146 BOTANICAL GAZETTE [FEBRUARY 
with the starch-filled cytoplasm surrounding them, but are occupied 
by cytoplasm similar to that of the remainder of the cyst, except 
that it is free from starch. This condition endures for a variable 
period; it sometimes disappears during the binucleate stage 
(cf. fig. 23), and sometimes persists into the octinucleate stage 
(fig. 24). 
The nuclei of the early stages of the coenocyte tend to remain 
in the central position originally occupied by the primary nucleus, 
_ but later scatter, finally becoming evenly distributed through the 
cytoplasm. The period at which they disperse varies, as would 
be expected. One case was found in which they were still closely 
bunched in the 16-nucleate stage (fig. 25), but they are usually 
dispersed a little before that time. 
SEGMENTATION.—On account of certain apparently conflicting 
processes observed, the writer has not been able to satisfy himself 
altogether concerning the mechanism by which the coenocytic 
cyst is cut up into spores. The account here given is therefore 
somewhat tentative. 
During the last mitoses in the sporangium, a change seems to 
come over the protoplasm of the coenocyte. Up to this time the 
nuclei have apparently lain freely in the common cytoplasm 
without any tendency to form separate cells. But during these 
mitoses the cytoplasm appears to contract around the spindles 
and to draw up closer to them, so as to leave vacuoles in the inter- 
mediate spaces (fig. 30). These vacuoles, surrounding, as they do, 
the separated masses, often resemble cleavage furrows cutting the 
coenocyte up into individual cells. The cytoplasmic edges of the 
segments do not present the sharp clean outlines seen in progressive 
cleavage, however, but appear more or less irregularly frayed, and 
frequently cytoplasmic strands cross the vacuoles and connect 
adjacent masses. 
These connections would seem to put aside any interpretation 
of the process as due to cleavage furrows, but one cyst was observed 
in which the margins of the individual masses wereeclear and sharp, 
without any bridges across the furrows (fig. 29). This case was 
difficult to interpret otherwise than as progressive segmentation 
by cleavage furrows. 
