6 The Plant World. 



similarly taken and the quotient expressed by three figures 

 as a whole number. Studies were made, as before, of the beans 

 as a population. Studies were also made of the beans in pure 

 lines. Fig. 4 shows the regression found in the population. 



This figure shows a regression of about one-fourth and is 

 much the same as that seen in Fig. 2. If we tak the means of the 

 parent lines and the means of the corresponding offspring, in 

 the study of the pure lines, a regression table may be constructed 

 similar to that shown in Fig. 3. 



Fig. 4. KeialiDii of oti>pii<^g lo p»,rtnt bean seeds in regard to relative width, as studied 

 by Johann^en in a population. 



In this instance, Fig. 5, we obtain a very interesting result. 

 Instead of there being any regression as seen in Fig. 4, 

 the correlation is nearly perfect and the line of regression nearly 

 parallels the line of 45 degrees. Taken aj a population, th2 bean 

 weights seem to corroborate Galton's law of regression. When 

 practically the same material is used, but treated by the method 

 of pure lines, then the regression disappears entirely and the off- 

 spring take the lead of ths parents. 



Now it is evident that the selection of individuals from a 

 population, on the basis of a certain character, will bring about 

 a very rapid result. But the rssult is not coincident with any 



