118 



Life and Letters of Francis Galton 



It may be difficult to get adequate appreciation of women's noteworthi- 

 ness, but it is still more difficult to measure heredity in ability, unless we 

 have some direct measure of whether ability can be transmitted through the 

 mother with strength equal to that of transmission through the father. We 

 know whether the father was or was not noteworthy, but if we have no 

 measure of the ability of the mother, we cannot determine whether an able 

 maternal stock transmits its ability equally through an able and through a 

 mediocre woman member. Further Galton does not discuss the sons of 

 Fellows as many might not have reached maturity ; 467 persons were 

 addressed, 207 of these sent serviceable replies, of which only G5 are treated 

 in Schuster's list of noteworthy families of F.R.S.'s (pp. 1-79). Galton's 

 data are numerically based on the 207 cases. He states that the real crux 

 of the problem lies in what the remaining 260 were like. Abstention might 

 be due to dislike of publicity, to inertia, or to pure ignorance; such causes 

 would hardly affect the randomness of the sample, but if the 260 did not 

 reply because they had no noteworthy kinsfolk this would influence the 

 sample, and badly influence it. The two extremes are that (a) we suppose 

 the 260 to share the richness of the 207 in noteworthy kinsfolk, (b) we con- 

 sider that the 260 had no noteworthy kinsfolk. Galton says he cannot guess 

 which of these hypotheses is the more remote from the truth, but considers 

 that actuality cannot be very far removed from their mean value. I cannot 

 find, however, that this is what Galton has really used. For example the 

 F.R.S.'s had 81 noteworthy fathers. The percentage of noteworthy fathers 

 on the first hypothesis is 81 x %$$ = 39 - 13 and on the second hypothesis is 

 81 x 4^$ = 17 "34; thus the mean of the two is 28 "24. Galton, however, does 

 not take the mean of the two hypotheses, but of the numbers 207 and 467, 

 and gets 337; then he finds M ^ u> - = 24-04, and this is the percentage he 

 actually uses. Taking 1 man in 100 as noteworthy — a somewhat arbitrary 

 assumption — he states (p. xl) that F.R.S.'s have 24 times as many note- 

 worthy fathers as the generality of men. Before we pass to Galton's final 

 table we may cite one or two points he makes which are of distinct 

 interest and importance for similar investigations. In Chapter IX he gives 

 the result of marking individual degrees of noteworthiness ; he made three 

 categories and gave to them in degree of noteworthiness marks 3, 2, 1. He 

 then reduced the total of marks for each degree of kinship (657) to the 

 total number of cases of noteworthiness (329). As a first appreciation the 

 two results differed very little ; thus (p. xxxvii) : 



The reason for this approximate concordance lies in the distribution of 

 triple, double and single marks being much the same in the different 



