COLOR AND COAT CHARACTERS. 27 



Both the expected zygotic classes are here represented and in the 

 expected equaUty. 



The results of cross 2 are given in detail in table 14. Using the same 

 zygotic formula for the dark, wild, hybrid, ticked-bellied females as 

 was used in cross 1, their gametes should be A' and a. The males were 

 agouti guinea-pigs with light bellies, but heterozygous. Their zygotic 

 formula would be Aa and they would produce gametes A and a.^ 



The gametic combinations expected to occur in cross 2 may be 

 expressed as follows : 



A'+ a gametes of hybrid wild. 



A + a gametes of tame. 



AA'+ Aa + A'a + aa zygotes. 



2 + 1 + 1 visible classes. 



According to this scheme, three visible classes of offspring result. 

 Four real zygotic classes are produced. The first two zygotic classes, 

 AA' and Aa, look alike because the tame agouti is epistatic, as has been 

 shown in table 3 and in the previous discussion. The real difference is 

 shown by breeding these two classes to non-agouti animals. The class 

 A'a is composed of animals of the dark wild agouti pattern with ticked 

 belly. The class aa is a non-agouti class. The visible classes should 

 occur in the ratio 2:1:1. The numbers actually recorded are 19 : 13 : 4 

 The most probable expectation would be 18 : 9 : 9. 



An examination of the breeding records of the visible classes gives 

 the final proof of the actual existence of the expected zygotic classes : 



Visible class, AA'-\-Aa. — These two zygotic classes are alike light- 

 bellied light agoutis, because A is dominant to A' and to a. The 

 zygotic class, AA', should produce gametes A+A', and when mated to 

 non-agouti animals should give Aa+A'a; that is, tame and wild 

 agoutis in equal numbers, but no animals without agouti. The zygotic 

 class Aa, when mated to non-agoutis, should give equal numbers of 

 tame light agoutis and non-agoutis. Of course, not all animals could 

 be tested, and the numbers were cut down by the sterility of the males 

 as well as by premature death of some females. The zygotic class, 

 AA', is represented by 9 399, 9 448, and 9 499, which produced only 

 agouti offspring, but of two sorts, dark wild and light tame, in approxi- 

 mately equal numbers (table 15). The segregation of the two sorts 

 of agouti from each other was complete and definite. No trace of any 

 tainting of one agouti by the other was observable. The dark, ticked- 

 bellied young of table 15 were of the darkest shade; the light-bellied 

 young were like a normal agouti guinea-pig. There was no segre- 



^It is evident that I am using "a" as the allelomorph of both A' and A. This may need 

 explanation, for it may be urged that the allelomorph of A' is a' ; but since A' and A are allelomorphic 

 to their absence (tables 6 to 11) and this "absence" is one and the same thing (visibly and in 

 crosses), we may designate this absence by a single sjTnbol. 



