GENETIC FACTOR FOR HEN-FEATHERING. 259 



the game type is expected in the back-cross in one fourth of the 

 classes, i.e., in 3 out of 12, which agrees fairly well with the re- 

 sults. For three factors for color, the game type is expected in 

 the back-cross once in eight times. The numbers are too small 

 to be decisive for either case, but as pointed out before, the main 

 color difference between the two races may not be more than 3 

 or less than 2 factor-pairs. 



In the earlier experiments one game appeared in F 2 out of 49 

 females, which is somewhat favorable to the assumption of three 

 factor-pairs. The new data are not sufficient in amount to settle 

 the question between two or three factors, but the latter is per- 

 haps somewhat more probable. 



Proof that the Character for Hen-Feathering is not Sex-Linked. 



The proof that the factor (or factors) for hen-feathering in- 

 volved in this cross is not carried by the sex-chromosome can 

 not be deduced from the evidence of the F x males in the cross 

 and its reciprocal, for, both the F males would get a Z chromo- 

 some from each parent, and therefore would be alike in both 

 cases. But in the F 2 generation the proof is given, for when the 

 cross is made one way, only hen-feathered males are expected, 

 and when made the other way (reciprocal) one hen-feathered to 

 one cock-feathered male is expected, if the gene concerned is 

 sex-linked. The F 2 results in both crosses show, however, both 

 hen-feathered and cock-feathered males; hence the factor (or 

 factors) involved is shown not to be carried in the sex-chro- 

 mosomes. 



