148 HEREDITY [ap. 



in more than one allelomorphic pair, all possible 

 combinations can be produced, except that no germ- 

 cell can contain both the members of one pair. 



The suggestion that this segregation of chromo- 

 somes, which can be seen to take place, is the 

 mechanism by which the members of an allelo- 

 morphic pair of characters are segregated, is quite 

 speculative; but it seems exceedingly unlikely that 

 machinery so exactly adapted to bring it about should 

 be found in every developing germ-cell, if it had no 

 connexion with the segregation of characters that 

 is observed in experimental breeding. There is also 

 the further fact in support of the suggestion, that 

 it is known in many insects that one pair of chromo- 

 somes is closely connected with sex, for in the males of 

 these species one chromosome of the pair is absent or 

 much reduced, but in the female both are similar. 

 These sex-chromosomes separate from one another 

 like the others (when both are present), and it has 

 been seen that there is experimental evidence for the 

 view that the sex-determiners behave like Mendelian 

 allelomorphs. One serious diflSiculty however suggests 

 itself at once ; the chromosomes are limited in number, 

 and it is undoubted that more allelomorphic pairs of 

 characters may exist in a species than there are pairs 

 of chromosomes, although in such cases there is no 

 evidence that members of different pairs are always 

 associated together. Several suggestions have been 



