THE MENDELIAN PRINCIPLES OF HEREDITY 99 



E, F, G, H). The Fi flies gave the expected re- 

 sults. These inbred gave the following F2 significant 

 classes : ^ 



GRL YWM GWM YRL GRM YWL GWL YRM 



2089 1361 17 23 887 817 5 



In these results the classes where single crossing over 

 is shown are GWM (17) and YRL (23) (Fig. 51, G, 

 G') and GRM (887) and YWL (817) (Fig. 51, F, F')., 



There are two classes, namely, GWL (5) and YRM 

 (0) (Fig. 51, H, H'), which involve double crossing over. 

 In order that they may take place, the two sex chromo- 

 somes in the female must break twice and reunite 

 between the factors involved, as shown in the diagram. 

 Such a redistribution of the parts of the homologous 

 chromosomes would be expected to occur rarely, and 

 the small number of double crossovers recorded in 

 the results is in accord with this expectation. 



In these questions of linkage we have considered 

 some of the most recent and difficult questions in the 

 modern study of heredity. We owe to Bateson and 

 his collaborators the discovery of this new departure. 

 In plants they have recorded several cases of linkage, 

 and other authors, notably Correns, Baur, Emerson, 

 East, and Trow have described further cases of the 

 same kind. Bateson has offered an interpretation 

 that is quite different from tlie one that I have here 

 followed. His view rests on the assumption that 

 separation of factors may take place at different times, 

 or periods, in the development of the germinal tissues. 



^ The classes omitted are those that do not bear on the question 

 in hand. 



