Nuclear Division in the Hepaticae. 507 
strain exerted on the nucleus takes place in four directions, 
each of which corresponds to the localization of a mass of 
protoplasm which is comparatively isolated from its neigh- 
bours by reason of the four pouch-like outgrowths of the 
spore-mother-cell. Furthermore the appearance of centro- 
somes in these cells, wherever they can be best made out, is 
not of such a character as to inspire confidence in their claims 
to individuality. They vary both in size and in number, even 
within the same cell, and this irregularity they share in common 
with animal centrosomes, as for example in the tissues inves- 
tigated by F. Reinke and by Heidenhain. But if they mark 
mere nodal points, as it were are granules which have been 
pushed or drawn into a position of stable equilibrium in the 
centre of those forces of which the radiations may not un- 
reasonably be regarded as the optical expression, there seem to 
be no grounds for expecting them to exhibit uniformity, either 
in size, number, or texture. In all of these respects we do, 
as a matter of fact, find differences and variations, both in 
animals and in plants. 
Again, it is by no means obvious, on the assumption of 
their individuality, why the centrosomes in these four-lobed 
cells should take up positions, and start systems of radiations, 
which in most cases are soon abandoned. In Fossombronia 
and in Aneura , in which plants I obtained the clearest cases of 
four equidistant centrospheres, these structures gradually 
approximate, and I believe (though I did not actually observe 
the process) that they finally coalesce in pairs to form the 
first karyokinetic spindle which, when mature, is bipolar , as in 
ordinary nuclear division. Those plants in which the poles of 
these spindles do not fuse to give rise to the bipolar condition, 
always disclose good reasons why this should not be the case, 
as for example in Pallavicinia decipiens \ in which the lobing 
of the cell is carried to such a point that no room is left for 
a normal straight spindle to be formed. 
Thus it would seem that in cases such as that of Fossom- 
bronia , the direction of the strains which are exerted on the 
1 Farmer, Studies in Hepaticae, Ann. of Bot. 1894. 
