78 Plant Genetics 



the same situation, so that the chances of a reappear- 

 ance of the original parent types would be very slight, 

 so slight that if they did appear they would be described 

 as "reversions." In this way many of the cases of true- 

 breeding hybrids can be explained. Of course we have 

 been dealing only with intermediate hybrids which are 

 intermediate quantitatively, due to the action of cumu- 

 lative factors. The gardeners and horticulturists might 

 claim that their hybrid is not merely different in degree 

 from either parent but that it is different also in kind; 

 that it exhibits an entirely new character, that is, it 

 differs quahtatively. If this is true of course the 

 explanation suggested does not apply. The explana- 

 tion can be modified, however, leaving the mathematical 

 possibilities the same. Suppose the original parent 

 strains differed, not by six cumulative factors, but by 

 six factors of different kinds. Under these conditions 

 it would be possible to bring together in a cross an 

 entirely new combination of factors which might result 

 in characters that would seem entirely new. If we are 

 dealing with complementary or supplementary factors 

 we might get a character in a hybrid that is entirely new. 

 If these factors were sufl&ciently numerous, the chances 

 of separating them again and recombining them exactly 

 as in the original parents would be exceedingly slight. 



These possibilities have been presented for two rea- 

 sons: (i) as explanations of true-breeding hybrids, 

 whether quantitatively or qualitatively different from 

 their parents; (2) to illustrate the method of using the 

 conceptions of the factor hypothesis. 



7. A PRACTICAL ASPECT OF THE CUMULATIVE FACTOR 



HYPOTHESIS.— Assume that a practical breeder crosses 



