Cell Structure, Growth and Division in the Antheridia of Polytrichum etc. 157 
a spirem begins ; and by the time the nucleus has expanded to its maximum 
size (Figs. 82 — 84), spirem formation is well under way. As in the earlier 
divisions, the exact time of disappearance of the nucleole varies some- 
what in different cells. I have not found in any nucleus of this generation 
so smooth and uniform a spirem as was shown, for example, in Figures 25 
and 26, Plate VI ; the nearest approach to this condition that I have observed 
is represented in Figures 82 and 83, Plate VIII. When the nucleus is in its 
most expanded condition (Figs. 82, 84, 86), the chromatin, in most in- 
stances observed, is loosely aggregated in a region midway between the 
poles, leaving the ends of the nuclear cavity nearly or quite devoid of 
stainable substance. It might be thought that the membrane, as the 
nucleus swelled, has withdrawn in the polar regions from the chromatin, 
leaving the position of the latter substantially unchanged. On the other 
hand, Figures 80 and 81, showing the nucleus in process of elongation, 
seem to indicate that the chromatin at first spreads out as the nucleus 
expands, and later aggregates in the central part of the nuclear cavity. 
Such an equatorial localization of the chromatin did not appear, so far 
as my preparations show, at the corresponding stages of any of the earlier 
divisions. These stages, however, axe relatively infrequent in the prepara- 
tions, and it is possible that further study may show a greater similarity 
in this respect than now appears to obtain between the different mitoses. 
Segmentation and some contraction of the chromosomes have ap- 
parently occurred in the cells shown in Figures 86 — 90, although the chro- 
mosomes are still irregulär in outline and are too closely crowded together 
to be clearly distinguished. 
The nuclear membrane now gradually contracts (PI. VIII, Figs. 86 — 89; 
PI. IX, Figs. 91, 92), as in the androgonial divisions, until the chromosomes 
are crowded into a compact knot. Commonly the membrane is present 
until a very late stage of this contraction (PI. IX, Fig. 91); but some- 
times (PL VIII, Fig. 90) it becomes indistinguishable at an earlier moment. 
Figure 88 represents a cell in which the membrane seems to have disap- 
peared unusually early on one side of the nucleus; but it is quite pos- 
sible that its disappearance in such a case is only apparent, due to a 
failure to take the stain. As Figures 86 and 87 show, the nuclear con- 
traction is sometimes at first asymmetrical, proceeding more rapidly at 
one pole than at the other. Whether the nucleus of Figure 84 has 
similarly begun to contract at one (the lower) pole, or whether the figure 
represents an earlier stage in which one end of the expanding nucleus 
has not yet reached the central body, the preparation from which it was 
drawn does not make certain. 
