G. H. SHULL 85 



(Gates, 1913 b). The original rubricalyx-pltmt, self-fertilized, yielded 

 9 0. rubricalyx, 1 rubrinervis, 2 undetermined 1 . Three self-fertilizations 

 of these rubricalyx-specimens produced in the next generation 57 

 rubricalyx and 22 rubrinervis, showing in one family of 44 plants a 

 ratio 3 : 1, and another family grown from seeds of an open-pollinated 

 rubricalyx-pl&ni in the same generation, yielded 71 rubricalyx, 38 

 rubrinervis and 4 doubtful 2 . Owing to various misfortunes the progenies 

 from several other self-fertilizations could not be classified, but both 

 rubricalyx- and rubrinervis-plants were apparently present. From these 

 results Gates concluded that rubricalyx originated as a heterozygote 

 which differed from the parent in a single, dominant, purely quanti- 

 tative, Mendelian character, and strangely enough at the same time 

 concluded that it would always split into rubricalyx and rubrinervis 

 (Gates, 1911). 



The small number of plants which reached maturity in Gates's 

 pedigrees aroused the desire to test more thoroughly the apparently 

 unique behavior of the rubricalyx-ch&r&cter. Since my experiments 

 with this form were undertaken, Gates (1912) has announced the 

 discovery of a " homozygous " individual of 0. rubricalyx, which breeds 

 true to this character, thus proving the incorrectness of his conclusion 

 regarding the continued splitting of the rubricalyx-progemes, but on 

 the other hand strengthening his assumption that the rubricalyx- 

 character is Mendelian in inheritance. His oft repeated emphasis of 

 the view that 0. rubricalyx differs from 0. rubrinervis in a purely 

 quantitative character which acts as a monohybrid Mendelian dominant 

 over the rubrinervis-type of pigmentation will have a special interest in 

 relation to the results described below. In a paper published since 

 this was written Gates (1914) wavers between the treatment of the 

 rubricalyx type of pigmentation as a Mendelian and as a non-Mendelian 

 character. His most positive declarations on the subject are that it is 

 non-Mendelian ; but if he sincerely holds to this conviction it is strange 

 that he should continue to treat the genetic behavior of this character 

 as if it threw valuable sidelights on Mendelian phenomena. 



During the past season (1913) I have had five pedigrees derived 

 from one of the three individuals of 0. rubricalyx grown by me in 1912. 

 These were 0. rubricalyx self-fertilized, 0. rubricalyx x rubrinervis, 



1 These two undetermined plants are listed in one place as 0. rubricalyx. See Gates, 

 1911, Table II, p. 365. 



2 In the same place (Gates 1911, Table II) this family is erroneously indicated as the 

 product of a self-fertilization and the four doubtful plants are included under 0. rubrinervis. 



