STUDIES OF THE DEVELOPMENT OF PEPEROMIA HISPIDULA 333 
thecium are represented in the mature stamen only by nearly collapsed 
cells with degenerating nuclei. The cells of the tapetal layer itself 
are well-developed at the time the microspore mother-cells are mature, 
and have dense cytoplasts with large nuclei (fig. 32). With the 
maturing of the anther the cytoplasts become thin, and the nuclei 
follow those of the other layers in gradually degenerating. 
The opening of the ripe anther is evidently accomplished chiefly 
by the change in form of the endothecial cells, though this process 
has not been studied minutely. The rupture of the anther walls 
occurs along the series of smaller cells at the juncture of the two micro- 
sporangia (fig. 34). In sections of the opened anther (fig. 22, above 
middle of spike) it is seen that the outer walls have bent in and in 
consequence the anther walls have become completely everted, and 
the inner surface strongly convex. Soon after the anther is emptied 
the filament breaks off, leaving only a short stump projecting from 
the base of the ovary (fig. 22). 
From the definitive archesporium, just within the parietal layer, 
are developed, chiefly by periclinal divisions, the one hundred or more 
microspore mother-cells of each microsporangium (figs. 31, 32). The 
sporogenous cells increase in size from a diameter of 8 or 10 ^t, when 
first cut off, to one of 12 to 18 at the time of synapsis in the spore 
mother-cell (figs. 29, 32). The nuclei in the meantime increase from 
5 or 6 )U to 10 or 12 )U in diameter, and the bulk of the chromatin also 
increases slightly up to the time of synapsis. The nucleoli of the 
definitive archesporial cells are usually single, and do not vary greatly 
from 2 or 3 /X in diameter in the resting nucleus throughout the develop- 
ment. 
It will be well to emphasize at this point the fact that though the 
number of sporogenous cells in each sporangium varies, as was sug- 
gested in the last paragraph, no cases were seen in which the number 
of microsporangia was greater or less than two. There is thus no evi- 
dence from the development that this stamen ever had more than two 
microsporangia, and no examples were seen of the marked variability 
in the number of degrees of development of these that occurs in 
Piper betel (see Johnson, 1910). The only suggestions of this are 
found in those rare cases where the terminal flower of a spike may 
be pistillate only, and in the fact that from 10 per cent to 25 per cent 
of the nearly matured microspores in a stamen may prove sterile 
(fig. 34). The latter condition is not unusual among angiosperms as 
