FLOWER OF GNETUM 
161 
flower) of female flowers 1 ’ - . In this connection Miss Berridge's discovery of 
small vascular complexes arising from the main vascular system between the 
outer and median envelopes in the complete female flower of G. Gnemon 2 is of 
interest. If these structures are vestiges of the traces which formerly 
supplied a ring of male flowers, as Miss Berridge supposes, they indicate that 
a grouping of flowers similar to that suggested above has been in existence. 
The identification of such a bisexual “flower” with the supposed primitive 
form of inflorescence would establish homologies between (a) the complete 
female flower and a uninodal inflorescence, ( b ) the cupule of the inflorescence 
and the outermost envelope of the complete female flower. However 
improbable such homologies may at first sight appear, they do not entirely 
lack support. In addition to the inference drawn from the frequent presence 
of a terminal female flower, we have the fact recorded by Mrs Thoday 3 that 
the vascular supply of the female flower of Gnetum is very similar to that of 
the whole cone of Cycadeoidea Wielandii and, it may be added, to that of a 
lateral inflorescence of Gnetum itself. A fact of even more significance is 
that, in a case in which the main axis of the inflorescence was injured, a 
lateral inflorescence replaced a female flower (Fig. 7). The injury to the 
main axis had taken place early in its development for the slightly concave 
surface of the wound was protected by two or three layers of suberised tissue, 
some of which was secondary. The secondary inflorescence was truly lateral 
in position, and its vascular supply passed out from the bundle system of 
the primary inflorescence in the same manner as the bundle supply to the 
female flower. The node from which the secondary inflorescence arises had 
previously borne several series of male flowers, as is indicated by the broken 
trunks of the traces which supplied them. The female flowers which fill up 
the ring on which the secondary inflorescence stands are incomplete. 
Although the evidence of earlier stages is wanting, the conclusion that the 
primordium which normally develops into a female flower may, under 
certain circumstances, produce a normal cone, appears to be abundantly 
justified. 
In both male and female inflorescences in G. scandens and G. Gnemon the 
normal nodal disposition of cupule and floral ring is frequently replaced 
partially or entirely by a continuous spiral. In the case figured a male in- 
florescence of G. Gnemon (Ceylon) — the three lower whorls are tilted but lemain 
distinct, while, in the upper part, the floriferous band with its subtending 
cupule traces a right-handed spiral (Fig. 1). Other cases have been seen in 
the male inflorescences of the same species from Buitenzorg. In a female 
inflorescence of G. scandens from Darjeeling there were two distinct " hoiG 
at the bottom, and, above these, 13 continuous turns of a right-handed spiial. 
s Berridge, 1912. 
1 Pearson, l.c. 
s Thoday, 1911, p. 1130. 
12—2 
