Sex-limited inheritance in Lychnis dioica L. 285 



(Tables I and III). The father was a well developed hermaphroditic 

 plant in a family of pure Melandrium album, being the only her- 

 maphrodite I have yet seen in this strain. The original material of this 

 line had been sent to me in September 1909 by Dr. BAUR, and had come 

 from nature in the vicinity of Friedrichshagen near Berlin, Germany. 

 It may have been rather closely related, therefore, to the line which 

 produced, in that vicinity, the angustifolia mutant. The rarity of her- 

 maphrodites in this M. album strain may be inferred from the fact that 

 in the same year (1912) in which this hermaphroditic plant appeared, 

 I had 57 families of pure first cousins of this plant, which contained a 

 total of 18$1 females and 1911 males and no hermaphrodites. The sibs 

 of the hermaphrodite plant consisted of 34 females and 36 males. 



The seeds from this cross, sown February 15, 1913, produced 

 96 broad-leafed and narrow-leafed rosettes. Of these plants 95 bloomed 

 and all were female! In many hundreds of families of Lychnis dioica 

 which I have grown during the past eight years, I have had no sizable 

 family before 1913 in which no males (or hermaphrodites) were included, 

 though there have been wide differences in the ratios. The result is 

 the more remarkable when it is recalled that both in BAUR'S cultures 

 (BAUK 1912) and in mine there has generally been an excess of males 

 in the pure M. album families, though in other strains of Lychnis dioica 

 there has been, on the whole, a marked excess of females. 



With some hesitancy I offer a hypothesis for this unexpected result, 

 since it may not be possible to determine experimentally whether the 

 hypothesis is correct or not. In my earlier papers (SHULL 1910, 1911) 

 I have shown that certain hermaphrodites gave the same results, on 

 breeding, as if they had been normal males. These I have called "somatic 

 hermaphrodites" because their genotypic condition was apparently the 

 same as that of the normal males. On the other hand, STRASBURGER 

 (1900, 1910) and DONCASTER (1912) have shown that females affected 

 by the anther-smut (Ustilago violacea) are changed to apparent (i.e. somatic) 

 hermaphrodites, and DONCASTER (loc. cit.) has found in one such case 

 pollen-mother-cells formed. I am now inclined to the view that the 

 hermaphrodite plant (11266 (51)) used in the cross described in this 

 section, was merely a somatic hermaphrodite, and that instead of being 

 derived from the male as in my former somatic hermaphrodites, this 

 one was produced by a somatic modification of the female, whose geno- 

 typic condition remained unchanged, so that all of its sperms were of 

 a single sort, all carrying both of the determiners B and F in their 

 normal condition. 



