SEGREGATION OF GERM-CELLS 349 



Then, if the fertilisation is fortuitous, the possibilities are — 



50 QQ + SofD)(R) + SO (d)^) + So(i^(r)» 



resulting in adults in the proportion, 

 I D + 2 D(R) + I R 



But even if we suppose that the germ- cells are all of one kind — 



viz. with both dominant and recessive characters in potentia — 



but that in some the dominant primary constituents gain the 



ascendency and that in others the recessive primary constituents 



gain the ascendency, chance fertilisations might still result in the 



Mendelian kinds and proportions of offspring. 



It may be that gametes which behave as if they were the 

 vehicles of only one character have the other in reserve. For 

 while purity means that only one character is expressed in 

 development, it may be that the other character is there all 

 the time in a latent state. 



Prof. Weldon (1905) pointed to the analogy of two blastomeres, 

 lying side by side in the 2-cell stage of the development of an 

 ovum. Normally one blastomere will develop into all the right 

 side, the other into all the left side of the embryo. They may 

 be described as dominant in these respects. But a dislocation 

 of the two cells may separate them, and each may develop into 

 a complete embryo. In other words, characters which would in 

 normal conditions have remained latent may as the result of 

 some shock become patent. There are many exceptional results 

 in Mendelian inheritance which suggest that the purity of the 

 gametes is not so thoroughgoing as the theoretical Mendelian 

 interpretation suggests. 



Furthermore, with every desire to follow out the simplest line of 

 interpretation, we must not forget that, even if all the fertilised ova 

 started alike, with both dominant and recessive primary constit- 

 uents, the expression of these in the course of development might 

 result — especially if there be a real struggle among homologous 

 determinants — in a victory here for the dominant characters, there 

 for the recessive characters, and a sort of compromise in a third set. 



