68 
A. B. Rendle . 
the later development. If however we imagine the typical Salix 
flowers to be derived from an hermaphrodite type we should not 
expect to find the glands, on the view that they represent a perianth 
derived from the andrcecium, more numerous in the male than in 
the female. 
Betulaceae comprises the two tribes Coryleae and Betuleae. 
In Coryleae (Corylus , Carpinus) the male flower has no perianth; 
it stands in the axil of a bract and consists of four stamens 
preceded by two bracteoles. In the female flower there are however 
in addition varying numbers of epigynous scales (i.e.: a rudimentary 
perianth). There is no suggestion that these perianth scales 
in the female flower represent the lost stamens of an hermaphrodite 
flower. 
In the other tribe Betuleae the state of affairs is reversed. In 
one section (Gymnothyrsus) of A Inns the male flower has a four¬ 
leaved gamophyllous perianth, and the four stamens are opposite 
the perianth leaves. In other species of Alnus (§ Alnaster and 
§ Clethropsis) and in Betala the perianth leaves are free and fewer 
in number, often only two. In the female flower there is no 
perianth. If the unisexual flowers had a common origin from a 
hermaphrodite type we might expect a similarity of arrangement in 
the two sexes. 
A comparison of the male flowers in the two tribes is sug¬ 
gestive. In Coryleae where there is no perianth the stamens are 
well protected by the bract and bracteoles; in Alnus a similar 
protection is afforded by the four perianth leaves, which are placed 
immediately behind each stamen ; there is no reduction in/ the 
number of stamens. In Betula where there are only two (median) 
stamens, often only the two median perianth leaves are developed. 
Associated with this development of perianth in Betuleae is a partial 
[A Inns) or complete ( Betula ) suppression of the bracteoles of the 
two lateral flowers of the dichasium. 
The suggestion is that the perianth is protective ; it is however 
not easy to associate it directly with the suppressed bracteoles, as 
its development is the same in all three flowers of the dichasium, 
whereas the bracteoles of the central flowers are both present. In 
the female flowers of the Coryleae where the bracteoles are present 
and help to form the involucre, the perianth consists of, as far as 
we can see, functionless outgrowths. If it is not possible to regard 
the perianth as a new foliar outgrowth sui generis, we must seek 
its origin in additional bracteoles which in the male flowers have 
