I06 THE PEEBI^Y INHIBITED. 



VII. SELECTION OF TEMPERAMENTS IN MARRIAGE. 



Of the 29 classes of matings recorded in table C, some contain many 

 more entries than others. Thus classes i, 3, 4, 5, 7, 8, 12, 26, and 29 

 include each only one mating, while classes 21 and 24 contain 22 and 20 

 respectively. Why this inequaUty in the frequency of matings in the 

 several classes? There are various hypotheses that are tenable. Thus 

 (a) since, as table A shows, in random matings certain gametic combi- 

 nations will probably be less common in the whole population than 

 others, such rare factor combinations may be expected to occur rarely 

 in mating combinations. Such relatively rare combinations are E2C2, 

 E2C2, e2C2, and e2C2; obviously the imion of two individuals, both of 

 whom are of these least common types, should be least common. Of 

 the 9 classes which contain only i entry each, 3 are of matings of these 

 relatively rare combinations. May not the variations in frequency of 

 matings in the various classes be fully accounted for by a diEference 

 in frequency of the parental zygotic combinations? The test of this 

 hypothesis will be found by noting if all matings of parents belonging 

 to classes of equal frequency are equally common or nearly so. This 

 comparison of theoretical and actual frequency of matings has accord- 

 ingly been made and results are given in table E. 



Table E demonstrates, indeed, as might have been anticipated, that 

 matings occur more rarely between two persons who carry the rarer 

 zygotic combinations than between those that carry the commoner 

 combinations. Indeed, except for the fact that group i in table E has 

 an unexpectedly small number of matings and group 5 a somewhat 

 unexpectedly large number of matings, the number of the matings in 

 the (hflferent groups is proportional to the product of the numbers 1,2, 

 4, which are the most probable frequency — ^proportions of the least, 

 mid, and greatest frequency, respectively, of the given zygotic com- 

 binations as shown in table A. Perhaps the most striking fact is that 

 the mating e2C2 X e2C2, or both parents calm-cheerful (or normal), is 

 not represented once in otur 146 matings. This is, however, easily 

 tinderstood when the nature of our material is considered; the great 

 mass of it starts with institutional cases of excitement or depression 

 (or both), and a search was made in the volunteer records of such as 

 showed such extremes of emotional disturbance. The zygotic factors 

 of both parents of the disturbed persons are, by hypothesis, not normal. 



When, now, one examines the relative frequency of matings inside 

 of one group, in place of the expected equality of the frequency one 

 finds great diversity. Part of this is chance; some of it strongly 

 suggests a selection in mating based on temperament. Thus in group 2 

 the mating EeC2 X E2C2 (nervous-cheerful mated to choleric-cheerful) 

 is far the commonest, the mating of a choleric-cheerful to a nearly 

 normal (e2Cc x E2C2) and of a nearly normal to a choleric-melancholic 

 (manic-depressive) are next in order. 



