30 



INHERITANCE IN GUINEA-PIGS. 



These are capable theoretically of forming 16 different combinations, 

 as follows, heterozygous combinations being omitted. The appearance 

 of zygotes containing each of these several combinations is indicated 

 opposite the respective combinations. 



ABCe, 

 ABCrE, 



ABC r e, 

 AbCrE, 

 abCE, 



AbC r e, 

 aBC r e, 



ABCE, golden agouti = wild type. 



Single mutations. 



AbCE, 

 aBCE, 



Double mutations. 



black-eyed red. 

 silver agouti. 



cinnamon, 

 black. 



red-eyed white, 

 silver cinnamon, 

 chocolate. 



AbCe, brown-eyed red. 

 aBC r E, red-eyed sepia, 

 black-eyed red 



red-eyed white, 

 red-eyed white. 



aBCe, 



Triple mutations. 



abCe, brown-eyed red. 

 abC r E, red-eyed chocolate. 



Quadruple mutation. 

 abC r e, red-eyed white. 



Of the 16 different combinations, 2 produce black-eyedred individuals 

 indistinguishable except by breeding test; the same is true regarding 

 brown-eyed reds. Four other combinations identical with these, except 

 for the substitution of C r for C, produce red-eyed whites, which visibly 

 are all alike but which breed differently. Three of the four kinds of 

 red-eyed whites have been identified by breeding test; no doubt the 

 fourth can easily be obtained. The fact that the several classes of 

 red-eyed whites look alike, and that the two kinds of black-eyed reds 

 look alike, and further, that the two kinds of brown-eyed reds look 

 alike, reduces the number of visibly distinguishable classes from 16 to 

 11, all except one of which have been recorded from this single experi- 

 ment. The experiment also produced albinos which theoretically 

 should be of 8 different formulae, if in the formula C a is everywhere 

 substituted for its allelomorphs C or C r . No attempt has been made 

 to distinguish the several expected classes of albinos by breeding tests, 

 the only certain means of identifying them. 



5. The close agreement observed between theoretical and recorded 

 numbers of F 2 offspring in this cross lends no support to the idea that 

 any association or linkage occurs among the 4 factorial variations 

 involved. 



