CROSSING OVER AND CHROMOSOMES 115 



ing stage, there can be given no explanation as to why 

 homologous genes should not recombine in all possible 

 combinations with other genes. But this is exactly what 

 does not happen. If it be supposed that the chromosomes 

 dissolve only partly into chains of genes, it is still not 

 obvious why the chains of one chromosome should be iden- 

 tical with those of the other (its homologue) as they must 

 be to recombine properly ; for, in neighboring nuclei other 

 chains are forming — as the crossing-over results indi- 

 cate — involving breaking at all possible levels. 



Bateson and Punnett have proposed a theory of cross- 

 ing over that is called reduplication. It is fundamentally 

 different from the one here adopted. Although I think 

 this theory outlawed by the evidence that Plough has 

 obtained, and made impossible by certain other considera- 

 tions that will be given later, the theory is so interesting 

 that it may be briefly stated. Bateson suggests that at 

 some time early in the embryo segregation may take place 

 involving heterozygous pairs of factors. In the actual 

 case presented only two such pairs are involved. As a 

 ^^ symbolic presentation'' of the situation Bateson gives 

 the diagram drawn in Fig. 51. 



Although the dichotomous method of separation is 

 utilized in the second line of figures to show reduction of 

 the two pairs at once, such figures could obviously bear 

 no relation to the ordinary process of cell division — nor 

 do they, I understand, pretend to be. After separa- 

 tion (segregation) the cells that get AB and ab are repre- 

 sented as dividing faster than the cells Ab and Ba, hence 

 there will be more of them in proportion as the two rates 

 of division differ. 



Bateson 's view is open to the following criticisms: 



1. The evidence from Drosophila, where many linkage 

 ratios: are known, gives no support to the view that these 

 ratios fall into relatively few dichotomous schemes, such 

 as Bateson 's hypothesis calls for. Other forms also fail to 

 fit such a view. On the contrary, the ratios fall into no 



