CROSSING. 89 



Size, weight, is a t37pically blending, quantitative character. 

 It is evident, that ahnost any gene which has any influence 

 upon the development of an animal or plant at all, must some- 

 what influence the final size, stature, height of the individual. 

 And for that reason, we must expect that, whenever we cross 

 two individuals of different strains, and when we ignore the 

 number of genes distinguishing their genot3T)e, we must ex- 

 pect those genes which influence size, and for which they are 

 not identical to be nvmierous. This means that we can only ex- 

 pect to find a very great variabflity in size due to variable gen- 

 ot3?pe in the second generation of almost every cross. Espec- 

 ially should this be true where two strains crossed differ them- 

 selves in size. 



In all those experiments in which the size of the individuals 

 of an Fa population has been studied, derived from hybrids 

 between species of different size, great variability in size has 

 been noted. But very often the fact that no clear-cut 3: 1 or 

 9: 3: 3: 1 segregation has been observed, has been taken as 

 proof for the theory that the genes which "determine" such 

 "quantitative characters" are different from other genes, and 

 are in themselves variable. 



We must remember, that several cases have been noted in 

 which two different sizes in one population behaved as alterna- 

 tives, and where size was inherited aaa. pair of Mendelian unit- 

 characters. We should find such cases whereever a single gene 

 has a very marked effect upon the development, and so on the 

 final size, when the variable effect of environmental variations 

 and heterozygosis for other genes does not interfere with our 

 observations of this presence and absence of the gene. 



We have the case of the dwarf guinea-pigs, studied by Miss 

 Sollas, that of the cupid sweet-pea, that of recessive dwarfs in 

 beans and dominant dwarfs in Shimo wheat. 



If our explanation of blending inheritance is correct, if ab- 

 sence of evident segregation is due to complexity and to the rel- 

 ative magnitude of variations in size induced by differences in 

 the evironment, it should be possible to find evidences of Men- 



