66 THE MENDEL JOURNAL 



If No. 1 ^ ^, the sister, is mated witli No. 2 B A 

 the brother, then the B sex-cells of the first may 

 meet the B and A sex-cells of the second, and give us 

 B B and B A respectively. Similarly, the A sex-cells 

 of No. 1 may meet the B and A sex-cells of No. 2 

 and give us B A and A A respectively. Thus we 

 shall obtain I BB + 2 BA + I A A. That is, on 

 the average in every four members of the offspring 

 of two B A parents, we sliall expect 1 pure black, 

 1 pure albino, and 2 impure or hybrid black indi- 

 viduals. And this is what experiment does give us. 



We have spoken so far of the segregation of these 

 two characters. But there is another feature which 

 we should note, and it is this : Segregation of charac- 

 ters alone would not be sufficient to explain the 

 experimental results. It is possible to have segrega- 

 tion and yet have both characters carried in the 

 same sex-cell ; they can lie side by side with each 

 other in the cell. But it is a fundamental part of 

 our proposition that any one sex-cell can carry one 

 only of the two alternative characters. To this 

 conception of the structure of the sex-cell we apply 

 the term " gametic purity." Necessarily, gametic 

 purity implies segregation. But segregation does not 

 necessarily imply gametic purity. We may now widen 

 the description of our principle of heredity, and 

 call it the theory of segregation and gametic purity. 



We have arrived, then, at this position : The 

 sex-cells are the carriers of the characters of the 

 race, from one generation to another. Whenever 

 two alternative characters, such as tallness and 



