84 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
both an apparent disparity in their number and that of the spindle 
threads, and the unlikelihood of the tenuous threads of the spindle 
fusing through the centrosome with the coarse rays, argue against 
any such relation between spindle threads and astral rays in Hydno- 
bolites. 
Throughout the division of the nucleus the centrosome appears to 
be considerably augmented by a granular mass. Both the solid 
and the granular parts are shown in figure 7 (pi. 7). It is likely 
that the granular portion consists in part of bases of rays lying in 
other planes than in the one represented in the drawings. 
The chromosomes are very small and are four or five in number. 
The daughter chromosomes separate after cleavage, and on reaching 
the terminals of the spindle the latter begins to elongate. The 
result of this is that the nuclear wall is distorted (pi. 7, fig. 10) 
Finally the nuclear wall disappears, leaving both spindle and nucle¬ 
olus in the cytoplasm. The spindle threads elongate still farther 
before their disappearance, the chromosomes in the meantime losing 
their identity and passing into a tight clue condition (pi. 7, fig. 13). 
Eventually the astral rays are no longer visible, new walls are 
formed about the nuclei, each centrosome shows as a small solid 
mass in a position 90° removed from its position during the division 
of the mother nucleus, the chromatin again forms a network and 
each nucleus acquires a nucleolus (pi. 7, fig. 14). The old nucleo¬ 
lus begins to appreciably diminish during the anaphase stages of the 
division. Its disappearance does not keep step with the develop¬ 
ment of nucleoli in the daughter nuclei as Gjurasin (’ 93 ) thought to 
be the case in Peziza vesiculosa. Another noticeable feature is 
that the cytoplasm does not now seem to be differentiated as it was 
during and just prior to the division of the nucleus, and has become 
rather coarsely reticulated (pi. 7, fig. 14). 
After a short period of rest the division of the two daughter 
nuclei is initiated by the same differentiation in the protoplasm sur¬ 
rounding them that was observed just before the first division. The 
protoplasm becomes finely reticulated and presents a radially striate 
appearance (pi. 7, fig. 15). As before, the change in the surround¬ 
ing protoplasm extends almost to the limits of the dense cytoplasm. 
The prophase stages are identical with those already observed. It 
may be noted in this connection that the centrosomes are frequently 
not demonstrable during the resting periods between the first and 
