258 PROF. K. PEARSON ON THE MATHEMATICAL THEORY OF EVOLUTION. 
may be spoken of as individual and tribal taste. Tribal taste manifests itself in the 
preference of one sex as a whole for mating with members of the other sex having 
special characteristics, or to the rejection as mates by one sex of members of the 
other having special characteristics. The preference and rejection being in neither 
case absolute, but relative. This type of sexual selection, which may be spoken of as 
preferential mating, is measured by the differences in mean and standard deviation 
between the variation-curves for the whole adult population of one sex, and for the 
mated portion of it. For example, the mean height and mean variation in height of 
women generally are not identical, or are not necessarily identical with the mean 
height and mean variation in height of wives. Preferential mating may have 
reference to any organ or measurable characteristic of either sex. 
Individual taste on the other hand does not denote the exclusion from mating of 
any section of the population of either sex. It is due to the preference of individuals 
with an organ or characteristic of given size for mates with the same or another 
organ or characteristic of a size, the average of which differs from the whole popula¬ 
tion average. This type of sexual selection which may be spoken of as assortative 
mating is measured mathematically by the coefficient of correlation between the two 
organs or characteristics in mated pairs. 
It will be obvious that preferential mating and assortative mating are fundamental 
ideas to be quantitatively allowed for in any theory of heredity. Their action may 
often be in entirely opposite directions. 4 ' 
(e.) Reproductive Selection .—One pair may produce more offspring than another, 
and in this manner give through heredity greater weight to their own characteristics. 
For example, the mean height of mothers is not identical, or is not necessarily 
identical with the mean height of wives, nor is the standard-deviation of fathers 
identical or necessarily identical with the standard-deviation of husbands. Further, 
the means and standard-deviations of mothers or fathers of sons may be different 
from those of mothers or fathers of daughters. The- quantitative measure of repro¬ 
ductive selection is the correlation between the size of any selected organ in either 
male and female and their reproductivity, the reproductivity being measured by the 
number of their offspring in either sex or both sexes. 
offspring fated to die out. Of course a might in certain cases after (1) mate with both A and Z. None 
of these possibilities corresponds exactly to what is described in this section as assortative mating, which 
in no way necessitates the exclusion from mating of z. a and z are not indeed competitors, but seeking 
different qualities in their mates. Thus, in man for example, the intellectual and non-intellectual might, 
and possibly do, sort themselves out in pairs, i.e., there is a correlation between intellectual capacity of 
husband and wife. 
* For example, preferential mating might lead in a highly social community to the rejection of 
consumptive mates, while assortative mating might, through localisation or community of habit, lead to 
considerable consumptive correlation. Thus sexual selection as a whole may influence in diverse ways 
the inheritance of the consumptive taint. 
