HERED] IT IN PURE LINES 21 







The first thing he did was to collect all the seeds (we would call 

 them beans) from a single bean plant, which is a male-female individu- 

 al. These beans were of many different sizes, some quite large, others 

 quite small, and many intermediate. In order to follow the heredity 

 of bean size with accuracy, he weighed each bean on a delicate balance. 

 What he determined was the weight of each bean, but we shall speak 

 of large and small size instead of heavy and light weight. Since some 

 few of the beans were very large and some few very small, what more 

 natural than to select the largest beans to be the seeds for the next 

 generation of plants, and to expect that the beans of the next and the 

 next and the next generation would be larger and larger and larger? 



If Johannsen was sanguine enough to expect to get such a rapid 

 improvement in beans, he was doomed to disappointment. He had 

 constructed a curve of variation, in the manner described in the last 

 chapter, for beans of the parent plant. This curve had a definite 

 shape and a definite mode. When he constructed a similar variation 

 curve for the progeny derived from the largest beans of the first gener- 

 ation this curve was practically the same as that of the parent genera- 

 tion. The beans were not all large, as he may have hoped, but some 

 Were large, some small, and many intermediate, just as in the parent 



oration. Continuing to select, for planting, the largest beans for 

 six generations, he found that there had been no change in the average 

 size of beans, and the number of largest beans had not increased. 

 Selection then had been powerless when operating upon individuals 

 with identical heredity. 



Another series of breeding experiments was carried out, select 



seed the smallest beans of the original parent plant of Pure Line I. 

 The bean progeny of the second generation differed not at all from 

 those derived from the largest bean of this pure line. They gave the 

 same variation curve and the same average as the latter, and this was 

 maintained for six generations. It made no difference whether the 

 largest, the smallest, or the average beans were selected for planting; 

 so long as they belonged to the same pure line, the same variation curve 

 and the same average w-ere maintained. 



A second pure line, vrhich we may call Pure Line II, was started 

 from another parent plant, chosen because the average size of beans 

 was distinctly less than that of the first pure line. The statistical 

 study of the beans of this second parent plant revealed another new 

 and significant fact, namely, that the beans, when arranged accord- 

 to sizes, gave a variation curve differing markedly from that of 



