i.] ANTECEDENTS. 71 



sense public men ; and (3) the number of those 

 who, not being publicly known, had never- 

 theless considerable reputation among their 

 friends. It is therefore only requisite (after 

 some previous revision) to add the returns 

 together, and to compare the number of dis- 

 tinguished kinsmen in the various degrees with 

 the total number of kinsmen in those degrees, 

 to obtain results whose ratio to one another is 

 the one we are in search of. These conclusions 

 are not materially vitiated by the fact that 

 different correspondents may have different esti- 

 mates of what constitutes distinction, so long as 

 each writer is consistent to his own scale. I 

 have tried the figures in many ways without 

 any revision at all, with moderate revision, and 

 with careful sifting, and I find the proportions 

 to come out much the same in every case. In 

 comparing these with previous results, obtained 

 from an analysis of men of much higher general 

 eminence ("Hereditary Genius," p. 317), I 

 find the falling off in ability from the central 

 figure, the hero of the family, to be less rapid 

 as the distance of the kinship increases. There 



