condition.

[Illustration: FIG. 128.--Later stages of nuclear divisions in the
pollen mother cell of wild onion, x 350. All the figures are seen from
the side, except _B_ ii, which is viewed from the pole.]

Although this is almost impossible to demonstrate, there are
probably as many filaments in the nuclear spindle as there are
segments (in this case about sixteen), and along these the nuclear
segments travel slowly toward the two poles of the spindle
(Fig. 128, _A_, _B_). As the two sets of segments separate, they are
seen to be connected by very numerous, delicate threads, and about
the time the young nuclei reach the poles of the nuclear spindle,
the first trace of the division wall appears in the form of isolated
particles (microsomes), which arise first as thickenings of these
threads in the middle of the cell, and appear in profile as a line
of small granules not at first extending across the cell, but later,
reaching completely across it (Fig. 128, _C_, _E_). These granules
constitute the young cell wall or "cell plate," and finally coalesce
to form a continuous membrane (Fig. 128, _F_).

The two daughter nuclei pass through the same changes, but in
reverse order that we saw in the mother nucleus previous to the
formation of the nuclear plate, and by the time the partition wall
is complete the nuclei have practically the same structure as the
first stages we examined (Fig. 128, _F_).[15]

[15] The division is repeated in the same way in each cell so that
ultimately four pollen spores are formed from each of the original
mother cells.

This complicated process of nuclear division is known technically as
"karyokinesis," and is found throughout the higher animals as well
as plants.

The simple method of fixing and staining, just described, while giving
excellent results in many cases, is not always applicable, nor as a
rule are the permanent preparations so made satisfactory. For
permanent preparations, strong alcohol (for very delicate tissues,
absolute alcohol, when procurable, is best) is the most convenient
fixing agent, and generally very satisfactory. Specimens may be put
directly into the alcohol, and allowed to stay two or three days, or
indefinitely if not wanted immediately. When alcohol does not give
good results, specimens fixed with chromic or picric acid may
generally be used, and there are other fixing agents which will not be
described here, as th