PLANT AND ANIMAL IMPROVEMENT 217 



tage of real improvement through the elimination of 

 certain undesirable characters. 



At present, this application of inbreeding to the im- 

 provement of cross-bred animals and plants is somewhat 

 of an unknown quantity. It has not been as thoroughly 

 tested as might be desired, but the basic principle is 

 sound. Although it is a drastic procedure, it is merely 

 utilizing to the fullest extent what practical breeders 

 have recognized as one of the most valuable benefits of 

 close mating. Accepting the doctrine that consanguinity 

 in itself is not in any way injurious and that good or evil 

 results from it solely through the inheritance received, 

 we can attack the century-old problem of inbreeding with 

 a clarity of vision heretofore impossible. Breeds of ani- 

 mals, and naturally crossed varieties of plants, which are 

 necessarily more or less heterogeneous in their hereditary 

 constitution, can be split up into their component parts 

 by this means. The pure types obtained can then be 

 selected wdth far more surety than is ever possible with 

 organisms in a continuously hybrid condition, thereby 

 presenting basic stock of tested value for further hybrid- 

 ization and recombination. 



With plants the application of this method would be 

 simpler than with animals. Most naturally crossed plants 

 can be artificially self -fertilized and constancy and uni- 

 formity reached in about eight generations if there are 

 no complicating factors such as self -sterility. The ex- 

 pense would not be prohibitive, although many pure lines 

 must be tested in order to have a high probability of 

 obtaining all that is best in a variety. After the most 

 desirable combinations are isolated, their recombination 

 into a new and better varietv, which could be maintained 



