528 
CLIFFORD DOBELL. 
possible to ascertain how the chromosomes divide, on account 
of their extremely small size. But since the chromosomes on 
the equatorial plate appear to be short rods (PL 27, fig. 14),. 
and those at the metaphase smaller granules (PI. 27, fig. 15), 
it seems probable that the cliremosomes themselves divide by 
a transverse constriction into two. 
Anaphases. — The daughter-groups of chromosomes now 
move apart, gradually passing towards the poles of the spindle 
(PI. 27, figs. 16, 17, 18). As they do so, they become still 
more closely packed together, and individual chromosomes 
can no longer be resolved. In the early anaphases (PI. 27, 
fig. 16) distinct spindle-fibres can be seen between the 
chromosome groups, but later the fibres become less distinct 
and more irregular (PI. 27, figs. 17, 18). Simultaneously, the 
differentiated structure of the spindle becomes less distinct, 
and finally vanishes. 
During the anaphases the centroplasts at the poles of the 
spindle also undergo certain changes. Their central granules 
are often much drawn out during the later stages, so that 
they appear in optical section as fairly long rodlets (PI. 27, 
figs. 17, 18). These then shorten, and thicken somewhat, and 
they then become bent into the form of incomplete rings, open 
on the side nearest the spindle (PI. 27, fig. 20). Finally, a 
closed ring is formed, which becomes converted into a 
minute, homogeneous, darkly- staining central granule in each 
daughter-centroplast (PI. 27, fig. 19). All these changes 
occur during the late anaphases and early telophases, and 
they do not always synchronize with the nuclear changes. 
For example, PI. 27, fig. 20, shows an earlier stage in the 
reconstruction of rhe centroplast than PI. 27, fig. 19, though 
the nuclear stage in PI. 27, fig. 19, is earlier than that in 
PI. 27, fig. 20. Centroplast and nucleus are thus independent 
of one another to some extent at this period. I have found a 
good deal of variation in this respect in different individuals. 
In at least one specimen I have seen the centroplasts com- 
pletely reconstructed, and spindle-fibres no longer visible, 
before the telophases had begun. 
