96 Negative Correlation in Oenothera Hybrids 



DISCUSSION. 



The behavior of the red pigments in these 0. rubricalyx-hylorids, 

 is in striking contrast with that reported by Gates (1909 to 1913 b), 

 whose reciprocal crosses between 0. rubricalyx and 0. Lamarckiana 

 appeared to consist predominantly of rubricalyx and Lamarckiana. To 

 what extent the differences between his results and mine are due to the 

 very slender basis for the conclusions regarding his crosses of these 

 species, cannot be determined, and consequently it is impossible to say 

 how much of these differences is to be referred to putative genotypic 

 differences between the particular individuals which entered into his 

 crosses and those which were used in mine. Whatever may be the fact 

 in regard to the lack of genotypic identity between his parent plants 

 and mine, the results described in this paper have an important bearing 

 on the several propositions set forth by Gates in regard to the origin 

 and genetic behavior of the rw&ri'ca /^-character, namely (a) that the 

 difference between 0. rubricalyx and 0. rubrinervis is a purely quantita- 

 tive one; (6) that the rubricalyx-character is a typical monohybrid 

 Mendelian character; and (c) that the method of inheritance of a 

 character is determined by the nature of that character itself. 



Many Mendelian color-patterns have been discovered in various 

 plants and animals, which are inherited quite independently of the 

 actual quantities of pigment present in the organism as a whole, the 

 same pattern being associated sometimes with weak pigmentation at 

 other times with intense pigmentation; dissimilar and independently 

 inheritable patterns may also affect different parts of the same indi- 

 vidual, and in every such case it is demonstrable that a " qualitative " 

 and not alone a " quantitative " difference is present. Such experiences 

 incline the geneticist to the interpretation of any striking change in 

 a color-pattern as something more than a quantitative change in the 

 amount of pigmentation present. Still, in a case as simple as that 

 which Gates's material seemed to him at first to present, in which the 

 new color-pattern completely includes the original one, the interpre- 

 tation of the change as a purely quantitative one is possible, though 

 perhaps not in any case particularly probable. Gates (1914) has now 

 shown that in the case of the rubricalyx-cha,ra,cter, also, the pattern is 

 measurably independent of the actual quantity of pigment, since he 

 secured pale red buds with the characteristic rubricalyx-distribution of 

 the pigment. The results from my crosses show even more strikingly 



