1oo6 NOTES ON THE PHYSICAL ANTHROPOLOGY [Appendix C
The position of the Wakhi will perhaps become a little plainer if we consider those tribes, measured on the last expedition, which according to 2A fall nearest to Wakhi (i). They are as follows:
Kirghiz (z) 4.48
Roshani . 657
Shughnani . 8.71
Ishkashmi . 9.13
Wakhi (I) . 9.41
The essential feature in which Wakhi (i) differ from Kirghiz (z) is length of nose, Kirghiz (z) being far shorter-nosed (and noticeably more platyrrhine). The essential features in which Wakhi (x) differ from the Roshani, Shughnani, Ishkashmi, and Wakhi (z) are nasal-breadth and nasal index, these last-named tribes being far narrower-nosed and more leptorrhine.
There is no great difference in any feature between the Tajik and Wakhi (I), the most noticeable being in nasal length and breadth, the Tajik being at the same time shorter-nosed, narrower-nosed, and more leptorrhine.
The tribes nearest related to the Mastuji and Chitrali, according to the differential index, are the following (where a EA contains a 0 of i•oo or over, the physical character to which it relates is mentioned in brackets):
Tajik . • 4'56
Roshani . 6.45 (N.B.)
Shughnani . . 656 (N.B.)
Ishkashmi . 7.35 (H.L., C.I.)
Kirghiz (z) . . 735 (N.L.)
Wakhi (z) . 754 (N.B., N.I.)
Tajik . . 5.61
Kirghiz (z) . . 5.70 (Biz., U.F.L.)
Roshani . . 6.7o (N.B., N.I.)
Shughnani . . 8-o6 (N.B., N.I.)
The same difference, therefore, exists between. the Chitrali and Mastuji on the one hand, and the RoshaniShughnani-Ishkashmi-Wakhi (2) on the other, as between Wakhi (1) and that group, viz. their noses are broader and more platyrrhine, and in several cases the A for these characters rises above Poo. On the other hand, Kirghiz (z) are shorter-nosed and more platyrrhine than either of them, though the chief feature which differentiates the Mastuji from the Kirghiz is the extremely narrow face of the former (a characteristic which they share with the Kafir). The inference is that above all the Roshani, who show so many affinities with the surrounding peoples in spite of the secluded nature of their habitation, probably represent in purest form the original population of the whole region ; the true Homo Alpinus type. To west (Tajik) and east (Taklamakan fringe) this type has been modified in various degrees by contact with a broad-nosed Mongolo-Turki type. The Karategin-Wanji group, while having basic affinities with the main Pamir stock, have been modified by contact with a narrow-nosed branch of Mongoloid peoples. The Sistani-Sayad group are basically Indo-Persian or Indo-Afghan, but contain, nevertheless, a leaven of the old Pamir strain.
Descriptive Characters. (Table VIII.)
It will be as well to say a word on the various descriptive characters ' of the peoples under consideration. These characters are, of course, from one point of view less satisfactory than measurements. In the first place they are not exact, in so far as they depend on the eye of the observer, whose standard may be affected by the observations taken on the people last studied. Further, the only method of comparing different series of observations is by. means of percentages, a system which may give misleading results in the case of a small series. In the present instance, for example, the number of Üzbeg observed amounts only to ten individuals. In the percentage, therefore, each individual counts as ten. One is described as ` fair-haired ', and consequently the figures