CH. VI, 8] HEREDITY AND VARIATION 313 



not likely to breed together repeatedly unless compelled by 

 experiment, though the same result is effected in some 

 plants which pollinate themselves ; characters are not passed 

 along singly, but commonly a number together in loose 

 aggregations; determiners seem to exert certain influences 

 upon one another directly ; and there are yet other compli- 

 cations. Hence in Nature the law is not manifest to obser- 

 vation, though discoverable by experiment ; but it operates 



9 



FIG. 220. Diagram to illustrate Mendel's Law of the segregation of 

 characters in heredity, using a single character of Fig. 219. 



If germ cells having the dominant character (black circle) breed with 

 others having the recessive character (white circle), then all of their off- 

 spring show only the dominant character but carry the recessive character 

 latent (black circle with white center) . If these forms breed together, their 

 offspring will show the distribution of characters represented in the diagram, 

 one-fourth will be pure dominants and one-fourth pure recessives, while 

 the remainder are dominants carrying the recessive character latent. If 

 those in this generation breed only with their own kind, the result in the 

 next generation is as shown in the diagram ; and thus indefinitely. 



as a steadily working principle which runs as a kind of 

 guiding thread through all heredity, while coming to view 

 now and then in such phenomena as "skipping a genera- 

 tion," elimination of characters from a race, and other 

 less obvious matters. 



Thus, on the basis of our knowledge of the performance of 

 the chromosomes in reproduction, in conjunction with 

 Mendel's law, heredity must rest upon the transmission of 

 determiners which, existing in each species in a certain 

 number, are distributed in different combinations in the 



