276 PROF. K. PEARSON ON THE MATHEMATICAL THEORY OF EVOLUTION. 
Male line. 
Female line. 
Grandson on grandfather. 
T986 
•0885 
Great-grandson on great-grandfather. 
T048 
•0307 
In the first case, the strength of inheritance is more than double through the male; 
in the second, more than triple through the male what it is through the female line. 
Were this law of inheritance true, not only of stature, but of other physical, and 
especially of mental characteristics, some justification might be found for confining 
hereditary peerages initially given for merit to the male line. Meanwhile, it cannot 
be too strongly emphasized that the present results apply only to one organ, are 
based on comparatively few families drawn from a special class of the community, and 
thus stand in need of careful criticism in the light of ampler statistical material. 
Another point already briefly referred to, which seems of significance, is the in¬ 
equality of regression in the case of ascent and descent in the direct line. It may 
seem paradoxical to assert that sons are more like fathers than fathers are like sons, but 
the solution is bound up in the statement that fathers of sons are less variable than 
sons, or, in another form, that every son is not to the same degree a potential father. 
Similarly, the opposite paradox that fathers are, on the average, more like their 
daughters than daughters are like their fathers, finds its solution in the relatively 
great variability of fathers of daughters. 
In Table IV. are tabulated alongside, in each case, the standard deviation for the 
corresponding general population, the standard deviations for inheritance from 
selected classes. Here again we see a general law for height, which deserves to be 
investigated for other organs, and for a variety of animals, namely, we breed “ truer 
to the type,” have less variability in offspring, if we breed from selected males rather 
than from selected females. We shall see later the effect of selecting both parents. 
(c.) On Further Relations between Correlation, Regression, and Variability. 
(i.) The Coefficient of Variation V. —In dealing with the comparative variation of 
men and women (or, indeed, very often of the two sexes of any animal), we have 
constantly to bear in mind that relative size influences not only the means but the 
deviations from the means. When dealing with absolute measurements, it is, of 
course, idle to compare the variation of the larger male organ directly with the varia¬ 
tion of the smaller female organ. The same remark applies also to the comparison of 
large and small built races. 
If the absolute measurements* have in the case of man to be on the average altered 
in the ratio of 13 to 1 2 to obtain those of the woman, if Mr. Galton has gone so far 
as to replace any woman by an equivalent man on this basis, then, clearly, to compare 
* The ratio 13 to 12 is not only true of stature, but approximately of several other organs, weight, 
brain-capacity, etc., &c. 
