302 THE BREEDING OF ANIMALS 



acter as to have great economic value to man. Varia- 

 tions of this character are invariably transmitted, and the 

 wise and observant breeder may often make rapid prog- 

 ress by making such chance variation the basis of his 

 selections. This is the method of De Vries. In following 

 this method the breeder does not consciously undertake 

 to cause variation but rather to take advantage of those 

 which have resulted from natural causes. The breeder 

 of the domestic animals will often have difficulty in deter- 

 mining whether a given variation is due to environmental 

 causes or is due to fundamental changes in the germ-plasm. 

 The skillful breeder, however, will conclude with reason- 

 able assurance that when an individual animal exhibits 

 a rare and unusual aptitude in the development of a 

 certain character or characters, in a herd in which the 

 individuals are all maintained under identical conditions, 

 the rare development may be regarded as a germinal 

 variation. Such germinal variations may under certain 

 conditions become the foundation of a new strain. 



286. The Burbank method. If the selection of varia- 

 tions is the road to success in improvement, then why not 

 systematically attempt to cause variations and thus 

 increase the chances for discovering a desirable mutation ? 

 This is the plan followed by Burbank. By crossing a 

 great number of individuals, variations are secured, and 

 by a process of gradual elimination the outstanding vari- 

 ants are retained and reproduced. This plan does not 

 create any new forms, but depends on the well-known 

 tendency of unit characters to rearrange themselves in 

 new combinations which for all practical purposes may 

 really become a new creation. This method involves 

 the propagation of the improved individual by budding, 

 grafting or similar asexual method and cannot therefore 



