SOCIOLOGICAL CONSIDERATIONS 95 



the winning records.* Mr. Robertson's conclusion 

 that the mares, in some cases, carry some inherent 

 factor which inhibits their staying powers, therefore 

 seems to be provisionally justified. I think the 

 further facts which he has so clearly brought out 

 and which show a continual decrease in the proportion 

 of winning mares as the severity of the test (length 

 of the race) increases, is strongly in favour of his 

 view, f Thus in a race of five to seven furlongs in 

 length the ratio of winning mares and horses, though 

 slightly in favour of the horse, is nearly equal ! But 

 in more severe races of twelve furlongs and upwards, 

 the winning horses are nearly three times more 

 numerous than the mares. That is an extremely 

 interesting physiological fact. I do not know that 

 in any problem turning about the question of sex, 

 any more significant fact could be found. For 

 it seems to demonstrate the important general 

 deduction that the more severe the test of staying 

 power becomes, the more clearly the males and 

 females are differentiated from each other. 



Though this deduction is derived from a series 

 of very accurate experiments with horses — for the 

 racecourse is virtually a place where a supreme 

 experiment takes place — it is probably true of human 

 beings. It would be of great interest to put the 



* It is additionally necessary to bear in mind the fact, as pointed out 

 to me by Mr. Robertson, that if the mares A\ere good enough, there are a 

 sufiScient number in training to win every race run. But while compared 

 to horses they are but of little use as stayers, as sprinters they seem to 

 excel. For in purely sprint races (five furlongs length) they gain a greater 

 number of victories than their proportion to horses entering the race with 

 them warrants. 



fSee p. 56. 



