Correlation and Application of Statistics to Problems of Heredity 99 



record making at least to the foals. In this way Galton obtained 716 sires 

 and 494 dams who had produced offspring satisfying the above conditions, 

 and he classified them by the number of times they had produced such marked 

 foals. Reduced to percentages of sires and dams the following table resulted : 



Percentage numbers of Standard Performers produced 

 by a single Sire or Dam. 



Galton explains the difference between sire and dam by remarking that 

 while the sire produces some 30 foals annually, the dam produces only one, 

 and therefore the chance of a large number of standard performers is much 

 less for her. He even allows that some of the exceptionally noteworthy per- 

 formances of the sires (Blue Bull, 60 ; Strathmore, 71 ; George Wilkes, 83 ; 

 Happy Medium, 92; and Electioneer, 154 standard performers) may be due 

 in part to the best mares being sent to famous sires. But he concludes 

 that the extraordinary "tail" of high-class offspring of the sires must be due 

 to some prepotency in some of the sires which enables them to impress their 

 character on their offspring, and he remarks : 



" My conclusion is that high prepotency does not arise through normal variation, but must 

 rank as a highly heritable sport, or aberrant variation ; in other words its causes must partly be 

 of a different order, or else of a highly different intensity, to those concerned in producing the 

 normal variations of the race. In a sport the position of maximum stability seems to be slightly 

 changed. I have frequently insisted that these sports or " aberrances " (if I may coin the word*) 

 are probably notable factors in the evolution of races. Certainty the successive improvements 

 of breeds of domestic animals generally, as in those of horses in particular, usually make fresh 

 starts from decided sports or aberrances, and are by no means always developed slowly through 

 the accumulation of minute and favourable variations during a long succession of generations." 



Here, I think, Galton has forgotten two things : 



(i) The average difference between the first and second individuals in a 

 group of 100 tabled to any character is no less than "36 of the variability of 

 the group, and in a random sample may be still higher, but this is no 

 adequate reason for treating the first individual (or the last) as a sport 

 because he is not, like mediocre individuals, practically continuous with his 

 neighbours. 



(ii) That the number of distinguished offspring any individual gives rise to 

 must be considered in relation to his total output. Galton merely says that 

 a sire produces " some 30 foals annually." I do not think this is adequate. 



Many years ago I saw a good deal of the working of a large thoroughbred 

 stud ; the stud contained a number of stallions, some famous for their racing 



* The word is quite good English, if Joseph Glanvill and Sir Thomas Browne are authorities. 



13—2 



