HEREDITY 147 



the most desirable strain and gradually eliminating the 

 less desirable. This is the pure line theory and is now 

 generally accepted as applying to many cases of improve- 

 ment, especially among plants. 



An acceptance of this theory recognizes the fact that 

 no amount of selection can improve a pure line after it 

 has been separated by continuous selection. 



An hypothetical example of pure line selection among 

 animals might be imagined in the case of the wool of sheep. 

 The wool produced under a given set of environmental 

 conditions in a flock of sheep might vary from eight to 

 twelve pounds. In the germ-cells of a given individual, 

 we may assume that determiners are present for the pro- 

 duction of eight, ten and twelve pounds. In breeding, 

 these varying tendencies may be separated. It is con- 

 ceivable that some of the offspring may have inherited 

 the tendency to produce twelve pounds of wool, while 

 others may have inherited the tendency to produce eight 

 pounds. Through many generations of intelligent selec- 

 tion, the flock-master may bring about a more or less 

 complete separation of the tendency to produce twelve 

 pounds of wool and may thus increase the average pro- 

 duction of wool from his flock. The application of this 

 theory to animal-breeding is more difficult than to self- 

 fertilizing plants, but the difficulties are partially removed 

 by close interbreeding. 



140. Hallett's wheat-breeding. The pure line method 

 of breeding probably explains Hallett's unusual success 

 in the improvement of wheat in Great Britain. However, 

 it must be said that Hallett believed that improvement 

 within a pure line of selection was possible. Hallett's 

 method may be best described by using his own words : 



" A grain produces a ' stool ' consisting of many ears. 



