34 On Inheritance of Weif/hf in Poultvii 



occur ill our F.. generation at all. W'c regard this as due to ihr fact 

 that an insufficient number of birds was bred in F., to give them. 

 Where F^ birds are heterozygous lor tour factors, the two extremes, i.e. 

 that homozygous for all four factors and that lacking all four, will only 

 occur once in 256 birds. The extremes were however readily obtained 

 in Fs and we have little doubt that we should have obtained theni in F., 

 had we been able to breed a sufficiently large number of biixls. 



Bearing these points in mind we have constructed a scheme in- 

 volving the following assumptions. 



(rt) That we are dealing with a case in which we are involved with 

 the presence or absence of four genetic factors each of which affects the 

 weight of the bird. 



(6) That when none of these I'actors are present the birds are of 

 the minimum size. 



(c) That two of these factors, A and B, i)roduce an increase of 

 66 °/o respectively on the minimum size when the birds are homo- 

 zygous and 38°/o when the birds are heterozygous for eithe'r of them. 



(d) That the remaining two factors, C and D, produce an increase 

 of 30°/j, respectively when the birds are homozygous, and 25 "/^ when 

 the birds are heterozygous. 



(e) That the constitution of the Gold Hamburgh is AABBCCdd 

 and of the Silver Sebright aabbccDD. 



If we take the minimum weight as 100 for birds of the consti- 

 tution aabbccdd, the maximum weight for birds of the constitution 

 AABBCCDD becomes 1 00 -t- 66 + 6(i + 80 + 30 = 292. Hence F, birds 

 of the constitution AaBbCcDd, themselves of grade 226, should when 

 bred together give a generation ranging from gi-ade 100 up to grade 

 292. Since four factors are concerned the extreme members of the 

 series will only make their appearance in this generation once, on an 

 average, among 256 birds. Table II shows the proportion and consti- 

 tution of these 256 birds which go to form the F^ generation. Opposite 

 the constitution in each case the weight grade is given, and in the 

 other two columns these theoretical weight grades have been translated 

 into expected actual weight on the basis of the range of variation as 

 found by experiment in the F,, and F^ generations. For example the 

 smallest ^ obtained in Pen 13, 1913, weighed about 550 grams and the 

 largest bred in Pen 12, 1913, reached a weight of rather over 1600 grams 

 (cf. Table I, p. 26). In order to obtain the actual weights of the 

 females as they should be on our scheme we have multiplied the male 



