434 BOTANICAL GAZETTE [JUNE 
at right angles from the nuclear surface and extend to the walls 
of the cell. They appear like ordinary cytoplasmic radiations 
with numerous microsomes (fig. 75). Whether these threads 
are the same as the longitudinal threads of earlier stages I did 
not determine. However this may be, there is no indication 
of such crossing threads at this stage. If they were present 
they should have appeared as well as the divergent ones. It 
may be that the function of these radiations is to carry the 
micronucleoli out into the cytoplasm. The micronucleoli are 
perfectly differentiated by various stains. With anilin-safranin 
and gentian-violet they have a brilliant red appearance, which 
makes them stand out more prominently in the sections than 
they do in the figures. With cyanin and erythrosin the nucleoli 
are blue, while the surrounding cytoplasm and nucleus are red. 
They are also differentiated by other stains. The mother 
nucleoli continue to become smaller, and sometimes some of the 
chromosomes are collected around the nucleoli in such a man- 
ner as to suggest that the nucleoli have something to do with 
the growth of the chromosomes (figs. 22a, 23a). This, however, 
is doubtless merely an appearance, since twelve chromosomes 
have not much opportunity to avoid contact with two nucleoli in 
so small a space. Usually the greater number of chromosomes 
ina nucleus do not lie in the proximity of the nucleoli at all 
(figs.21-23). By the time the chromosomes are ready to be 
arranged at the equator into the mother star the original nucleoli 
have entirely disappeared, the small daughter nucleoli or micro- 
nucleoli being scattered through the surrounding cytoplasm 
(figs. 24, 25). The micronucleoli show a tendency to become 
_ placed near the periphery of the cell and away from the nuclear 
a spindle. However, they are often seen quite close to the poles 
and on the spindle, where with improper staining they might be 
confounded with the centrospheres. With proper staining, how- 
ever, there is no possibility of such confusion, for the nucleoli 
in the -cytoplasm> have an entirely different structural appear a 
"ance from the centrospheres, and also show a different staining 
ee : reaction. There i is no doubt i in ~ mind that the ees - 
ret 
