HUMAN GENETICS 335 



are made as to the degree of dominance shown by the genes in ques- 

 tion, and as to the frequency of the genes in the population. It turns 

 out that the assumptions which are necessary to derive, from Men- 

 delian principles, correlation coefficients similar to those actually 

 found, are quite reasonable ones. The sort of correlations which are 

 found, therefore, give no reason for denying the Mendelian analysis. 



But actually very httle can be deduced from correlation coefficients 

 of this kind, because they are affected by too many enviroimiental 

 factors.^ The biometrical assumptions assume that all genotypes develop 

 in the same environment, and this is certainly not true. If we consider 

 a correlation between brothers, there will be differences between the 

 environments of the two brothers of a pair, and the greater these 

 differences are, the less alike will the brothers be and the lower the 

 correlation between diem. On the other hand, the environment of one 

 pair of brothers will differ from that of another pair, and the greater 

 these differences, the more boys of one family will resemble each other 

 rather than boys of another family, and the greater the correlation 

 between brothers will be. The actual correlation found will therefore 

 be the resultant of the true genetical correlation in a uniform environ- 

 ment, with the differences within a family tending to lower it and 

 differences between famiUes tending to raise it. There is no way of 

 assessing how important these environmental factors have been in a 

 given case, and even if we find exactly the correlation which would be 

 expected on certain genetical assumptions, we carmot jump to the 

 conclusion that those assumptions are justified. 



6. The Genetic Analysis of a Continuously Varying Character: Intelligence 

 We cannot derive any insight into the actual genetical situation 

 unless we can find some way of simplifying the usual complicated state 

 of affairs. As an example of what can be done in this way, we shall take 

 the analysis of human intelligence.^ 



The range of variation is apparently quite continuous between idiots 

 and geniuses. The first necessity is to attempt some method of classifi- 

 cation. In the lower grades, some more or less specific types of idiocy 

 or insanity can be distinguished. Thus Huntington's chorea is a disease 

 involving progressive mental degeneration; it has been clearly shown 

 to be due to a simple autosomal dominant. Amaurotic family idiocy is 



^ Cf. Hogben 1933. 



2 General references: Freeman 1934, Hogben 193 1, 1933, Holmes 1934, 

 Penrose 1936. 



