46 



THE INHERITANCE OF COLOR IN MICE. 



On a basis of association of characters both classes (2) and (3) should be 

 in excess. The observed result follows: 



It will be seen that while type (2), pink-eyed black, is slightly in excess, 

 the browns (which must also on the association hypothesis be in excess) are not. 

 The sum of classes (2) and (3) is 136, while on the supposition of random segre- 

 gation we should expect 132.6. 



IV. Cross 96. Brown Crossed with Pink-Eyed Dilute-Brown. 



Brown possesses both factors D and P; pink-eyed dilute brown has neither 

 of them and is djp in formula. 



Brown DP x Pink-eyed dilute broivn dp, 



Fi Brown, DdPp. Fi (1) Brown, DP, (2) Dilute brown, dP, 

 (3) Pink-eyed brown, Dp, (4) Pink-eyed dilute brown, dp. 



On an association hypothesis both classes (1) and (4) should be in excess. 



Classes. 



(1) 

 Brown. 



Observed . 



Expected on a basis of random seg- 

 regation 



49 



41.4 



(2) 



Dilute 

 brown. 



21 



13.8 



(3) 



Pink-eyed 

 brown. 



13.8 



. (4) 

 Pink-eyed 



dilute 



brown. 



4.6 



Sum of 



classes (1) 

 and (4). 



50 



46 



The sum of classes (1) and (4) is onlj^ 4 in excess of the expectation on a 

 basis of random segregation. This may be considered as showing no associ- 

 ation, because class (4) is not in excess of the expectation. If now we add 

 together the numbers of individuals observed in the classes that should show 

 "association," if it exists, we have 225, while the number called for by the 

 mendelian hypothesis of random segregation is 220.6, a striking approximation 

 to the observed numbers. We may then conclude that these experiments show 

 no sign of association in the gametes between the factors D, P, A, or B or 

 between the four conditions characteristic of their absence. 



