RELATING TO SECONDARY SEXUAL CHARACTERS. 21 



uals may grade off into one class, in other parts into other classes. 

 An almost continuous series of types might be arranged from black 

 to pale yellow. 



The difficulty of matching the hen-feathered males to their genetic 

 mates is almost insuperable. In table 1 an attempt was made to put 

 these males with their respective females. The difficulty is, of course, 

 greater for the cock-feathered birds, even with the castration evidence 

 (that is too meager at present for the purpose) , but a few of the males 

 may be placed with certainty, and the rest guessed at. 



One bird appears to be a hen-feathered game male resembling in 

 many respects the female game, but darker and redder. There is more 

 shafting on cape and wing-bow. The breast is unusually dark-salmon. 

 The hackle is darker than is the game female. Upper wing-coverts 

 broadly laced with black. (Plate 10, fig. 3.) 



The occurrence of this hen-feathered jungle-fowl is so unique and 

 the coloration of the bird so interesting that I have added to the plates 

 three feathers of such a bird, viz., a stippled saddle feather, a feather 

 from the back, a hackle feather, and a wing covert with stippled center 

 and a black border. The neck hackle departs somewhat from the 

 hackle of the jungle-fowl hen, but in the same direction as does the 

 neck hackle of the Sebright cock from his hen. 



Looking over the F 2 group, the most noticeable thing is the large 

 number of blacks (E and G), all of which are stippled. Probably the 

 factor came from the game, because group E was present in the back- 

 cross as well as in F 2 , and because these black birds are always stippled. 

 The yellow color (land J) may have come from both, each breed having 

 then a black factor that, as a pattern, covers over most of the yellow. 

 It is difficult to distinguish penciling from stippling in the F 2 yellows. 

 Without figuring each of these types, their description in detail is not of 

 much value. The skins will be deposited for reference in the Zoological 

 Laboratory of Columbia University. 



C. Back-Cross of Fj to Game. 



As the back-cross of the Fj to the game might appear more likely to 

 reveal the kinds of germ-cells present in the individual, the results 

 from such a cross may be given before discussing the genetic data. If 

 it were certain that the "game" contained all of the recessive factors 

 that are involved in the experiment, this method of testing the result 

 would be ideal, but there is no way of determining a priori whether 

 this is the case. The question will be taken up later. The pre- 

 sence of two kinds of males with corresponding but largely uncor- 

 related differences in their plumage makes their classification as a 

 group impossible. It is simpler, therefore, to put the females into their 

 classes first, after which the hen-feathered males may be expected to 

 fall into the same groups (or nearly so), while the identity of the cock- 



