i.] ANTECEDENTS. 29 



The foregoing tables show results bearing on 

 the question whether harmony or contrast pre- 

 vails in the physical characteristics of the parents. 

 I think they must be accepted as decidedly in 

 favour of harmony. The grand totals which they 

 give are 78 cases of harmony, 31 of contrast, 

 and 56 of indifference. In short, there is more 

 purity of breed in scientific men than would 

 have resulted from haphazard marriages. In the 

 temperaments of their parents, harmony strongly 

 prevails over contrast, the proportion being 5 to 

 1 in favour of the former. In colour of hair, 

 harmony is twice as frequent as contrast. In 

 figure it is equally common, because " corpulent, 

 stout, or plump " persons of one sex seem to have 

 a peculiar and reciprocated liking for " spare, 

 neat, or small " persons of the other. This is 

 literally the only case in these tables where a 

 love of contrast equals that of harmony. I came 

 to much the same conclusions by giving appro- 

 priate marks for harmony, contrast, and indif- 

 ference to each quality in each case, thus 

 obtaining aggregate marks for every pair, which 

 I treated on much the same principle that the 



