INHERITANCE—IMPROVEMENT BY BREEDING i 



93 



111 the maize plant the factors which interact to cause the CHAP. 



transmissible differences in the size of the organs are very 

 complex, as has been pointed out by East and Hayes, who 

 refer to the consequent difficulty of working out in detail their 

 inheritance. 



" It is perfectly obvious to one familiar with the maize plant 

 that it is almost impossible to work out in detail the inherit- 

 ance of the complex factors that interact to cause the trans- 

 missible differences in the size of the organs. 



" That size characters are complex in themselves is shown 

 by the numerous varieties grown commercially. They each 

 vary from their own means, but different variety means in 

 height are found all the way from two and one half to fourteen 

 feet, with but little actual difference between the most similar 

 strains. Further to complicate matters, all size characters 

 respond to environmental stimuli, and these non-inherited 

 fluctuations obscure the analysis of pedigree cultures in a still 

 greater degree. 



" For these reasons we do not attempt to analyse our results 

 further than to say that they do show segregation in every case. 

 And segregation is held to be the important and essential feature 

 of Mendelism. Therefore we believe that size characters Men- 

 delize. . . . But in size characters dominance is probably very 

 incomplete or absent. . . ." 



" Several genes for the same character may exist in the germ 

 cells of one organism, the number being limited possibly by the 

 number of chromosomes. The limited number of cases, thus far 

 found, presumably is due to the fact that few size characters 

 have been investigated, for nowhere would these phenomena 

 be so likely to occur as in quantitative characters. . . . Several 

 independent allelomorphic pairs may produce the same somatic 

 character." 



"A heterozygous combination presumably produces half 

 the effect of a homozygous combination. Then as dominance 

 becomes less and less evident the Mendelian classes vary more 

 and more from the formula (3 + 1)" and approach the normal 

 curve of error (•£ + \) n . When there is no dominance and open 

 fertilization, a state is reached in which the curve of variation 

 simulates the fluctuation curve, with the difference that the 

 gradations are heritable. The heritable variations are al- 

 ways more or less obscured, however, by the ever present 

 fluctuation." 



1.3 



V. 



