VARIATION AND HEREDITY 17 



number of candidates is large, the marks are found 

 to group themselves round a mean value in just the 

 same way as do the figures expressing the stature of a 

 number of men of the same race. Most candidates 

 obtain marks in the neighbourhood of 50 per cent, of 

 the total, while fewer and fewer candidates are found 

 as we get nearer to zero at one end and to 100 per cent, 

 at the other. If the results group themselves in an 

 irregular manner either at one end or the other of the 

 scale, it is fair to assume that either the papers, the 

 candidates or the examiners were unsuited to the 

 occasion. 



To estimate the intensity of inheritance in these 

 mental characters, the positions in the Oxford Univer- 

 sity class-lists of a large number of fathers and their 

 sons were compared, and the relative position in the 

 forms of public schools of brothers, between whom, of 

 course, there should be correlation if heredity is strong, 

 since they have the same parentage. In the case both 

 of parental inheritance and of fraternal relationship, 

 the coefficients of correlation were found closely to 

 agree with those for physical characters. Thus we 

 obtain one class of evidence bearing on the important 

 result that mental qualities are inherited in the same 

 way, and with the same intensity, as are physical 

 characters. We may not be able to predict the mental 

 powers of any given family in the same exact way in 

 which we can foretell its probable composition with 

 regard to the physical characters for which definite 

 Mendelian inheritance has already been made out ; but, 

 if we consider, not one family, but a large number of 

 families, the results statistically are no less accurate and 



2 



