724 
DUNCAN S. JOHNSON 
over the male, female and hermaphrodite spikes, female flowers, 
which show a wide range of stages of degeneration (figs. 46-51). 
Moreover, in any flowers with normal ovaries, the ovules seem 
to degenerate after reaching various stages of development. 
This often begins before the time of pollination and hence can- 
not in these cases, at least, be attributed to the lack of a stimulus 
due to this process. The first stage noted in the series showing 
progressive degeneration of the ovary itself is one in which, 
though the carpels have closed together above the ovarian cavity, 
and the stigmas seem normally developed, the ovary is without 
a trace of an ovule (fig. 46). In other cases the ovarian cavity 
may be as reduced as in those just mentioned, and in addition, 
the stigmatic lobes markedly retarded in growth (fig. 47) . In the 
flower shown in figs. 48 and 51 the ovarian cavity has entirely dis- 
appeared, while the stigmas are represented by a small spine, 
with no trace of three distinct constituent lobes. The stigmas 
of the flower shown in fig. 49 are represented merely by a slight 
mound. Finally, in the flower shown in fig. 50, there is no trace 
whatever of carpellary tissues. 
The above mentioned cases were taken from spikes in which 
most of the flowers seemed functionally hermaphrodite, and in 
all the flowers figured the pollen seemed to be perfectly normal. 
Of the functionally male spikes studied, only two or three spikes 
were found in which the ovary was not represented by a slightly 
or considerably developed rudiment, in from 5 per cent to 30 
per cent of the flowers. Often an ovarian cavity is present, and 
rarely a well-organized ovule. No case, was seen, however, where 
fertilization seemed to have occurred in an ovule on a male spike. 
The ovule of P. betel arises as a mound on the floor of the ova- 
rian cavity. It is first distinguishable just after the carpels have 
closed in to surround the cavity (figs. 2, 4). 
The inner integument is initiated at about the time the primary 
sporogenous cell is undergoing its first division (figs. 33, 52). 
The tips on the carpels have at this time closed tightly above the 
ovule to form a rather long stylar canal (figs. 45, 52). The outer 
integument starts soon after the inner (fig. 53), but never grows 
above the latter to take part in the formation of the micropyle. 
