PROF. K. PEARSON ON THE MATHEMATICAL THEORY OF EVOLUTION. 271 
Table IV.—Variation in Selected Groups. 
Class of Selected. 
S.D. in 
inches. 
S.D. in 
inches. 
Unselected. 
Matage :— 
Wives of selected husbands .... 
2-293 
2-303 
All wives 
Husbands of selected wives .... 
2-617 
2-628 
All husbands 
Parentage :— 
Fathers of selected sons. 
2-135 
2-325 
All fathers of sons 
Fathers of selected daughters . 
2-548 
2-731 
All fathers of daughters 
Mothers of selected sons. 
2-225 
2-334 
All mothers of sons 
Mothers of selected daughters . 
2-180 
2-274 
All mothers of daughters 
Fraternity :— 
Sons of selected fathers. 
2-403 
2-617 
All sons 
Daughters of selected fathers 
2-189 
2-347 
All daug-hters 
Sons of selected mothers. 
2-495 
2-617 
All sons 
Daughters of selected mothers . 
2-250 
2-347 
All daughters 
Table V. — Sexual Ratio. 
Class. 
Ratio of Means. 
Ratio of S. D.’s. 
Ratio of V.’s.* 
Husbands to wives .... 
1-082 
1-141 
1-055 
Males to females. 
1-081 
1-115 
1-032 
Sons to daughters. 
1-080 
1115 
1-032 
Fathers to mothers . . , . 
1-079 
1-084 
1-005 
* V = the “ coefficient of variation ” or percentage of variation in organ 
= 100 S. D. 4- (mean). See below. 
N.B.—Mr. Galtox excluded from his calculations the larger families, but it seems to me that large 
families form an essential feature of the community. Two brothers are more likely to be two brothers 
of a large than of a small family, and, accordingly, large families ought to be given their proportionate 
weight. The whole problem, indeed, of reproductive selection turns upon the inclusion of large families. 
Explanation of the Tables. — These tables were calculated in the following manner: 
Table I. A father or mother appears once for each child in this Table. The mean 
heights of each group were then calculated, as well as their standard deviations 
(S.I).) or deviations of mean square. The probable errors of the means and standard 
deviations were then found by means of the formulae 
p.e. of M.H. = ‘674506 X S.D 
p.e. of S.D. = -674506 X S.D./v/2n, 
where n is the number of cases recorded in the second column of the Table. 
