300 F. B. SUMNER AND H. H. COLLINS 



Further evidence for the same interpretation of these color 

 varieties is derived from back-crosses. Matings between Fi 

 hybrids and 'a' yellows resulted in ten offspring. Five pre- 

 pared skins are available from this lot. These give us 3.2 as 

 the mean ratio of red to green — a figure lower than all but four 

 of the twenty-four 'h' yellows and almost identical with the 

 average of the 'a' race. Notes were made, furthermore, upon 

 the other specimens, either when living or freshly killed. These 

 indicate that they were, for the most part at least, of a dull 

 yellow or buff appearance. In some cases they were expressly 

 likened to the 'a' yellows. 



Matings of some of the foregoing back-cross individuals inter 

 se — ((a-b) - a) - ((a-b) - a)^^ — resulted in the birth of thirteen 

 young, all of which were listed as 'yellows,' though they are 

 described in much the same terms as their parents. The ground- 

 color was a very dull yellow or buff, darkened by a considerable 

 admixture of black hairs. They resembled the duller specimens 

 of the 'a' yellows, and in no case approached the more ruddy 

 hue of the 'b' strain. 



The other .back-cross (i.e., between the Fi hybrids and the 

 'b' yellows) gave a quite different result. Of the two mature 

 skins which were preserved, both are closely similar to the bright- 

 est 'b' yellows in appearance, giving red : green ratios of 3.5 and 

 4.0, respectively. These lie altogether outside the range of the 

 'a' series, and the larger figure almost reaches the extreme for 

 the '6' series. Furthermore, notes made upon the entire lot 

 (thirteen in all), when living or freshl}^ killed, show that at least 

 six resembled the '6' strain more nearly than the 'a,' while several 

 others are listed as 'intermediate.' There is no record of a speci- 

 men's having the predominant appearance of the 'a' strain. 



" We have employed hyphens instead of multiplication signs in designating 

 these various crosses, since reciprocal crosses were commonly made, and we wish 

 to be non-committal as to which parent belonged to which race. When the 

 multiplication sign is used, it is commonly understood that the female parent 

 is named first. 



