146 GENETICS 



parently furnishes evidence of modifications of an he- 

 reditary characteristic through selection following a 

 homozygous cross. 



Castle succeeded in selecting two extreme races of 

 rats from his hooded stock, one possessing almost no 

 pigment and the other with the "hood" so extended 

 that it covered practically the entire body. 



Is then the germ a variable thing that makes it possi- 

 ble to select effective differences out of a pure line, to 

 the discomfiture of Mendelians who build their house 

 on the rock of constancy of the germplasm, or can 

 these perplexing results be somehow explained? 



7. CONCLUSION 



At any rate it would be gratifying scientifically to 

 discover one fundamental law to which all these various 

 cases of pure line selection are accountable because in- 

 tellectual satisfaction always follows upon finding the 

 common denominator of things. 



A unifying explanation that makes a single harmoni- 

 ous interpretation of these apparently diverse results, 

 based on the idea that all are reducible to Johannsen's 

 conception of the ineffectiveness of selection within a 

 pure line, has perhaps been reached in the theory of 

 modifying genes which will be considered in the next 

 chapter. 



Certainly the pure line concept is a very useful tool 

 for the geneticist since with it the hereditary upset of 

 outside germplasm is eliminated. Consequently it is 



