THE MECHANISM OF HEREDITY 65 



The peculiar fact tliat there is no crossing over in 

 the males need not concern us here. In other species 

 there is evidence of crossing over in both sexes. What is 

 important is that crossing over does occur with a definite 

 frequency and this frequency is constant for any par- 

 ticular pair of characters except when modified in various 

 ways which can be given concrete explanations. It does 

 not matter, moreover, whether the two factors enter the 

 cross as ^'^ or as ^j^ , crossing over is the same in each 

 case; that is, the tendency is just as great for Ah to stay 

 together after they are in that combination as for AB to 

 stay together when that particular combination charac- 

 terizes the individual. 



The gross result of several thousand experiments of 

 this character, therefore, is to associate Mendelian in- 

 heritance more definitely than ever with the chromosomes. 

 What seemed to be an exception, furnished striking evi- 

 dence in support of the theory. It is true, a few isolated 

 instances of characters which appear to be carried by 

 cell substances other than the chromosomes have been 

 discovered ; but it is pretty clear that in the main all of the 

 varied characters which differentiate the individuals 

 within a single species — and these are the only ones 

 which can be studied profitably through crosses — are con- 

 trolled by the distribution of factor units which lie 

 within the chromosomes. 



We can visualize the whole process of heredity by 

 means of chromosome diagrams just as we can visualize 

 the whole process of chemical recombination through 

 models of the atoms constructed to fit the facts furnished 

 by chemical reactions. The result is that we can do much 

 toward predicting what will happen under given con- 



5 



