284 Shull. 



86 broad -leafed rosettes and 9 narrow -leafed. The 91 plants which 

 bloomed consisted of 67 broad -leafed females, 15 broad -leafed her- 

 maphrodites and 9 narrow-leafed hermaphrodites. The narrow-leafed plants 

 were indistinguishable from the angusti folia males of other pedigrees 

 in any respect other than the development of well-marked ovaries, some 

 of which were functional. Not only were the leaves identical at all 

 stages with those of angustifolia males, but the petals also were narrow 

 and frequently exhibited the virescent lines mentioned above in the de- 

 scription of the angustifolia form. 



This demonstration that my "genetic" hermaphrodites have the same 

 constitution with respect to the broad -leaf character as the males, 

 being like them, heterozygous for the gene 5, is of much interest 

 because of the strong support it gives to my conclusion, arrived at on 

 other grounds, that the hermaphrodites are modified males ; at the same 

 time it strengthens the indication given by the preceding family, that 

 the males of Lychnis dioica may be generally heterozygous with respect 

 to the gene B as well as the gene F. MORGAN (1913, p. 173) assumes 

 that my hermaphrodite mutants were derived from the female, by a 

 modification of one of the female sex -genes, F. He gives the female 

 the formula FF and the hermaphrodite FH, just as I gave them in one 

 of my formulae (SHULL 1911, pp. 357 and 365), though he states that 

 the symbols he uses are not those used by me. The only difference 

 consists, not in the symbols, but in MORGAN'S assumption that the H 

 is a modified jP, while I supposed that it might have an origin inde- 

 pendent of F and was inclined to the view that it resulted from the 

 modification of a male gene M, or perhaps a sexually indifferent "synaptic 

 mate" of F. If the genetic hermaphrodites were modified females, would 

 it not be strange that in the process of modification the sex should be 

 only half way converted to the male condition, while the accompanying 

 broad-leaf character is completely reduced to the condition seen in the 

 angustifolia males? While such an interpretation is possible, it appears 

 to me simpler to conclude that the mutation which gave rise to the 

 hermaphrodite took place in that element of the male which already 

 lacked the gene B and which may possibly be a male gene, M. A further 

 convincing evidence that hermaphrodites are derivable from the males 

 is the occurrence of a narrow -leafed hermaphrodite mutant in family 

 No. 12240 (see Table I), for in that family no narrow-leafed females 

 were due to appear. 



(3) Pedigree Numbers: 11335 (14) X 11266 (51) = 12261. The 

 mother plant has been shown to be heterozygous with respect to B 



