Heredity in Population and in Pure Lines. ' 



change in the race indicating a shifting of type. It consists 

 merely in the complete or paitial isolation of a form. If small 

 seeds arc chosen from a m.ixed lot, we will be apt to secure small- 

 seeded offspring, as a whole. But in our choice of small seed, 

 we are also apt to include some minus variants of large-seeded 

 lines. On the other hand, the selection of large seeds from a 

 mixed lot will most likely result in the securing of large-seeded 

 offspring, but in this case we will be apt to include some plus 

 variants of the small-seeded lines. A consideration of these 



:^j^ 



\ 



LIBRAE 





Fig. 5 . Relation of offspring to parent bean seeds in regard to relative width as studied 

 by Johannsen in pure lines. Compare with Fig. 4. 



facts will lead us to understand how it is that the relations oj 

 offspring to parents are such as shown in Figs. 1, 2, and 4, when 

 no regard is paid to the transmissive power of the parent seeds; 

 in short when no attention is paid to pure lines. We can say in 

 any case that "a race with six generations of selection will breed 

 true within 1 . 2 % truth ever af tei wards. ' ' As a matter of fact, a pure 

 race may be secured in a less period of time, even in one genera- 

 tion. In short, the increase of one plant brings about a pure 

 type. If, however, selection is made to one line or biotype, 

 which continually crosses with another line, then the purity of 

 the line through selection may be indetinitely postponed. 





