348 EXPERIMENTAL STUDY OP INHERITANCE 



more stable — namely, by separating out incompatibilities (domin- 

 ant and recessive potential unit characters) and becoming the 

 vehicles of either the one or the other ? The adult organism 

 cannot have both ; intermediate forms or blends do not occur 

 in these Mendelian phenomena. Are there differential di- 

 visions during the development of the germ-cells which lead 

 to there being two camps of gametes which we may briefly 

 describe as pure potential dominants and pure potential re- 

 cessives ? Is this not a possible expression of a struggle among 

 hereditary items or homologous determinants, and in line with 

 Weismann's theory of germinal selection ? 



A more precise suggestion, to which it seems too soon to attach 

 great significance, is the fascinating hypothesis that the segregation 

 occurs during the maturation divisions. If we assume that the 

 chromosomes are the vehicles of the hereditary quahties, which 

 seems highly probable ; if we assume, further, that a particular 

 potential unit character is contained in each germ-cell in one chromo- 

 some, and not in others, which seems a difficult assumption : 

 then it is possible that Sutton may be correct in his suggestion that 

 the segregation of gametes into two sets occurs in the course of the 

 maturation divisions. (See T. H. Morgan, Experimental Zoology^ 

 1907, p. 72.) 



Is it necessary to assume a Segregation of" Pure " Gametes ? 



— Mendel's theoretical interpretation of his results was based, 

 as we have seen, on the fascinatingly simple assumption of the 

 segregation of the germ-cells of the hybrid offspring into two 

 contingents, each set the vehicle of only one of the two antithetic 

 characters. To repeat the scheme : 



D?R,^ D?R(?. Pure-bred parents. 



D(R) $ D(R) ? . Hybrid offspring, 



producing producing 



sperm-cells, egg-cells, 



T T 



say 100 ( D J and 100 ( R \ say 100 ( D ) and 100 I R )» 



