190 Bulletin No. 155.— 1913. 



In conclusion, it may be said regarding the above cases that the- 

 Fi results of both Davenport's and Hurst's matings of W. L. stock 

 with various black breeds are fully explainable on the hypothesis 

 advanced above: that the W. L. cT cf are normally homozygous 

 for B and /, while the 9 9 are homozygous for I but heterozygous 

 for B. Many of the birds used by both Davenport and Hurst were 

 manifestly impure with respect to several factors. 



Cases la, 2a, 3a and 4a. — What now happens when the Fi cross- 

 breds from any of the matings presented in Cases 1, 2 and 3 are 

 mated among themselves? The white cross-bred cf cf as we have 

 seen possess the zygotic constitution 



C2Bbf2li 

 while the 9 cross-breds are 



C2BhFfIi 

 The cf forms gametes, 



CBfl ' CBfi • Chil ' Chfi 

 Since the 9 is heterozygous for three pairs of characters, eight 

 sorts of gametes might be expected : 



CBfl • CBFI ' ChFI • CBFi 



CBfi • Cbfl • CbFi ' Chfi 

 But it has been pointed out on a previous page that in the case of 

 the Barred Plymouth Rocks there is good reason for assuming that 

 the factors B and F never pass together into the same gamete; that 

 there is some sort of a repulsion between these factors. In tlie case 

 of the Barred Plymouth Rocks this results in black, or at least non- 

 barred 99 when B. P. R. 99 are crossed by the cf of a non- 

 barred breed. Since we have observed in the data already presented 

 a certain proportion of black 99 are produced in F2, we may 

 tentatively assume that in the White X Black crosses being described 

 there exists a similar incompatibility between the factors, B and F, 

 It may therefore be supposed that of the 8 sorts of gametes that might 

 be formed by the 9 cross-bred (with the zygotic formula, CiBbFfli)^ 

 there are actually formed only four; in other words, we may eliminate 

 from consideration all possible gametic combinations containing both 



