Raymond Pearl 
35 
ditions as endocarditis, pneumonia cruposa, and nephritis chronica. Yet the 
diflferences found in the appearance of the brain in different cases were clearly 
associated with differences in age, and not with causes of death. On this whole 
question more data are needed. Exact measurements of the correlation between 
skull capacity and brain-weight are much to be desired. We also lack definite 
biometrical evidence regarding the effect of morbid conditions on the weight of the 
brain. The evidence at present available seems to lead to the conclusion here 
indicated. 
On the other hand, it can hardly be maintained that natural selection has 
acted on the size of the brain to the degree that it has on skeletal structures. 
The brain, in this respect, comes nearer to the organs included in our most variable 
group. So that, on the one hand with a less marked dependence of the measure- 
ment on the previous metabolic condition of the organism than is found in the 
most variable group of organs and characters, and on the other hand, with a less 
stringent selection with reference to size as compared with the least variable 
group, an intermediate condition in variability for the brain is to be expected. 
Of course, too much stress cannot be laid on this argument as to the relative 
variability of the characters of the human body in view of the meagreness of the 
data at present available. The data at hand, however, plainly suggest some such 
an explanation, and it is only as a tentative suggestion worthy of being kept in 
mind as further data are available that the matter has been developed here. To 
summarise ; when a series of the various organs and characters of the body whose 
variation has been measured are arranged in the order of their relative variability, 
those characters falling at the most variable end of the series are those whose 
measured value is affected markedly by the previous general metabolic condition 
of the organism, and in which ability to function properly is not closely related to 
size of organ, and in which natural selection, if it has acted at all on the character, 
has not acted directly upon the thing measured. The opposite end of the series 
includes the least variable organs and characters, in which both the above 
mentioned conditions are reversed. Brain-weight and skull capacity occupy an 
intermediate condition both with reference to variability and the conditioning 
factors discussed. 
Attention may be turned next to the question of the relative variability of the 
different races in respect to brain-weight. In the following scheme the races are 
arranged in descending order of mean brain-weight in the left-hand columns, and 
of variability (measured by the coefficient of variation) in the right-hand column. 
The lists are based on " total " series. 
From these lists it would appear that as a general rule the higher the mean 
brain-weight is, the lower will be the variability. But it must be kept in mind 
that with our present series none of the differences in the variability columns is 
significant. Taking the extremes, Bavarians and Swedes for the males, and 
Bavarians and Bohemians for the females, the differences between the coefficients 
of variability with their probable errors are respectively "526 + "217 and '958 ± -337 . 
5—2 
