Correlation and Application of Statistics to Problems of Heredity 119 



kinship groups. Galton concluded that marking for different degrees of 

 noteworthiness would be a waste of energy in such a rough inquiry as that 

 he was undertaking. But I think it would have been of great interest had 

 Galton divided his material in another way, i.e. classified his F.Ii.S.'s into 

 the three categories of noteworthiness, and tested whether their kinsmen 

 had the same or different totals of marks. In other words he would have 

 answered the question of whether ability leading to noteworthiness is 

 inherited in quality as well as quantity. 



The next point is very important. Most men know beside their own 

 name that of their mother, i.e. her maiden name. Hence both the numbers 

 and achievements of the uncles and aunts in both paternal and maternal 

 lives are known and there is no difference of a sensible kind in Galton's 

 totals. This holds also for the achievements of the grandparental generation. 

 But when we come to the great grandparents and great uncles, there have 

 been further changes of name in fa me fa, me me fa, fa me bro and me me 

 bro, and Galton attributes the ridiculously low number of cases of note- 

 worthiness compared with those for fa fa fa and fa fa bro with a loss of 

 record owing to change of name. This probably has a good deal to do with 

 it, but it does not account for the successes of me fa fa and me fa bro, who 

 of course bear the mother's maiden name, being only half those of fa fa fa 

 and fa fa bro, who bear the father's name. I am inclined to think that the 

 factor of assortative mating to which I have referred on p. 1 09 is at least a 

 contributory cause. 



I now reproduce Galton's final table of results, to which I have added 

 percentages*: 



Numbers and Percentages of Noteworthy Kinsmen recorded 

 in 207 Returns of F.R.S.'s. 



* Obtained from Galton's assumption that we may take the mean of the extreme cases, i.e., 

 we multiply by - (— + —J = -3486. I prefer this to his actual method. 



