The Origin of the Perianth in Seed-Plants. 67 
evolution of the angiosperm, and there is therefore no apriori 
reason for deriving the floral envelopes in the latter group from 
sporophylls. 
The writer concludes with the statement “ that everywhere 
both calyx and corolla, may lay claim to a similar place of birth, 
viz: in the androecium, however distinct or dissimilar from each 
other in almost every character they may at times appear.” 
It is difficult to accept this statement in its entirety. The 
flowers which we are inclined to regard as the most primitive are 
unisexual and have presumably not been derived from an 
hermaphrodite type ; yet we find in the female, as well as in the 
male, instances of a well defined perianth, which in the case of the 
female could not have originated from an androecium. Hence the 
origin of perianth from androecium cannot have been universal. 
It may be of interest to note the relation which the perianth 
bears to the sporophylls in some of these “ early ” simple-flowered 
orders. 
Salicaceae comprises the two genera Populus and Salix. In 
the poplars, which are wind pollinated, we find in both sexes a cup 
development of the axis below the sporophylls. This cup is the first 
developed part of the flower ; its position and development suggest 
a simple perianth, and there is no reason for associating its origin 
with either stamens or carpels ; it is an outgrowth of the floral axis 
preceding the sporophylls in time and space. In Salix, where 
entomophily has arisen, we find in a similar position one or two 
small honey-secreting glands. If these are homologous with the cup 
in Populus , they are in comparison much reduced and late in 
development. Their degree of development shows considerable 
variations, and in many species the reduction is more complete in 
the female flower than in the male. Thus, while in the male of 
many species a posterior and anterior gland is present, the anterior 
is absent in the female. In some cases the glands are united 
laterally below to form an inconspicuous lobed hypogynous ring (as 
in Salix tetrasperma). 
We may compare this reduction of the perianth in Populus to 
a nectary or nectaries in Salix with the similar reduction of petal to 
nectary in Ranunculaceae, and associate it with the entomophily of 
the latter genus. The occasional appearance of hermaphrodite 
flowers in the willow has suggested the derivation of the unisexual 
flowers from an hermaphrodite -type, but the argument may be used 
in the other direction and the hermaphrodite flower be regarded as 
