(By Ethyme Tschepourkowsky, experts from Biometrika, Vol. 15, No. 3/4. (Dec., 1923), pp. 254-270.)
APPLYING the geographical method to measurements of nearly 50,000 Russian peasants* I was led to the chief types given in the map, Plate I.
(i) There exists in the East of European Russia (Riazan, Tambof and Penza governments)? an island of relative dolichocephaly, namely 16 contiguous districts, the mean cephalic index of each not exceeding 79.9. Such a low mean
index is nowhere else to be met with. The empirical distribution of those districts is provided in the accompanying Table I(7816 men). The hair- and eye-colour in this region is relatively dark.
(ii) In the other parts of Great Russia the watersheds are inhabited by a much more brachycephalic type : the mean index of districts is there nearly 83-87. This type is relatively blond. The same population with much more accentuated blondness and brachycephaly occupies the swamps of White Russia along the rivers Pripiet and Berezina, effluents of the river Dniepr. The river valleys in Great Russia are occupied by an intermediate or mixed type of index: 81-82.
(iii) The Little Russian population is also brachycephalic but very dark.
(iv) Along the valleys of the Nieman and West Dwina (w. Duna) and in government Moghileff there are remnants of the Scandinavian type, blond and relatively dolichocephalic (80-81).
According to considerations I will not reproduce heref I think that the first of these types (East Russian) is the remnant of an ancient population of Turco-Finnish and partly slavic origin; the second invaded Great Russia much later (700-1000 years ago) from the swamps of White Russia and has an affinity with the round-barrow population of Britain$; the third colonized Little Russia quite recently from the Carpathians (Homo alpinzis) after the Tartar invasion. I have used biometric methods in the following cases :
(1) To determine if the number of observations wss sufficient to calculate the mean index of districts or subdistricts (communities) I calculated the coefficien of correlation : (a)between two mean indices for the same districts deduced from two serie? of observations: by myself, or by myself and another author. It was 0.912 + ,022; (b) between the lowest number of observations in each district and the above mentioned differences of the mean indices, deduced from two series of my own observations. The result was negative (r = -0,195 + .077), i.e. with the increase of the number of observations the differences decrease (see Tables II and III).
(2) To determine if my appreciation of hair-colour was sufficiently exact, I calculated the correlation for the relative number in percentages of extreme blonds and extreme dark-haired in each government (Table IV) and found r = -*883 +- *030.
(3) To solve the question if in the Russian peasant population the hairand eye-colours are correlated with the cephalic index I observed the colour of 27,000 peasants (soldiers) in the same conditions of light, divided them into geographical groups (districts), in every group calculated the mean index for each subdivision of hair- and eye-colour” and found (Tables V and VI) : (a) that the differences between those mean indices are distributed according to the laws of probability; and (b) that they also diminish with the increase of the number of observations (r-negative?). I t confirmed my previous assertion that the fundamental types cannot be discovered with certainty by the sole method of establishing correlations among characters;.
(4) Anthropologists often describe a people (in most cases without any attempts to resolve it into physically different con~ponents) by a series of arithmetical means for different characters. But if we take within the limits of the same race many determinations of a given character (for instance, if we take from a race ten groups of individuals and deduce a mean cephalic index for each group) these determinations will vary according to a probability curve. Each race is characterized by its own probability curve of this kind and we must compare these curves but not the means of characters deduced from one or two observations. In Table VII we see for instance the curves of mean indices of subdistricts (communities) in different districts of Russia and in Table VIII the curves of percentages of blond type in different governments of the Great and Little Russias.
(Rest to be posted in future)
Appendix: Composite photograph
I have constructed a special photographic apparatus which enables me to take rapidly nearly 2000 portraits on a cinematographic film without recharging. With it I have photographed under the same conditions of light nearly 5000 men 26-27 years old in full face and profile. Using Galton’s method I made with a reflex-camera the combined portraits of the above-mentioned East Region population, i.e. of the Riazan, Tatnbof and Penza Governments (see Plate 11, Fig. I), of White Russians of Mariupol District (Fig. 2) and of Little Russians of the Government of Minsk (Fig. 3). We see from these portraits combined from a large number of individuals (451, 147 and 140 individuals) that only a slight longheadedness characterizes the Eastern type. As a result I found no sensible differences between these types using Bertillon’s scheme. The composite portrait of Mongols, combining only 25 individuals, shows a striking difference (Fig. 4).
I have also used my apparatus to make composite portraits of the profiles of the ears (Figs. 5a, 5b) and of the nose (Fig. 6) which as far as I know have not previously been made by anybody else. The events of the last few years have
interrupted my investigations.