DOCTRINE OF CLEISTOGAMY. 85 
resemblance in many respects to a cleistogamic flower that led me 
to notice what I thought might be a peculiarity due solely to the 
separation of the sexes—namely, the expansion of the corolla—and 
I reasoned in this way. We know that the first external sign of 
fertilisation is withering of the floral envelopes or petals; also that 
if pollination of an expanded flower is prevented, the perianth 
remains fresh for some time, a fact so well recognised that the 
Mr. Scott’s cases are not in point; neither do I know anything 
further of Campanula colorata than the mere statement of the fact 
that it bears flowers intermediate between open and closed (Darwin, 
l.c., p. 880). There is, however, the genus Viola, which answers 
of the antedimorphic condition of the genus. At some period, but 
under conditions which it is impossible to dogmatise upon, it woul 
appear that trimorphism set in, afterwards followed by orphism 
by the elimination in some case or cases of the closed flowers, ite 
i i ery 0 
