1002 
Gates . — Somatic Mitoses in Oenothera . 
a delicate network of threads fills much of the nucleus, yet the chromosome 
centres are still visible and can be approximately counted. It is possible that 
when the mitoses follow each other very rapidly the nucleus begins a new 
division before it has passed much farther than this stage into the resting 
condition. But if the growth is not too rapid, a completely resting nucleus 
is formed (Fig. 36), corresponding exactly with Fig. 1 except in its smaller 
size. In the subsequent growth of the nucleus up to the size of Fig. 1, its 
chromatin content remains in the form of a uniformly distributed reticulum. 
It is impossible to see in these facts any support for a theory of pro- 
chromosomes. That each section of the resting reticulum is derived, in 
large part at least, from a given chromosome is probable, but there is 
no centre or core of the chromosome which retains its condensed forpn 
through the completely ‘resting 5 nucleus. 
Other Observations. 
For comparison, I have shown in Fig. 38 a single megaspore mother- 
cell in O. lata in metaphase of the heterotype mitosis. Not only the cell, 
but the spindle and chromosomes are enormously larger than in the sur- 
rounding nucellar tissue. Fifteen chromosomes can be counted, though 
they are rather closely grouped. As is well known, allotypic chromosomes 
generally differ from somatic or typical chromosomes, not only in size but 
in shape. The heterotypic chromosomes of Fig. 38 are not slender rods 
or loops, but are short and stout or almost spherical. But for unknown 
reasons the chromosomes in occasional somatic cells may approach the 
shape of the heterotypic chromosomes, as in Fig. 13, or may take other 
forms, such as that of Fig. 14, which resembles the ordinary shape in early 
telophase of somatic mitosis (Figs. 27, 28). It should be stated that the 
cells represented by Figs. 13 and 14 were both surrounded by dividing cells 
in which the shape of the chromosomes was normal, so that the variations 
in shape cannot be attributed to the fixation or subsequent treatment. 
Another observation which is worth recording here, as showing the 
relative stability of the spindle as a structure in the cyloplasm of the cell, 
and that it is not merely a reflection of osmotic stresses or electro-magnetic 
lines of force, is the following. By an accident which occasionally happens 
to every cytologist, a preparation which had been mounted in Canada 
balsam some time previously was slightly crushed, so that the cell-walls 
of a small area of tissue were destroyed by the pressure. On the margin 
of this area were several cells which were burst open and their contents 
extruded. One such cell contained a spindle in metaphase, and this was 
pressed out of the cell through the broken cell-wall. The spindle itself 
remained perfectly intact, with the chromosomes attached and undisturbed, 
though the spindle had been moved out of the ruptured cell and turned 
through an angle of 90°. This spindle remained floating freely in the 
