国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

Its position is perhaps midway between the Desert and Kirghiz groups, with a slight leaning towards the latter,
owing to the great variability of head-circumference.

Table 11 shows the inverse order of relationship of the various peoples, and perhaps the differential index
affords a better indication of remoteness than of affinity. The highest ΣΔ exists between the Kafir and Dolan,
viz. 16·28; and, as far as ΣΔ of 13 and over are concerned, the antithesis between the Pamir group and the
Kirghiz group, and between the latter and the Chinese, alone appears. It is true that the names of
Karanghu-tagh and Nissa occur, but these, as we have seen, have a very mixed population, and it is evident
that the population contains a large Pamir element. With regard to ΣΔ of 12 and over, we find certain of the
Desert group, viz. Khotan, Keriya, and Korla, opposed to the Kafir. As we have seen that the Pamir group
have a certain relationship with the Desert group, we may take it that the Kafir, who are after all only related
with the former through the Mastuji, constitute an extreme, and bear a certain relationship to some people whose
influence does not extend as far as the desert. The Loplik, too, appear as widely divergent from the Kirghiz
group, as might be expected owing to their relationship with the Chinese. The difference, which tends to
separate the people of Faizabad from the Pamir peoples and the Chinese, accentuates the traces of relationship
which have already been found for them with the Kirghiz group. It is unnecessary to pursue the investigation
further in great detail, but attention may be called to one or two points. When ΣΔ of 10 and over are considered,
it is seen that the Chinese, who have already shown a divergence from Niya, are differentiated from Keriya, and,
what is more important, from Korla, one of the Desert group. Among the Pamir group the Chitrali must be
regarded as influenced by some extraneous element, probably that which enters into the composition of the Kafir,
since they are brought into opposition with Turfan and Korla. Kök-yar, which has been related most closely
with Khotan, and next with the Sarikoli, is differentiated from the Kafir, Kirghiz, and Loplik, the last being,
as we know, closely related to the Chinese. To establish the position of the last named and of the Wakhi, it is
necessary to consider the ΣΔ of 9 and over. We then find that the Chinese are widely separated from the
Kirghiz and Desert groups; that they are not related to the Pamir peoples is evident from the fact that their ΣΔ
for the Chitrali and Sarikoli are over 8. They, therefore, constitute an extreme. As for the Wakhi, they are
shown to have no relationship with the Kirghiz group by being opposed to the Kirghiz themselves, while their
ΣΔ for the Kafir is over 8. The position of Keriya and Niya is interesting. They show the greatest divergence
from the Pamir group and the Chinese, but differ, though in a lesser degree, from Faizabad and the Dolan.
Keriya, however, has been shown to bear some sort of relationship to Kelpin, while its ΣΔ for the Charkhlik,
one of the Desert group, is over 8. Keriya probably, therefore, occupies an intermediate position, or, together
with Niya, which is further removed from the Kirghiz group, contains some extraneous element. The ΣΔ of
Faizabad and the Kirghiz, which is over 8, shows that the former cannot be definitely included in the Kirghiz
group, and, therefore, probably stands half-way between the latter and the Desert group, especially as it bears
some slight affinity with Khotan.

Conclusions from the Measurements.

The foregoing examination would seem to establish the fact that the various peoples with which this paper
deals may be divided into four groups. First, a group of mountain folk, all closely allied, in the extreme west of
the area under consideration. The nucleus of the group is formed by the Sarikoli, Mastuji, and Pakhpu, with an
easterly extension into the Desert area in the shape of the Bagh-jigda. Closely akin to them are the Chitrali
and Kafir, who, nevertheless, exhibit certain differences without approaching any of the other peoples, and,
therefore, probably contain some element foreign to this district. Second, a group of desert peoples, the nucleus
of which is composed of Turfan, Khotan, Korla, and Charkhlik. This group has a westerly extension into the
mountains in the shape of the Wakhi; it possesses certain affinities with the peoples forming the nucleus of
the former group, and the inhabitants of some localities, such as Kök-yar and the mixed populations of Karanghu-
tagh and Nissa, appear to stand half-way between the two. With the Desert group should be classed the people
of Polu and the rather mixed population of Hami; Niya and Keriya also have some affinity with certain of its
members, but seem to contain some other element. However, their apparent aloofness may be due to the
small number of measurements available. The third group, one which is very distinct, is formed of the Kirghiz,
Kelpin, Dolan, and Ak-su. Traces of relationship with the Desert group, however, are not wanting, though these
are slight, and the people of Faizabad may represent a mixture of the two elements, or, indeed, all three, Pamir,
Desert, and 'Turki'. The fourth group, the Chinese, seems to stand practically alone, though the Loplik are