Notes. 
149 
One isolated case was found, in a cupule with four flowers, in which the most 
centrally placed of these was situated on an upgrowth from the base of the cupule 
(see 9 and 10, Fig. 1). Here the distal portion of the upgrowth was continued 
laterally past the flower, giving rise to two structures of exactly the same nature as 
the exterior cupule segments. It may be stated here that all the flowers examined in 
the later stages proved to be abortive, although ovules were present, even in the 
rudimentary ones. 
Androgynous Inflorescences. 
In these a floral series may be present leading from normal female flowers on the 
one hand, through hermaphrodite forms with reduced androecium and a more or less 
rudimentary inferior ovary, to typical male flowers with or without a vestigial superior 
ovary and the full complement of stamens on the other (Fig. 1,1, 2, 3 and 4). As 
described by Eichler, 1 the male flowers may or may not have such a vestigial pistil 
present within the perigone, and when it is present may vary in structure from the 
less reduced type, with three stylar arms and a slightly enlarged base representing the 
ovary, to the extremely reduced condition with two stylar arms and scarcely any 
enlargement of the basal ovarian portion. The degree of approximation of the extra 
flowers to the normal male or female type can be correlated with their position on the 
cupule. Thus one may say broadly that the higher the insertion of the flower on the 
cupule segment, the closer does it approach normal male structure. Fig. 1, 5, is 
a somewhat diagrammatic representation of an androgynous inflorescence with one 
female flower in the normal position : of the two rudimentary flowers on the lateral 
walls of the cupule, one is hermaphrodite, while terminally on the segments there are 
present several male flowers. The one on the right hand, although having a perigone 
and vestigial pistil, is destitute of stamens. No extra flowers were found on the 
outside lateral walls of the cupule in any of the material examined. The anthers of 
these flowers were all in the post-dehiscence stage at the time of examination, but they 
appeared to be of the normal build. * 
Transitional Inflorescences. 
Fig. 2, 1, is a vertical section of a specimen with eighteen male flowers situated 
peripherally on the much-reduced capule. The latter was unsegmented and the slight 
peripheral upgrowth sfirrounded a shallow 
depression in which were situated one her¬ 
maphrodite and one female flower, both 
being poorly developed. A curious stage is 
shown in Fig. 2, 2. The much-attenuated 
stalk, scarcely thicker than that of a normal 
male inflorescence, divided distally into two 
short arms, each of which terminated in 
a flower with a rather large perigone (sic) 
but entirely destitute of androecium and 
pistil. Numerous male flowers were clustered on each of the two arms (only a few 
of these are shown in the drawing). 
Fig. 2, 
1 and 2, transitional inflorescences > 
3, weakly developed male do. 
1 Bliithendiagramme. 
