124 



HEREDITY AND SEX 



of the next generation, their offspring will make a new 

 curve that has moved, so to speak, in the direction of 

 selection (Fig. 63). 



If again two more extreme individuals are selected, 

 another step is taken. The process is assumed to go 

 on as long as the selection process is maintained. 



So the matter stood until a Danish botanist, Johann- 

 sen, set seriously to work to test the validity of the 

 assumption, using a race of garden beans for his meas- 

 urements. He discovered in the first place that popu- 



FIG. 63. Schematic representation of the type-shifting effect of selec- 

 tion from the point of view of Galton's regression theory. The * marks the 

 point on the curves of A, Ai, Az from which the selection is supposed to be 

 made. (Goldschmidt.) 



lations are made up of a number of races or "pure 

 lines." When we select in such a population we sort 

 out and separate its constituent races, and sooner 

 or later under favorable conditions can get a pure 

 race. Selection has created nothing new; it has 

 picked out a particular preexisting race from a mixed 

 population. 



Johannsen has shown that within a pure line selec- 

 tion produces no effect, since the offspring form the 

 same group with the same mode as the group from which 

 the parents came. The variability within the pure 

 lines is generally ascribed to environmental influences 



