THE MENDELIAN PRINCIPLES OF HEREDITY 91 



in the same way as the flies ; thus the two Z's in the 

 red-eyed gray female are half black (for red) and half 

 gray ; the single X in the white-eyed yellow male is 

 half white and half stippled. 



In the Fi generation the X chromosomes are first 

 represented as they came in (second hne), i.e. with 

 their original composition. The next line gives the 

 three large classes that result, viz., 2 GR9 ~l GR$ 

 — 1 YW $ . But if free interchange takes place in the 

 female, some of the eggs will have chromosomes like 

 those in the fourth fine, viz. YR and GW . Such 

 eggs will give the classes represented in the lowest hne, 

 viz., 2 GR^~\ GW$~\ YR$. Thus, as already 

 explained, there results one kind of female and four 

 kinds of males. 



I said that the proportion 4:1:1:1:1 is the ideal 

 result in the cross between the yellow-white and the 

 gray-red flies. This ideal scheme is not realized because 

 of a complication that comes in. The compUcation 

 is due to Hnkage or a tendency to hang together of the 

 characters that go in together. We must now take up 

 this question. It is one of the most modern develop- 

 ments of the Mendehan theory — one that at first 

 seemed to throw doubt on the fundamental idea of 

 random assortment that gives Mendel's proportion 

 9:3:3:1. But I beheve we can now offer a reasonable 

 explanation, which shows that we have to do here with 

 an extension of Mendelism that in no sense invalidates 

 Mendel's principle of segregation. It not only extends 

 that principle, as I have said, but gives us an oppor- 

 tunity to analyze the constitution of the germ-plasm 

 in a way scarcely dreamed of two or three years ago. 



