J. Blakeman 
143 
Looking generally at the brain-weight correlations we see that predictions 
based on the diametral product and the circumference will give us the best results. 
Age and stature have formed such a large part of the anatomical discussions on the 
subject that we feel compelled to retain them, but clearly a knowledge of any 
diameter of the head is of more importance in predicting brain- weight than either 
stature or age. After careful consideration we determined to make our most 
complex prediction formula depend upon a knowledge of the diametral product, the 
horizontal circumference, the stature and the age of the subject. 
Before considering prediction formulae, however, it is well to compare our 
correlations with any available material. 
In the first place we take the English skull. 
TABLE XVIL 
Comparison of Skull and Head Correlations. 
Character 
General Hospital Population 
Whitechapel Crania* 
s 
? 
$ 
? 
L with ... 
w 
•598 -t- -040 
•545 + -049 
C 
•597 ± •051 
•691-1- -040 
B with ... 
w 
•4894- -048 
•607 + -044 
C 
•631 ± -048 
•646 -1- -044 
^with ... 
w 
•544 + -044 
•673 + -038 
G 
[•501 -f -059 
■600 + ■049'] 
i/with ... 
w 
•623 + -032 
•627 + -042 
C 
•820 + •oae 
•848+ -021 
Pwith ... 
w 
•703-1- ^032 
•749-f031 
c 
[•750+ -035 
•823+ ^025] 
We notice that the correlations of capacity and external measurements are all 
greater than the corresponding correlations for the brain-weight. The apparent 
exception (C and L, ^ is within the limits of probable error), namely G and H. for 
both sexes, arises, we think, from the fact that H is here the basio-bregmatic and 
not the auricular height of the skull, and no stress can thus be laid upon this 
exception-f-. The conclusion, which we might have anticipated, is therefore forced 
upon us that it is easier to predict capacity from external measurements on 
the skull than brain-weight from external measurements on the living head. 
The values of the correlation between C and P, although obtained indirectly, 
are probably very close : thus A. Lee| gives for the corresponding values in 
Bavarian crania: ^ , •TOO ± "034 and %, •SH ± -023, the differences for the two 
races being well within the limits of random sampling. 
* Cited from Macdonell, Biometrika, Vol. iii. p. 232. Except those for P and C which are deduced 
from the formula : r,.p = ( F,, >•( + Fs)'c8+ Viircn)IVp. All these constants are known for the Whitechapel 
crania except r^,/ the correlation of capacity and auricular height, we have used for this the known cor- 
relation of capacity and basio-bregmatic height, cited in brackets under C and H. 
t If the auricular lieight be more highly correlated with capacity than the basio-bregmatic height, 
the result is very satisfactory, because the former is all that is obtainable on the living head. The 
point should be investigated as ample data exist. 
+ Phil. Trans. Vol. 196 A, p. 237. 
