Nuclear Division in the Hepaticae. 493 
I hope, however, to be able to settle definitely this point, 
both in Fegatella and in those other liverworts in which I have 
so far failed to reach a position of certainty as to the exact 
mode of procedure. The omission here is due to the great 
difficulty of getting enough material suitably fixed at exactly 
the right time. This being so, it seems better to avoid specu- 
lation, which further observation may only show to be unsound. 
But whatever the exact mode of origin of the chromosomes 
previous to the final karyokinesis may be, there is no doubt 
whatever as to the further course of events. The eight 
chromosomes possess, as has been said, exactly the same 
shape as those in the first division, and when they divide 
at the equatorial plane, they exactly resemble the chromo- 
somes of the first heterotype division. The daughter- 
chromosomes now retreat to the spindle poles, and radiations 
extend from each daughter-nucleus to the two others which 
are nearest to it, and, further, a spindle appears connecting the 
two nuclei which are situated at the obtuse angles of the 
rhombus formed by the peripheral spindles (Fig. 59). In this 
way five spindles are formed altogether, and in each one 
a cell-plate is formed. These all unite ultimately, and thus 
the spore-mother-cell is divided into the four special mother- 
cells. The elements of the nuclei have now begun to enter 
into a state of rest, and they do so in the reverse order in 
which they came out of it. The nucleoli reappear in all cases, 
several in number, and they are always associated with the 
linin thread which has been re-formed, by the fusion of the 
chromosomes. The filament shows frequent anastomoses, 
and perhaps this points to the fusion of the chromosomes not 
being so regular as is sometimes supposed. As the nucleoli 
increase in number and in total bulk, the linin becomes thinner 
and diminishes in its staining capacity. Gradually the nucleoli 
run together, and though vacuolated at first, they lose this 
character as the whole process nears its termination, until 
finally there only remains in each spore-nucleus a single large 
nucleolus 1 , and a diffuse and somewhat indistinct tangled 
1 It is not meant to imply that the nucleolus becomes the receptacle for the 
M m 
