44 



MASS STUDIES IN BUILD. 



On the hypothesis of 6 zygotic factors for build, the possible com- 

 binations in the progeny are much more numerous. Seven classes of 

 zygotic combinations are possible. We recognize in our work only 5 

 classes of build. Accordingly, it would be necessary to redistribute 

 our classes of build into 7 or else assume that the two lowest classes 

 are both comprised in "very slender" and the two highest in "very 

 fleshy." The former operation would require an amount of work 

 hardly justified by the possible advantage; so the latter procedure 

 was adopted as perhaps a sufficiently close approximation. The dis- 

 tributions are given in table 18 which is given in detail only in part. 



MATE SELECTION IN BUILD. 



Statistics on temperament and stature of consorts (Davenport, 

 1915, p. 106; 1917, p. 329) seem clearly to show that there is an 

 assortative mating in respect to these traits. The question arises: 

 Is there assortative mating in respect to build? The inquiry is ren- 

 dered the more difficult, inasmuch as build changes to such an extent 

 with age. Nevertheless, as there appears to be a considerable corre- 

 lation (though not yet calculated) between build at 25 and at 50 

 years, it is fair to assume that some degree of the mature build is 

 already indicated at the period just before marriage. 



Table 19. — Percentage distribution of parents of each sex among the various classes 

 of build as found in 531 selected matings. Based on Appendix tables. 



If, now, there is no assortative mating in respect to build, we should 

 find that persons of any given build, say slender, would have very 

 slender, slender, medium, fleshy, and very fleshy consorts in the respec- 

 tive proportions in which such classes of build occur in the whole 

 population of parents. A marked deviation from this expectation 

 would indicate the falseness of this hypothesis and that there is an 

 assortative mating in respect to build. 



To test the hypothesis we can make use of 531 matings, including 

 those which are employed in the main tables. We find the male and 

 the female consorts in these matings to occur in the different classes 

 in the numbers and proportions shown in table 19. 



