Cell Structure, Growth and Division in the Antheridia of Polytrichum etc. 133 
are fairly numerous; sometimes the plate forms a smaller arc about the 
nucleus than is shown in Figures 2 and 3, sometimes the nucleus is more 
nearly surrounded. The bending of the plate evidently accompanies an 
increase in its superficial area, which in turn may result either from a 
Stretching out of the kinoplasmic material already present or from an 
increase in its amount. Very likely both processes are concerned; it will 
appear, however, from facts to be described later, that the growth of the 
kinoplasm during the life of a particular cell is not in general equal to 
the diminution caused by its halving in the preceding division; hence 
the increase in the superficial area of the plate is probably due, at least 
in part, to a rearrangement of its material. 
The first visible indication of an approaching mitosis is the transverse 
division of the single plate into two (Fig. 4) and the moving apart of the 
daughter plates (Figs. 5 — 7) until they he on opposite sides of, and at 
some distance from, the nucleus (Figs. 8 — 10). Stages in the divergence 
of the daughter plates are few in my preparations; on the other hand, 
cells that show the plates in their final positions are numerous. Apparently, 
after their Separation they move rather quickly to their destined places, 
where, as we shall see, they remain while the nucleus prepares for division. 
Not infrequently, one of the daughter plates is replaced by two 
(Figs. 5, 8, 9, 10), which lie close together and parallel. A somewhat 
similar, apparently fragmented condition at an earlier stage is shown 
in Figure 4. Such figures illustrate the possibility of a breaking up of the 
material of the polar plates into smaller bodies, a process which, as will 
be seen, is carried much further in the later androgonial generations. But 
in the earlier generations now under consideration, the typical appearance, 
in spite of not infrequent exceptions, is that of a single plate in the young 
cell and of two plates in the cell which is preparing to divide. 
Even before the mother plate divides (Fig. 2), it is sometimes con- 
nected by a few achromatic fibers with the nuclear membrane. But after 
its division, and while the daughter plates are moving apart (Figs. 5 — 7), 
the fibers increase in number and length, running now from that face 
of each plate which is turned toward the nucleus, into the cytoplasm 
and usually in the general direction of the nucleus. None of the fibers 
has been seen to penetrate the nuclear cavity at any time before the dis- 
appearance of the nuclear membrane; the fibers which, in Figures 6, 10, 
16, 17, 20 and 21, seem to enter or cross the nucleus, are those which 
are visible at an upper focus, lying above and close to, or touching, the 
membrane. These fibers at no time form cones; they are attached indis- 
criminately to all parts of the surface of the polar plates (Fig. 6). The 
