134 
Charles E. Allen 
fibers from one plate cross those from the other at various angles; onlv 
rarely, as yet, does a fiber connect the two plates. 
When the divergence of the polar plates has reached approximately 
180° (Figs. 8 — 10), a very considerable number of fibers are present, 
determining already the position, shape and extent of the future spindle. 
Many fibers run, still in a rather haphazard manner, from either plate 
toward the other. Some extend from plate to plate, others end either 
in the cytoplasm or in contact with the nuclear membrane. There is no 
evidence as yet of any connection between the cytoplasmic fibers and the 
nuclear contents. Occasionally short fibers connect a plate with the ad- 
jacent plasma membrane (Fig. 8); but in the main the fibers are confined 
to that part of the cell which lies between the polar plates. 
If the plates lie in the plane of the section, they are, as might be ex- 
pected, transparent enough not to be plainly visible. When they are 
distinctly seen, therefore, it is usually in edge view. But if (as in Figs. 5, 
6 and 8, and in the upper cell of Fig. 1) the plates are viewed diagonally, 
something of their breadth and outline is apparent. In Figure 1 (upper 
cell), the irregulär outline of the plate is marked, and its edges are seen to 
run out into fibroid prolongations. Something of the same sort appears in 
Figure 5. Such prolongations are often evident also in focusing upon edge 
views of plates. In other instances (PI. VI, Fig. 7; PI. VII, Figs. 31, 32), 
the appearance of the plate suggests not merely that it is irregulär in shape, 
but that it may consist of a number of smaller bodies lying close together 
or in contact. In still other cases, the cell contains, according to its 
stage of development, one or two groups of plainly separate bodies (e. g., 
Figs. 11 — 18, PI. VI). There may even be an apparently continuous plate 
at one spindle pole, and at the other a group of smaller bodies (PI. VII, 
Figs. 32, 34, 58; PI. VIII, Fig. 59); and all possible transitions may be found 
from the condition of a compact plate to that of many distinct bodies. 
It seems plain from the appearances described that the plates and the 
groups of smaller bodies — which will be referred to as kinetosomes — 
are mutually equivalent, being merely different forms assumed by the 
same substance. In general, polar plates are present from the period 
of the earliest formed androgones to a time when a median longitudinal 
section of the antheridium contains in its widest part five or six rows 
of androgones ; later tlian this, they are commonly replaced by groups of 
kinetosomes. But to this rule there are many exceptions. For example, 
Figure 6, Plate VI, shows a cell with polar plates found in an antheridium 
with seven rows of androgones at its widest part; Figure 10, one from 
