284 BOTANICAL GAZETTE [APRIL 
reduction division of both megaspores and microspores of one form, 
in this case the short-styled form. Whether this is accompanied 
by any morphological difference in the chromosomes or not is of 
course another question. 
The resemblance of dimorphic heterostylous plants to dioecious 
plants suggested to DARWIN (p. 285) that heterostyly may have 
been one of the ways by which the dioecious condition among flower- 
ing plants was attained. He cites several cases of plants which are 
dioecious, but show indications of a heterostylous ancestry. Aspe- 
rula scoparia, an inhabitant of Tasmania, is dioecious, but the male 
flowers have large anthers and a very small pistil with rudimentary 
stigma and style, while the female flowers have a large, well- 
developed ovary and rudimentary anthers apparently quite desti- 
tute of pollen. Discospermum, of Ceylon, is apparently heterostyled, 
but one of the two forms is always barren, the ovary containing 
about two aborted ovules in each loculus; while in the other form 
each loculus contains several perfect ovules. The species is there- 
fore really dioecious. Most of the species of the South American 
genus Aegiphila are heterostyled. In Aegiphila obdurata, how- 
ever, the anthers of the long-styled form are entirely destitute 
of pollen, while the pistil is perfectly developed; in the short- 
styled form, on the other hand, the pistil is aborted, while the 
stamens are perfect. 
There are a number of facts which indicate (BLAKESLEE 3, 
p. 371) that in all dioecious plants one sex is dominant and makes 
its appearance while the other remains latent. Male and female 
willow plants are frequently found with flowers of the opposite sex. 
Lychnis dioica is normally dioecious, but STRASBURGER (35, p. 692) 
‘found in his cultures at Bonn occasional hermaphrodite plants. 
These were in every case affected by a smut, Ustilago violacea, and 
he attributes the hermaphrodite condition to the action of the 
fungus. Ustilago violacea fruits only in the anthers of the host 
plant. If it attacks a male plant it fruits in the anthers, and if it 
attacks a female plant, in some way it stimulates its host to the pro- 
duction of stamens, in which it fruits. 
Recently SHULL (32, p. 112) has described occasional hermaph- 
rodite plants occurring in a pure bred normal race of Lychnis, 
