418 BOTANICAL GAZETTE [JUNE 
nucleolus to form chromosomes on the spireme, as the compound 
appearance of that body in C. Ehrenbergii especially would indi- 
cate might happen. There seems to be no question that this 
body loses part of its material during the time of spireme formation, 
as its diminished size clearly shows, but it is no more possible here 
than it is in the flowering plants to prove directly that it is used 
to form the spireme. The spindle fibers are being formed at either 
pole, and it is entirely possible, as STRASBURGER believes, that it 
may be utilized in their formation. The apparently small quantity 
of chromatin in the reticulum, combined with the small size of the 
chromatin granules observed on it as compared with the density 
and size of the spireme, would certainly lead one to suspect, how- 
ever, that the material from the diminishing nucleolus was being 
transferred to it. Of one fact there can be no question, and that is 
that the substratum of the spireme itself arises in the extra 
nucleolar part of the nucleus, and not, as MirzEKEWITcH, BERGH, 
KARsTEN, and others have found in Spirogyra, inside the nucleolus, 
and that while the relation of the nucleolar material to this 
structure may not be clear, it does not seem to differ essentially 
from that which is found in the higher plants. : 
The spireme cuts transversely into chromosomes, no previous 
longitudinal split having been found, although it probably occurs: 
At the time this segmentation takes place, the spireme thread still 
has an irregular outline, due to small projections from its surface. 
These projections appear (fig. 24) also on the chromosome at first, 
but later disappear, and the chromosomes become long rods with 
a smooth surface. 
In the meantime the ellipsoidal nucleus is undergoing chang® 
in shape. Its smooth outline is lost and the nuclear walls are draw? 
out in places (figs. 19-21). The walls (fig. 24) are very much 
thickened at the future poles of the spindle, and a dense layer of 
fibers lies outside of them, while the equatorial walls are thin an 
irregular. Some of the figures obtained at this time resemble 
somewhat those of Hertwic (24) for Actinosphaerium, although 
the thickening in the present case is not so great as is shown 17 his 
end plates. Some of these fibers extend out and are apparently 
attached farther back in the -chromatophore, evidently ae 
