though perhaps slight effect upon the population of Nan-hu and Tun-huang. The position of the Loplik is a little difficult to fix. I am inclined to regard them as a very early Mongolian offshoot, who found their way into the Loplik marshes long before Nan-hu and Tun-huang were colonized by Chinese of kindred stock. Owing to long residence they have become affected by contact with the Desert folk. The other alternative is to regard them as a Desert people affected by contact with the Chinese, but, having regard to the slight effect which Chinese influence has had upon Hami and Turfan, in spite of long contact (which has greatly modified their ethnography), I think this is not nearly so probable. The great differentiation of the Chinese and Turki groups is interesting, since both are regarded as Mongolian'. It is evident that they belong to widely different branches of the Mongolian race, and it must be concluded that the Turki are allied to the Southern Mongolian, the Chinese of Nan-hu and Tun-huang (and also probably the Tibetans whose measurements are given) to the Northern Mongolian stock. If this is so, and the Turki peoples do, in fact, contain a large Southern Mongolian element, their stature has been greatly increased in the course of their wanderings, by contact, probably, with Iranian peoples. This leads us to the question to what extent we may regard the Turki as a distinct branch of Mongolians, and whether it would not be more correct to look upon the various tribes which fall under this heading as being originally mixtures, in varying proportions, of Mongolian and Iranian elements, which time has reduced to comparative homogeneity. Finally, the point which emerges most clearly from the welter of measurements and descriptive data contained in this paper is this : that the original inhabitant of the Pamirs and Taklamakan Desert, including the cities now buried beneath the sand, is that type of man described by Lapouge as Homo Alpinus, with, in the west, traces of the Indo-Afghan ; and that the Mongolian has had very little influence upon the population. In using the term Homo Alpinus, I wish it to be understood that I employ it merely as the name of a certain type already described, and do not necessarily imply that the actual population of the Alps is closely allied to the population of Chinese Turkestan.
In conclusion, I will quote from my previous paper (already cited) the description of these types
I. A white-rosy race, very brachycephalic, stature above the average, with thin, prominent nose, varying from aquiline to straight, long, oval face, hair brown, usually dark, always abundant and wavy (I think this should now be altered to curly), eyes medium in the main. This is Lapouge's Homo Alpinus.
` z. A race, also white, but with a slight tendency to brownish, also very brachycephalic and with stature above the average, nose broader and usually straight, cheek-bones broad, hair straighter, darker, and less abundant, eyes dark. The "Turkish" race.
A brown, mesaticephalic, tall type, thin, prominent, and aquiline nose, long, oval face, black, wavy hair, dark eyes. This race may be termed the Indo-Afghan.
A brownish, brachycephalic race, stature under the average, nose straight, thick, and broad, black, wavy hair, little on face, brown eyes. The Tibetans.' (I now doubt whether the Tibetans can be said to constitute a race ; there seems to be a great difference between the inhabitants of Northern and Southern Tibet respectively, a difference which corresponds in the main to that between the Northern and Southern Mongolian. Possibly the population of Tibet consists of both the latter elements, with an infiltration from the desert of modified Iranian in the north. More information, however, is necessary before we can speak of Tibet as a whole.)
` 5. A yellowish, brachycephalic, short race, short, flattened nose, with broadish nostrils, straight or concave, short, broad face, straight, black hair, scanty on face, dark, oblique eyes, with fold covering the caruncula. The (Southern) Mongolian race.'