18 
Olive, Mitotic division. of the imclei of tlie Cyanophy ceae. 
figs. 2, 3, 11, 13, 77—79. It will be seen by contrasting figs. 8 
and 10, e. g., with figs. 11—13, that tlie central body in tlie 
former drawings is made np botli of deeply staining cbromatm 
substance, and of an acliromatic portion: wliereas, in tbe latter 
preparations, no sucli differentiation is clearly visible. Especially 
in figs. 11, 12, and 13, which are from slides colored with. 
methylene blue, iron haematoxylin, and Flemming's triple stain 
respectively, tlie dense central body appears to be homogeneous, 
as is clainied by Palla. These preparations were simply over- 
stained and not sufficiently washed out. The same is trne of 
hgs. 2 and 3, with the difference that the washing out of the 
stain has been carried on a step further, so that a portion onlv 
of the acliromatic substance remains deeply stained. Tliis resnlts 
in dark streaks of chromatin and achromatin, whicli may sorne- 
times give the appearance of long chromosomes, such as are 
figured by Kohl and "Wäger. 
In all the forms studied by the writer, including species of 
Oscillatoria , Phormidium, Calothrix , Kostoc . Gloeocapsa and Cy- 
lindrospermum, botli chromatin and achromatin could be made 
out in the central body, in properly differentiated preparations. 
Two striking pecnliarities were at once noted. First, that the 
acliromatic portion appeared ,to be often made up of an unu- 
sually dense substance; and, secondly, tliat the chromatin gra¬ 
nulös seemed relatively very minute. Particularly in Cylindro- 
spermum were these pecuharities noticable, for it could not be 
determined, even with the highest available magnification, that 
the achroniatic portion was made up of fibrous protoplasm. as 
could be demonstrated in nearly all the other cases; and further, 
the chromatin granules were so minute that they long escaped 
detection (figs. 80, 82—85, 89, 90). 
The extreme density of tliis kinoplasmic. fibrous mass (called 
by Palla Füllsubstanz“), which makes up tlie bulk of the cen¬ 
tral body, is prcbably mainly responsible for the inability of the 
majority of investigators to detect the chromatin granules en- 
closed within it. Particularly in cross - sections of the actively 
dividing nuclei of Oscillatoria princeps and 0. Frodiehia can the 
fibrous nature as well as the density of the achromatin, after 
careful examination, be made out (note, e. g., the mass in the 
middle of hg. 18). It will be further seen that, in the longi¬ 
tudinal sections (figs. 10 and 11), the density of the kinoplasm 
varies, although, in tliis view, its fibrous nature is not so easily 
demonstrable. 
In all the forms studied, with the one exception of Cylmdro- 
spermum , the writer has discovered that the number of the mi- 
nute chromatin granules is constant for the same species. The 
fact that, in Cylindrospermum , the ceils are comparatively long 
may liave prevented me findmg a cross-section in which the 
dir omo so nies, as I have called them, were grouped favorably for 
counting. For example, in some cross sections, as few as fotir 
can be counted, in otliers, six, or even teil by focussing up and 
