172 INBREEDING AND OUTBREEDING 



many other crosses which showed hybrid vigor. Further- 

 more, not only was the union of such simple factorial com- 

 binations inadequate to account for the frequency of the 

 widespread occurrence of heterosis, but there was another 

 seemingly insurmountable objection to the interpreta- 

 tion. It was pointed out that if heterosis were due solely 

 to dominance of independent factors, the distribution of 

 the second generation would be unsymmetrical in respect 

 to those characters in which an increase was shown in the 

 first generation. This criticism has its basis in the famil- 

 iar fact that Mendelian expectation in the second hybrid 

 ' generation where there is complete dominance is always 

 an expansion of the form (3 + 1) to a power represented 

 by the number of factors. Even with partial dominance 

 the criticism holds, although the lack of symmetry is not 

 so marked. 



But in the vast amount of data accuniulated upon the 

 inheritance of quantitative characters no such tendencies 

 toward asymmetrical distribution in the second genera- 

 tion are evident. In the majority of cases recorded where 

 hybrid vigor is shown in the first generation, the distri- 

 bution of the individuals fits the symmetrical curve, com- 

 monly known as the Curve of Error, remarkably well. 



It is evident, therefore, that the objections raised 

 against the hypothesis of dominance as a means of ac- 

 counting for heterosis, as outlined by Keeble and Pellew, 

 are valid. But both these objections to dominance as an 

 interpretation of heterosis were made before the facts of 

 linkage were known. With linkage these criticisms based 

 upon Mendelian expectancies with independent factors do 

 not hold. 



Abundant evidence is fast being accumulated to show 



