Raymond Pearl 
51 
the sexes in body-weight and stature account for 39'246 gr. or 28 of the observed 
difference in mean brain-weights. In other words, by selecting individuals on the 
basis of body-weight and stature alone we shall produce about the same degree of 
change in the mean brain-weight as by selecting on the basis of stature and age 
alone. 
Taking all the results together we reach the general conclusion that the 
difference between the sexes in mean brain-weight is only in part to be accounted 
for by differences in other bodily characters. While in general it is true that 
a small body has a brain of low weight, yet in order for the observed difference in 
mean brain-weight in men and women to be due to this factor alone either the 
women ought to have very much smaller bodies than they actually possess, or the 
men ought to have larger bodies, or both sexes ought to be different from what 
they actually are in size of body in the directions indicated. Of course it is very 
easy to say by way of speculation that the smaller brain of woman is due to the 
fact that the female human organism is subjected to less strenuous demands along 
the lines of motor and intellectual activity than the male. But such speculation 
leads to nothing on account of our lack of definite scientific evidence as to the 
degree of correlation between the weight of the brain and amount and intensity 
of psychic activity. 
One other question remains to be considered in connection with the relation 
of brain-weight to sex. Does either sex show any considerably higher degree of 
correlation between brain-weight and other charactei's than the other ? Table IX 
furnishes data on this point. With the exception of the Swedish brain-weight 
and age correlations and the brain-weight and skull-length correlations it is seen 
that the female coefficients are uniformly larger than the male. The differences 
are in many of the cases not significant in comparison with their probable errors 
but the general tendency towards higher correlatioiis in the females is clearly 
evident. The meaning of this tendency is in this case difficult to conjecture. 
It falls in line with the result which has been pointed out by Lee and Pearson in 
several papers*, viz., that for physical characters generally women usually show a 
higher degree of correlation than men, in civilized races. 
9. Brain-iueight, Age and Stature. 
The coefficients of correlation between brain-weight and age and stature 
respectively have already been given in Table IX, but before proceeding further 
in the discussion of their significance it is necessary to determine whether the 
regressions are linear. One might on general grounds expect the bi-ain-weight and 
age correlations to be skew, since this is generally the case with growth correlations 
where one of the variables is age and the other the absolute magnitude of some 
organ or character. If in any case the correlation turns out to be skew, it of 
course greatly complicates the problem. 
The method of analysis which I have followed in determining the degree of 
* E.g., "On the Kelative Variability and Correlation in Civilized and Uncivilized Baces," Hoy. 
Soc. Proc. Vol. Lxi. pp. 343—357. 
7—2 
