58 THE FAMILY AND THE NATION 



such examinations are more or less selected. It is 

 doubtful how far the normal curve would represent 

 the facts of the actual distribution of general ability in 

 such a heterogeneous nation as our own, with so many 

 different types in the population. We might find, 

 could we make actual measurements, that the experi- 

 mental curve was much distorted. Still, broadly con- 

 sidered, the results would be of the kind indicated by 

 Galton ; the different types of the people mingling 

 to some extent with each other. In dealing with 

 specialized ability, particularly in selected samples of 

 the population, a near approach to the theoretical 

 normal curve would be expected, and is found where 

 measurement is practicable. 



Galton's method enables us roughly to classify the 

 differences in ability of the nation. He also showed 

 that another estimate of differences in ability of one 

 special kind may be obtained from the results of com- 

 petitive examinations, and especially from the old Cam- 

 bridge Mathematical Tripos, in which the names of 

 successful candidates were arranged in order of merit. 

 All these candidates are good enough mathematicians 

 to obtain mathematical honours, and even the worst of 

 them must be considered as of moderate ability. Yet 

 the Senior Wrangler usually got more than thirty times 

 as many marks as the lowest man on the list, and, 

 owing to the limited time allotted to the examination, 

 it is allowed that the results under-estimated the differ- 

 ences. The candidates were trained in similar con- 

 ditions, at all events for the last three, the most 

 important, years. Hence it follows that we must 

 regard the ability of an average Senior Wrangler as 



