164 PHYSICAL BASIS OF HEREDITY 



mosomes was found to be as a iiile somewhat smaller than 

 the sum of the parental haploid numbers, indicating that 

 one or more had conjugated. To the extent to which such 

 union, and the consequent reduction, takes place, the 

 characters of the second hybrid generation may differ 

 from those of the first — at least if the conjugating pairs 

 have different factors in them. 



A similar behavior of the chromosomes has been 

 described by Doncaster and Harrison for two species of 

 moth of the genus Biston (Fig. 24). The hybrids were 

 sterile, and no further generations were raised. 



Federley later made similar crosses ^vith three other 

 moths. A cross between SnierintJius ocellata (with 27 

 chromosomes as the haploid number) and Dilina tilice 

 (with 29) he regards as a cross between genera. A cross 

 between S. ocellata and S. populi (with 28) he regards as a 

 species cross. A cross between 8. ocellata and S, ocellata 

 var. planus he regards as a racial, or varietal, cross. 

 As before the spermatocytes of the hybrid have the sum 

 of the two parental numbers of chromosomes (or a few 

 less at most) . In other words, conjugation of the chromo- 

 somes does not take place. The most unexpected result in 

 these combinations is that the types that are so alike as to 

 be classified as varieties behave as regards conjugation 

 like the other two combinations. The results suggest that 

 ordinary conjugation may not be due to the similarity 

 of the sets of genes carried by the chromosomes so much 

 as to other peculiarities of the combination. 



