416 THE AMERICAN NATURALIST [Vol. LI 



as chocolate to black. It is to be remembered that in this 

 group black is dominant to agouti. In all we obtained six 

 wholly new characters from our matings, clearly as the 

 result of the absence of two genes in every instance. 



Matings of white-bellied animals with gray-bellied 

 gave either only white-bellied or a minority of gray- 

 bellied in Fj. Grray-bellied rats clearly have a gene less 

 than white-bellied. This is the same result which Morgan 

 obtained in his work with animals of this group. Black 

 was dominant over agouti and clearly there were two 

 kinds of blacks, with or without the gene which makes the 

 difference between white-bellied and gray-bellied agoutis. 

 We never obtained white-bellied black ones. But the 

 blacks with the gene under discussion had a much more 

 deeply black color, very often with a green or a violet 

 sheen. We obtained yellow-bellied yellows, and, just as 

 in the agouti series, white belly was dominant over yellow. 

 Male 28 and female 34, both white-bellied yellows, gave 

 three white-bellied and one yellow-bellied young. Our 

 chocolate and cinnamon rats died on the steamer bring- 

 ing them to Java. The character white tail tip proved to 

 be recessive. We obtained pearl-gray young and yellows 

 from matings between yellows and pearl grays, but yel- 

 lows never produced pearl grays. Two agouti animals 

 sometimes produced yellows, but never pearl gray. 

 These were only obtained when one parent was either 

 pearl gray or black. In other words, th© factors which 

 produce the difference between black and agouti animals 

 are the same which make the difference between pearl 

 gray and yellow. 



Our new rats, waltzers, and animals with new colors, 

 such as they are can not be called species. We can make 

 species out of them by continuing the breed. If we sell 

 a number of animals of one color to rat fanciers, and they 

 get sufficiently enthusiastic over them to provide classes 

 for them at pet-shows, we will be justified in calling such 

 a breed a domestic species. 



We saw that in our experiments with rats no new 

 dominant characters originated, unless we want to call 



