. JLUME LII NUMBER 5 



THE 



BOTANICAL GAZETTE 



NOVEMBER 19 if 



REVERSIBLE SEX-MUTANTS IN LYCHNIS DIOICA 1 



GEORGE HARRISON SHULL 



(WITH FIFTEEN FIGURES) 



Six hermaphrodite specimens of Lychnis dioica L. were found in 

 cultures of 1908, and eight in 1909. With respect to their heredi- 

 tary behavior in the first generation, when used as pollen parents, 

 these hermaphrodites proved to be of two kinds, the individuals 

 A and B being capable of determining the hermaphrodite character 

 in their male offspring, while individuals C and D behaved exactly 

 like normal males, giving progenies consisting of females and nor- 

 mal males. 



The conclusion was reached (SHULL 26) that the hermaphrodites 

 are modified males, because (i) in all families in which the first 

 mentioned type of hermaphrodite was used as the pollen parent 

 the offspring consisted of females and hermaphrodites in the same 

 ratio as would have been expected of females and males if a normal 

 male had been used as the pollen parent, and because (2) the second 

 type of hermaphrodite when used as a pollen parent gave the same 

 result that a normal male would have given. 



Accepting tentatively the Mendelian explanation of sex first 

 clearly enunciated by CORRENS (6), which recognizes the one sex 

 as homozygous and the other sex as heterozygous with respect to a 

 sex-producing gene, it was decided that these hermaphrodites (and 

 therefore also males) must be heterozygous, because (i) the males 

 are capable of being modified in such manner as to display function- 



1 Read at the meeting of the Botanical Society of America, Minneapolis, Decem- 

 ber, 1910. 



329 



