Nuclear Division in the Hepaticae. 471 
plasm or kinoplasm), from which extends a beautiful system of 
radiations, some passing towards the cell-periphery, whilst 
others curve over the surface of the nucleus. A single central 
body could in some cases be easily distinguished, although 
I failed to find it in others. Shortly after the appearance of 
these structures, the nuclear elements begin to alter in character. 
The linin becomes more chromatic, and disposed more regu- 
larly just beneath the nuclear wall. The nucleolus, so far as 
I could determine, does not begin to change until after the linin 
has begun to do so ; but in the stages which rapidly follow, it 
is seen to alter its form. It then becomes angular, and, in 
extreme cases, almost star-shaped, whilst at the same time the 
linin begins to exhibit a very striking increase in the amount 
of chromatin which it contains. This substance is not evenly 
distributed, but is aggregated along certain tracts of the 
filament, and eventually the intervening parts give way and 
thus set free the chromosomes. But this is only accomplished 
when a considerable interval of time has elapsed after the 
chromatic tracts have been well mapped out. The young 
chromosomes then appear as rods, which are often bent in an 
elbow-sort of fashion, and they lie much more irregularly in 
the nucleus than in both previous and subsequent stages. 
The nucleolus, which now is very much distorted in form, is 
often connected with several of the chromosomes, and it soon 
-afterwards disappears. The centrospheres at this time become 
very prominent, and indeed they are the finest I have seen in 
any plant-tissues. The chromosomes split longitudinally, and 
almost immediately afterwards the achromatic spindle is com- 
pletely differentiated, and the chromosomes, or rather the 
still closely approximated respective halves, become regularly 
arranged equatorially around its periphery. I could not follow 
out with certainty the stages in the formation of the spindle, 
owing to the suddenness with which it appears in its fully 
formed condition. Whether it is in this case largely of nuclear 
origin, as seems not improbable, or whether it is of cytoplasmic 
nature, must remain for the present undecided. The spindle 
when fully formed still exhibits radiations at its poles, though 
