352 The Theory of Natural Selection 



place (Fig. 91). So thoroughly, however, had the Darwin- 

 ian mode of thinking become assimilated among the biologists, 

 especially in England, that for many years Johannsen's work 

 made no impression whatever. It was impossible for Dar- 

 winists to see the implications of these studies. 



Jennings in this country had cultivated Paramecium 

 through many generations and obtained results similar to 

 those of Johannsen with beans (Fig. 92). In later experi- 



A Aj A2 A3 



Fig. 91. Galton's Law of Filial 

 Regression 



The sons of tall fathers will be taller than 

 the population in general, but not as tall, on 

 the average, as the tall fathers. There is a 

 tendency for the offspring of deviates to ap- 

 proach the general run of the population. If 

 from a general population, A, extremes are 

 selected as the parents of the next generation, 

 A] , the new generation will have a mode and 

 a range different from the previous generation, 

 in the direction of the selection; but the mode 

 will not correspond to the characters of the 

 selected parent. Similarly in successive gener- 

 ations, Ao, A3, and so on. The regression is 

 represented by the length of the arrow — that 

 is, the distance between the selected parent and 

 the actual mode of the following generation. 



ments, however, Jennings obtained different results in another 

 one-celled animal, Difflugia, which may be likened to an 

 ameba with a sandy outer shell (see Fig. 17) . Careful selec- 

 tion yielded several distinct strains recognizable by the 

 character of the shell, the number of protuberances, and the 

 appearance of the mouth of the shell. 



Dr. W. E. Castle also produced a noteworthy differ- 

 entiation of type by means of selection (Figs. 93 and 94) . 

 This was apparently a clear case of establishing a character 



