142 
Biometi^ic Constants of English Bram-weights 
Dr A. Lee*. Thus she found it sensibly zero for $ Theban mummies and — 'IS 
for male Thebans. A list of values is given by Levvenz and Pearson-|- in which 
it is shown that English crania give a practically zero value for the correlation 
between cephalic index and capacity. That value was, however, obtained 
indirectly, i.e. from the L and B correlations with each other and the capacity, 
and accordingly it seemed worth while to investigate the matter directly on the 
Whitechapel crania. Accordingly we extracted from Macdonell's tablesj the 
cajoacities and cephalic indices of 70 and 80 % crania and worked out directly 
the correlation between the two characters. We found : 
Correlation of Cranial Capacity and Cephalic Index for Whitechapel Crania : 
Males: + -093 + -080. Females: - "099 ± -075. 
Neither value is really significant having regard to the probable error, and we 
might therefore conclude that this correlation is sensibly zero for English crania. 
Such a result would be in good accordance with the sensibly zero correlation 
between brain-weight and cephalic index for males, but not in accordance with the 
result for English females. Assuming that there might be an error somewhere in 
the arithmetic the correlation between brain-weight {iv) and cephalic index / was 
tested indirectly by the formula 
ruo = ( Vb ^bw - Vl riw)/ Vi (2), 
with the following results : 
TABLE XVI. 
Cephalic Index and Brain-weight. 
Sex 
Direct Calculation 
By Formula (2) 
Males 
Females 
+ -0014 
+ •2110 
+ •0.373 
+ •2084 
The degree of closeness is as great as we might expect from the two methods, 
the square of F/lOO being neglected in obtaining formula (e). There is, therefore, 
we think, no doubt that, while the correlation between cephalic index and capacity 
is sensibly zero for English crania of both sexes, the females of a General Hospital 
Population have a sensible correlation between brain-weight and cephalic index. It 
is conceivable that this is due to some exaggeration of breadth in the measurement 
of the heads of the younger females, for whom the brain-weight would be larger 
and as a rule the hair interfere more than would be the case for older women or for 
the males§. On the whole we think we have very adequate grounds for omitting 
the cephalic index entirely from our prediction equations for the brain-weight. 
* Phil Trans. Vol. 196 A, p. 232. t Biometrika, Vol. iii. p. 374. 
J Biometrika, Vol. iii. Tables after p. 244. 
§ The hair difficulty occurs twice in the case of measuring B and only once in measuring L. Hence 
B would tend to be more exaggerated than L. 
