50 HELEN DEAN KING 



individuals of later generations, which will be given in a subse- 

 quent publication, show a still further decrease in the variabiUty 

 of body weights as the inbreeding advanced. 



As the coefficients of variability were not determined for any 

 individuals in the first six inbred generations, the extent of varia- 

 bihty in the body weights of the inbred series as a whole is not 

 known. A comparison of the average coefficients for body weight 

 as given in table 15 and in table 17 show that the animals in the 

 seventh to the fifteenth inbred generations were, as a group, 

 about 7 per cent less variable in body weight than the animals 

 in the control series. This difference would doubtless be greater 

 if the inbred rats had not undergone unusual changes in body 

 weight due to altered conditions of nutrition. 



On the basis of the calculations made by Fish ('14) regarding 

 the amount of homozygosis in different generations of animals 

 produced by brother and sister matings, the group of rats com- 

 prising the seventh to the ninth inbred generation was, on the 

 average, about 83 per cent homozygous. These animals, as 

 table 16 shows, had a range of variability in body weight that 

 was greater than that in any other generation group. It is in- 

 teresting to note that, although the body weights of this group 

 of inbreds greatly exceeded the norms for stock animals of like 

 age, their variability was not correspondingly increased, since the 

 average coefficients for the group are about the same as those 

 for the stock controls (table 17). The group comprising the in- 

 dividuals of the tenth to the twelfth inbred generations was, 

 according to Fish's table, about 91 per cent homozygous, while 

 the last group was, on the average, 95 per cent homozygous. 

 Since body weight probably depends to a considerable extent 

 on 'Hhe presence or absence of definite genetic factors segre- 

 gating from one another in gametogenesis on lines with which 

 we are already familiar" (Punnet and Bailey, '14), one might 

 expect to find a definite correlation between the amount of homo- 

 zygosis and the variability in body weight in animals obtained 

 from brother and sister matings. On referring to the average 

 coefficients for the body weights of the various generation groups, 

 as given in table 16, it is found that for both males and females 



