50 THEORY OF EVOLUTION 



The question as to whether a given character 

 is dominant or recessive is a matter of no theo- 

 retical importance for the principle of segre- 

 gation, although from the notoriety given to it 

 one might easily be misled into the erroneous 

 sup23osition that it was the discovery of this re- 

 lation that is ^lendel's crowning achievement. 



Let me illustrate by an example in which the 

 hybrid standing between two tyj)es overlaps 

 them both. There are two mutant races in our 

 cultures of the fruit fly Drosoj^hila that have 

 dark body color, one called sooty, another which 

 is even blacker, called ebony (fig. 20). Sooty 

 crossed to ebony gives offspring that are inter- 

 mediate in color. Some of them are so much 

 like sootv that thev cannot be distinguished 

 from sooty. At the other extreme some of the 

 hybrids are as dark as the liglitest of the ebony 

 flies. If these hybrids are inbred there is a con- 

 tinuous series of individuals, sooties, interme- 

 diates and ebonies. Which color here shall we 

 call the dominant? If the ebonv, then in the 

 second generation we count three ebonies to 

 one sooty, putting the hybrids with the ebonies. 

 If the dominant is the sootv then Ave count three 



