78 
Variation and Correlation in Brain-Weight 
for brain-weight and " size of body." It has been shown (p. 66) that of a given 
change in mean brain-weight between 25 and 30 per cent, can be brought about 
by a selection of " size of body," measuring this both by stature and weight of 
body. As it is fair to assume that natural selection has acted on " size of body " 
we thus reach some idea of how much of an effect such selection has had on the 
weight of the brain. 
A vital point at issue is as to what are the underlying causes of brain-weight 
differences. We have seen that there are definite racial types in brain-weight 
apart from differences in the other bodily characters which could be studied in 
this connection. To take a concrete example, we find that groups of male and 
female Bohemians having the same mean length and breadth of skull have probable 
mean brain-weights differing by about 69 grams, the female mean being lower 
than the male, of course. To what is this difference to be ascribed ? Evidently 
there are a considerable number of possible factors which may enter into the 
matter. In the first place these males and females would still differ in stature by 
a certain amount so that a part of the remaining brain-weight difference would 
be accounted for in this way, but probably less than 25 per cent. Selecting too 
the same skull size would undoubtedly reduce the gross sexual difference in 
stature to some extent. The mean age of the arrays is essentially the same, so 
that nothing is gained from this source. To account for the still remaining 
difference there are at least the following possible factors : 
1. The specific gravity of the female brain may be lower. 
2. A unit increase in skull length and breadth may be associated with 
a greater change in skull height in the male than in the female. 
3. The walls of the female skull may be thicker than in the male. 
4. The female brain may fill the skull cavity less closely than does the male. 
(The opposite has been stated to be the case. Cf Donaldson, loc. cit. p. 118.) 
5. The ventricles of the brain may be proportionately larger in the female 
than in the male brain. 
With the material at present available it is impossible to determine which 
of these possible factors really play the important role. It is my purpose merely 
to call attention to the kind of problem that confronts one at this point in the 
analysis of brain-weight constants. It at least will serve to show some of the 
directions in which more definite information is desired. 
Donaldson* has well said: "The search for correlation between the size and 
form of the brain and the degree of the intelligence has interested all who have 
worked on this organ, and although it might be designated as the psychologist's 
standpoint, it has, from the very first, been in some measure before the minds of 
all." This point has so far been avoided in this paper for the reason that it 
seemed to me desirable to settle the question of the purely physical variations and 
* Loc. cit. p. 85. 
