1903] MITOSIS IN PELLIA 35 
rays converge in jig. 75 is also a centrosphere, and the densely 
staining masses at the poles of the spindle in fig. 6, although not 
organized into a definite body, consist of the same material as 
centrospheres and, at an earlier stage in mitosis, may have had 
amore definite form. We have not intended to represent a 
centrosome in any of our figures. Bodies which have the super- 
ficial aspect of centrosomes are shown in figs. 14, 16, and 17, 
but here the sharply staining body at the center of the centro- 
Sphere is, without doubt, the cut end of an astral ray. The 
Structure at the upper pole in fig. 9 certainly looks like a centro- 
sphere containing a centrosome, but such an appearance is so 
rare that it seems safer to regard the sharply staining body as a 
chance granule. Still, it is evidently just such a body as this 
that Van Hook (38), in his recent study of Marchantia, inter- 
prets as a centrosome. 
In the very early prophases a beautiful system of radiations 
becomes quite conspicuous. This system we regard as an aster, 
comparable with the asters of thallophytes and of animals. 
The system first appears as a few fibers converging to a point 
which is usually in contact with the nuclear membrane or very 
near to it (jigs. rz—13), but, in some instances, may be at a 
considerable distance from the nucleus (jigs. 14-16). Persist- 
€nt search failed to reveal any body which could be identified 
positively as a centrosome or centrosphere before the appear- 
ance of the aster, and even after the appearance of the aster and 
centrosphere, no centrosome could be distinguished. Granules, 
like those shown in all the figures, were frequently found in con- 
tact with the nuclear membrane after the nucleus had begun to 
enlarge, and it is probable that some of the granules were cen- 
trospheres, although no method was found for identifying them 
before the appearance of the rays. Bodies which bear remark- 
able resemblance to centrosomes (figs. 14, 16, 77) and which. 
for atime, were interpreted as genuine centrosomes, proved to 
be merely the cut ends of coarse fibers. Sometimes several 
deeply Staining points may be seen; such an appearance might 
€asily be mistaken for a centrosphere containing several gran- 
ules, In cases like those shown in figs. 14-77, the “granules” 
