THE PRACTICE OF 'BREEDING 301 



breed of fixed characters, that is, characters which are 

 represented by definite determiners in the germ-plasm 

 in such a way that the individual animals of the breed 

 are able to transmit these desirable characters to their 

 offspring with a reasonable degree of certainty. Thus 

 many breeders have surrounded their animals or plants 

 with exceptionally favorable conditions and have selected 

 those which have developed the prized qualities most 

 perfectly under such conditions. This was the method 

 of Hallet in developing improved varieties of wheat. Un- 

 doubtedly many of the early breeders based this practice 

 upon a very deep-seated but mistaken belief in the in- 

 heritance of acquired characters. This method has been 

 very successful in a considerable number of cases, both 

 among plants and animals, but not because the environ- 

 mental factors involved had fundamentally changed the 

 real character of the germ substance. In all such cases 

 the improved environment acted merely as an efficient 

 selective device and indicated those individuals which 

 actually possessed the capacities valued by the breeder. 

 This method has often failed in accomplishing lasting 

 improvement, because the conditions surrounding the 

 breeding stock are not average conditions and the appar- 

 ent improvement may be wholly due to a better food 

 supply or more room and not due to fundamental differ- 

 ences in the germ. Another method of improvement 

 which involves selection of the best is to place the plant 

 or animal under ordinary or even unfavorable conditions 

 and select those individuals which appear best able to 

 develop the desired qualities under such conditions. 



285. Selecting chance variations. We know that 

 sudden and important variations often occur in the germ- 

 plasm. Some of these variations may be of such a char- 



