40 



15 males. The figure used for each male represents the mean for all of his mates. 

 These males were used on the following years: — 2 in 1913 and 1914, 4 in 1914'and 

 1915, 2 in 1915 and 1916, 2 in 1916 and 1917, 1 in 1917 and 1918, 2 in 1919 and 

 1920, 2 in 1922 and 1923. 



Mean first year, .6128±. 016043; Mean second year, .6580 ±.01 5825; Difference in 



means, .0452±.0225. 



Although the data are meager in table 17, we can give it no other interpretation 

 than as indicating that the male does partly control the hatching power of his mate 

 through dominant factors. 



The mean hatchability for the fifteen males during the first year is .6128zt 

 .016043, for the second year .6580 it .015825. There is a difference of .0452±.0225. 

 This difference is just double its probable error and can therefore be of no conse- 

 quence. The point we wish to emphasize in table 17 is the striking constancy in 

 phenotypical hatching power of the same male, even when mated to different hens 

 on two successive years. Such a degree of constancy was not found to exist in 

 hens, as table 3 shows. The mean pullet-year hatcliing power of the hens was 

 .5678zb.011313. The mean second-year hatching record of the same hens was 

 .4791 ±.012963. The standard deviation is nearl.y three times as great for the hens 

 as for the males. The difference in the mean hatching power for the same hens on 

 two successive years is .0887±.0172, which is significant. The genetic interpreta- 

 tion given below will serve to elucidate several apparent complications. 



Genetic Factors Concerned ^ 



One dominant gene seems to be concerned in the production of high hatchability. 

 We use the symbol H to designate this gene. There is no sex linkage and all results 

 obtained are to be expected in a simple mono-hybrid ratio. With this hypothesis, 

 three possible genotypes of males and females exist, namely, HH, Hh, and hh in-' 

 dividuals. The genotype is obscured in most cases for both males and females. 

 Such being the case, onl}^ the breeding test can be used as a guide for matings. 



Hatching records on 886 females studied in this report show that these birds 

 divide themselves into three general classes or phenotypes: — (1) Those showing 

 hatchability of 85 per cent or above, we call high. (2) Those with a hatchability 

 of 55 to 84 per cent, we call medium. (3) Those below 55 per cent, we call low. 

 Since factor H has a cumulative effect, the range for the medium class is twice as 

 great as for the high class. The minimum for the low class has not yet been de- 

 termined. Below are summarized the males' phenotypical and genotypical classes: 



1 A detailed report on the genetics of hatchability will appear in another publication. 



