COAT CHARACTERS IN GUINEA-PIGS AND RABBITS. 45 



eliminated altogether, while others are kept at their maximum inten- 

 sity. How this is possible we could understand better if we knew in 

 what condition pigment characters are transmitted in the germ. From 

 the experimental evidence alone it would seem probable that the dif- 

 ferent color patches are represented in the germ by different elements 

 separately heritable, possibly in different chromosomes though of this 

 we know absolutely nothing and that as a result of long selection 

 these separately heritable elements may become firmly associated with 

 particular body regions. From the mode of development of the hair 

 pigments as described by Leo Loeb ('97) and confirmed by observa- 

 tions of Mr. Howard and myself, it is clear that the pigments are 

 formed in the Malpighian layer of the epidermis. A pigment patch, 

 then, is, morphologically, simply the pigmented epidermis of a particular 

 body region, and its character is determined by whatever determines the 

 character of the epidermis in that body region. 



That black-eyed white animals are essentially animals of the centrip- 

 etally pigmented type is shown by breeding them inter se or by mat- 

 ing them with albinos. In every case thus far I have obtained only 

 young with one to several, or even with a//, of the typical pigment 

 patches. The task of establishing a black-eyed white race which will 

 breed true is similar to that of establishing a Dutch-marked race which 

 will breed true, or in cattle a white-faced race like the Hereford. It 

 is not to be accomplished in a single generation, but I doubt not its 

 possibility. When guinea-pigs have been bred to a particular color 

 pattern for many generations, we can form a better estimate of the 

 magnitude of the task involved, and possibly meantime may learn 

 something about the mechanism of transmission of the color patches. 



The specific pigments transmitted by black-eyed white animals are 

 probably determined by the same rules that govern in the latent pig- 

 mentation of albinos. These pigments probably differ according to the 

 parentage in different cases, or more correctly according to the latent 

 pigmentation which each individual received from its parents. Indeed, 

 it is not impossible that vanished pigment patches are merely latent 

 pigment patches, and that the reason why these keep reappearing in 

 such a seemingly erratic fashion is that they keep emerging from 

 latency under the influence of cross-breeding. In the black-eyed white 

 animals with which I have experimented, red as well as black spots 

 are apparently transmitted in the gametes formed, but I should expect 

 that in animals different in origin red might be transmitted apart from 

 black or vice versa. 



I have not examined with the microscope the eye-pigments of the 

 different self-colored varieties of the guinea-pig. From a superficial 

 examination, however, I should say that black pigment is present in 



