1887] The Significance of Sex. 155 
cell-division is purely facultative, and has not been inseparably 
associated with nuclear division. When it takes place, as in 
Opalina, any number of nuclei may be separated away in the 
daughter-cell. In many cases the nuclear divisions affect only 
the chromatin, while the hyaloplasm and its nucleolemma remain 
as funiculi uniting the segments. (See Figs. 42, 31, 29, 28, 41, 
etc.) The microsomata of a nucleolus or of a karyosoma behave 
in the same way; as in Figs. 19, d, 43, 114, 115, 123, g, 124, 9. 
In Radiolaria we get a multitude of small “ massive” nuclei 
that divide by amceboid constriction (Fig. 25), and in the central 
capsule vesicular nuclei, whose different metamorphoses are 
shown in Figs. 16, 17, 18, and I9. 
In division of vesicular nuclei we get, as the simplest form, the 
Remakian scheme. (Fig. 100.) Next we may get the granular 
contents arranged in fibrils that are bisected by the constrict- 
ing nucleus (Fig. 101), or they may remain in the granular state. 
(Fig. 99.) The most complex case is given by Actinospherium 
eichornii. (Fig. 103, a-g.) Here the nucleolus separates into 
two bodies, one containing chromatin and oné the parachromatin ; 
then each body segments into fine granules; these granules get 
arranged in fibrils; the chromatin-granules fuse to karyosomata 
lying in an equatorial plate in the spindle formed by the parachro- 
matin; the karyosomata divide into daughter-karyosomata that 
pass towards the poles; on its way each daughter-karyosoma 
segments to microsomata; the microsomata segment to granules, 
which, however, form separate karyosoma-like masses until they 
are absorbed by the polar plates. The latter are due to the fact 
that the external protoplasm had gathered at two opposite poles 
of the nucleus and had attracted the parachromatic cortical layer 
of the nucleus. Why the protoplasm should gather at the poles 
and so induce nuclear division is unknown. Possibly substances 
in the nucleus have first passed to the poles and attracted the 
cytoplasm. These substances may be the segments of the para- 
nucleolus, for it is possible to derive the spindle-fibres from a dif- 
. ferentiation of some of the chromatin-granules into hyaloplasm. 
It is rare to find the nucleolus (or the granules that represent 
it when it is segmented) in the same condition during nuclear di- 
vision as during the resting phase. There is a cycle of changes, 
so that one condition has to be assumed for division, and then 
the chromatin returns through inverse stages to its resting state. 
