Species crosses in Rats. 105 



Mated to their brother, one of the females produced in two litters: 

 8 blacks, 1 darkbellied agouti, 1 yellow whitebelly, 1 chocolate. 



These litters furnished us (luite a list of novelties, 'i'lie yellows 

 were to be expected from th(> experieuco of Dr. Bon bote. 



The silvercoloni-ed animals had a peculiar shade, which was de- 

 sciibed as dove-coloured by the ^^nglisli rat Fanciers, to whom we sent a 

 skin for inspection. It was very neaily like the colour described by 

 mousebreeders as plumsilver. By subseciuent brooding work it was 

 discovered, that this colour is related to black in the same way as 

 yellow to ag-outi. 



The origin of the novelties requires some comment. They (>ould 

 be described as due to mutation, but, apart from the fact that this word 

 does not mean very much if we are looking for causes, the numerical 

 relationship of the novelties to the old colours is significant. If these 

 novelties should be due to what the workers with Drosophila call "factor- 

 mutations" or point-mutations, they would be produced from heterozygotes 

 for this one factor, and the expectation should be one novelty in four. 

 The proportion in which the novelties appeared however varies, but is 

 much nearer to one in sixteen than to one in four. This is significant. 

 It points to the solution, that in every case the novelty appeared when 

 animals were born, which simultaneously lacked two genes, of which 

 one is present in- each pareutspecies, and responsible for the corresponding 

 dominant colour. The male must have been heterozygous for a good 

 number of sets of two genes, and his children from wild Rattus females 

 must have been heterozygous for several. Let us provisionally give 

 names to these genes. We will then call C and D the two genes which 

 are absent from yellow rats, but of which one at least must be present 

 in non-yellows, agoutis, or blacks. At first sight it looks as if it should 

 be easy to discover, whether a novelty is due to loss of one gene or to 

 simultaneous absence of two genes, as in the latter case a pro|)ortion 

 of fifteen normal to one novelty should be of fairly common occurrance 

 in crosses. But this is not the case. If non-yellow rats of the rattus 

 group possess either C or I), crossbreeding them with cd-yellows results 

 in a monohybrid ratio in F^. For the yellow mated to the animal with 

 C differs from it only in lack of D and vice versa. In our later work 

 we repeatedly found, that our novelties behaved as simple recessives in 

 crosses with wild rats. This is of great importance in the light of the 

 interpretation of other novelties, especially in the fruitfly. 



