in order to call attention to another feature which differentiates this group very strongly from either group A or group B. The members of group C are the only people (with the exception of the Kirghiz) who show any ten-
dency to skin-pigmentation. While other groups show loo per cent. of members described as white-rosy', the
Sistani, Sayad, and Biloch show a major percentage of individuals described as ' brownish-white '. The figures
are as follows : Sistani 64 per cent., Sayad 71 per cent., Biloch 86 per cent.
These figures are highly significant, not only as marking off this group very clearly from the others, but as suggesting the possible presence of an Indo-Afghan strain. As mentioned above, the Kirghiz alone of the other
peoples under consideration show a tendency towards pigmentation ; but the number of brownish-white individuals amounts only to 7 per cent., while a consideration of the EO produces results which practically preclude any relationship between them and group C.
In order to provide some comparison between the measurements taken by Sir Aurel Stein upon this journey and those obtained on his previous expedition (see Serindia, vol. iii, and ,journal of the Royal Anthropological
Institute, vol. xlii, 1912), I have calculated the differential indices for all of the former with five selected groups of
the latter, viz. 19 Wakhi, 22 Chitrali, 28 Mastuji, i 8 Kâfirs, and 38 Kirghiz. To take the Kirghiz first. This group of 38 individuals shows little or no relationship to any of the peoples measured on the last expedition, the
ED ranging from 10.92 (Karategin) to 20.27 (Sistani). I am omitting the group of 54 Kirghiz measured on the
last journey, but in regard to these the E6, amounts to 9.52. This is a most remarkable fact, and seems to imply that the two groups of Kirghiz have little in common save the name. On an examination of the factors which
compose the E0, it is seen that no less than three amount to more than Poo, viz. those for bizygomatic-breadth, facial index, and upper facial index, indicating that the Kirghiz measured on the previous journey were far more euryprosopic than those encountered on the last expedition. Further, they are noticeably broader-headed and more brachycephalic ; their noses are also longer and broader, though the indices of the two groups fall close together, and show them to be the most platyrrhine of all the peoples measured by Sir Aurel Stein.
Now, since the group first measured constitutes the extreme, not only in platyrrhinity, but also in brachycephaly and euryprosopism, it is a fair inference that the group measured last has been influenced by contact with
Alpine or Iranian stock. Hence its comparatively close affinity with the Tajik and, to a less degree, with the Roshani. Now it is clear that the Alpine stock is distinguished by relatively broad nostrils, so an admixture of this element would not disturb the nasal proportion so much as the facial and cranial.
Another interesting point is that this extreme platyrrhinity of the first group of Kirghiz, supported by the only slightly less platyrrhinity of the second group, when compared with the nasal-measurements and index of
the Özbeg, reinforces very strongly the view that there are two definitely divergent Mongolo-Turki stocks in Central Asia : one very broad-nosed and relatively very platyrrhine ; the other very narrow-nosed and, though also relatively short-nosed, comparatively leptorrhine.
On his previous journey, Sir Aurel Stein also took measurements of a small group of Wakhi, 19 in all, on the Taklamakan slope. I shall refer to them as Wakhi (i). On the present journey he secured measurements of no less than 54 farther in the hills. These I shall mention as Wakhi (z). Now the first series was small, and we are brought up against the question of random sampling, but nevertheless the difference between the two groups
is very marked. The differential index amounts to no less than 8.99, and contains two O over Poo, viz. those for nasal-breadth and nasal index, Wakhi (1) being broader-nosed and more platyrrhine than Wakhi (2). Further,
Wakhi (I) are broader in face, more euryprosopic than Wakhi (2). Now in all these respects Wakhi (I) hold an intermediate position between Wakhi (2) and the peoples on the edge of the Taklamakan desert, whom in my previous survey I regarded as the nucleus of what I termed the desert group ' : the inhabitants of Korla, Turfan, Khotan, and the Charkhlik. In that survey I regarded Wakhi (1) as an extension of the desert group into the Pamirs, modified by contact with what I called the Pamir group ', of which the nucleus was the Sarikoli,
Mastuji, and Pakhpo. I think that my point of view was wrong. The measurements of the group Wakhi (2) show that the Wakhi as a whole, with the Shughnani, Ishkashmi, and Roshani, constitute the true ` Pamir group ', and
are more truly representative of the Homo Alpinus type ; while Wakhi (I) represent an extension of this type towards the ` desert group ', who have been modified by contact with that particular branch of the Mongolo-Turki stock which is represented by the Kirghiz, especially the group of Kirghiz first measured, which I will call Kirghiz (I). The group of Kirghiz measured on the last expedition, which I will call Kirghiz (2), take their place as a branch of this Mongolo-Turki people who have been modified by contact with the Pamir group.
II 6 N