374 
Pigmentation, Selection and Anthropometric Characters 
height. Both the methods used and the colour categories adopted to classify 
pigmentation are in this case different from those employed at Birmingham. 
There are four hair colour groups — red, fair, medium and dark, and three eye 
colour groups, light, medium and dark. In this case again the usefulness of the 
data suffers from lack of precision in classification. The height was taken to the 
nearest inch. 
The data have been treated in the same way as those derived from Birmingham. 
Table XIII corresponds to Table XII ; but the mean height and probable error have 
here been found not only for the hair colour groups, but also for the eye colour 
groups, and every combination of hair and eye colour. An inspection of this Table 
shows how little deviation there is from the general mean. When we look to see 
whether there is any indication of the same deviations from the general mean 
commented upon in connection with the Birmingham data, we fail to find any 
correspondence whatever. In this case, on the whole, there is less deviation than 
at Birmingham, and there is nothing to call for attention. 
In order to make this more clear, figures have been constructed on the same 
principle as that followed to form Figures I A and B. Here again in Figure II 
A, B, C and D the mean of the whole is represented together with the means of 
the various groups and twice the probable error calculated with reference to the 
mean of the whole. 
Figure II A and B gives the height and hair colour shown according to this 
method. In this case we have the only mean which exceeds twice the probable 
error. The mean of the red haired men is considerably in excess of the general 
mean, but not so much as to be of itself very markedly significant ; and further, 
when it is found that among the women in Figure II B no such very marked 
deviation is indicated, then we are probably justified in thinking that it does not 
represent any true connection between red hair and stature in excess of the mean. 
A glance at Figure II is enough to show that the differences are, except for 
reddiaired men, absolutely insignificant, and that there is no indication whatever 
to be found of those deviations which we observed in the Birmingham data. 
Height and Pigmentation. Summary. The data, therefore, which have been 
examined in this paper do not afford any grounds for the belief that pigmentation 
and height are correlated. As has been pointed out, there are some vague indica- 
tions of possible association between a certain pigmentation class, namely dark 
brown, and a stature in excess of the general mean. But this scarcely does more 
than to provide hints for future research ; and the main essential for such research 
is that pigmentation should be measured with greater accuracy. With data such 
as have been at our disposal, which in the first place are derived from populations 
that would not be deliberately chosen for such an investigation, and which are 
lacking in the accuracy of pigmentation classification that is desirable, it would be 
surprising to reach any markedly definite result if the relation be really small ; 
the failure to do so, therefore, does not make it less probable than before that, 
