64 
Variation and Correlation iii Brain- Weight 
age ? The mean braiu-weight of the Bohemian males with a mean age of 45'699 
is 1454"839 gr. From equation (19) we have then 
1454-839 = 1091-021 + 2-288 S - (1-755 x 45-699) 
in which S is the mean stature of the selected group of Swedes. Solving we find 
.S'= 194-06, 
or, in other words, our group of Swedes would have a meaii height of 6 ft. 4 4 in. ! 
Now as a matter of fact the Swedish males with a mean age of 45-02 years have a 
mean stature of 169-79 cm. In order then to get from the adult male Swedish 
population a group having the same mean brain-weight as the adult male 
Bohemian population, by a selection of stature alone, it would be necessary to 
raise the mean stature of the Swedes 24-3 cm. or roughly 9^ inches. Of course 
it is not for a moment to be supposed that evolution in brain-weight proceeds 
by the indirect selection of stature alone. Indeed when it is remembered that the 
mean stature of Bohemian males is only 169-4 cm. the example just cited shows 
how little the indirect selection of stature has to do with the matter. In the 
discussion of evolutionary problems, however, it is important to know just how 
much or how little effect each factor involved is capable of producing. It is in 
this direction of measuring the relative effect of different combinations of factors 
that regression equations such as those given above have great value. 
It is apparent from equations (19) to (26) that stature and age in a measure 
compensate one another so far as their effect on bi-ain-weight is concerned. The 
influence of stature is however in all cases greater. It is of some interest to know 
in the particular cases how great an increase in stature compensates a given 
increase in age and vice versa. In order to show this I have prepared the following 
table. 
An increase of 10 cm. in stature is compensated in its effect on brain-weight by 
an increase in age in the 
Swedish males of IS'O years 
Swedish females of 43-4 „ 
Hessian males of 25-3 „ 
Hessian females of 10'3 
v- Total series only. 
10. Brain-weight and Body-iveight. 
In only one of the four series of brain-weighings studied (the Bavarian) was 
body-weight tabulated in the returns. Recent workers on the subject have left 
this character out of their data on the general ground that it is of very doubtful 
significance or reliability in post-mortem returns from a general hospital popula- 
tion. Certainly it is true that it is influenced to a very marked degree by the 
