J. Blakeman 
157 
Here Vj denotes the coefficient of variation of any character or Vj= lOOajjJ, 
and ./ is a mean vahie. 
For the practical purposes of the present inquiry we may take % = PY>§ 
although this is not absolutely exact*. As P is in cm.^ we must reduce S to cm. 
We then find : 
TABLE XXII. 
Capitulo- Statural Index. 
Sex 
Mean = 100 x 
Standard Deviation 
= 100(Tx 
Coefficient of 
TT ■ 100 cr. 
Variation = — 
X 
Correlation 
with Stature 
Regression 
with Stature 
9-1340 
+ -0255 
■4095 
± -0180 
4^4828 
± -1977 
- ^8366 
± ^0187 
- -0442 
± -0018 
? 
9-4136 
± -0273 
■3924 
± ^0193 
4-1686 
± -2051 
- -7957 
+ ^0255 
- -0441 
± -0024 
From the regression coefficients we at once find : 
For cfs : 
Capitulo-Statural Index = 16-6674 - ■04416/Sf, 
For %■&: 
Capitulo-Statural Index = 16-4812 - -044138, 
8 being measured in centimetres. 
Comparing this with Gladstone's Table on p. 116 of this volume of the 
Journal we have : 
TABLE XXIIL 
Capitulo-Statural Index for Different Statures. 
Stature 
Gladstone's Observed 
Values on Males 
Values deduced from 
Post-mortem data 
185-3 cms. ... 
8^66 
8-48 
177^7 cms. 
8^98 
8^82 
170^0 cms. 
9^00 
9-16 
162-4 cms. ... 
9-69 
9-50 
154"8 cms. 
10^05 
9-83 
Probably the fair homogeneity of the post-mortem material gives sensibly 
better results, especially as it is smoothed by the correlation process, than the 
See R. S. Proc. Vol. 60, pp. 491, 492. 
