470 TRANSACTIONS OF THE CANADIAN INSTITUTE. [VoL. VI. 
in the cytoplasm which now attains a light greenish-yellow colour. The 
mass of picrocarmine-staining substance persists, but otherwise the 
cytoplasm is homogeneous, and it does not stain with hematoxylin, picro- 
carmine, or with any of the aniline dyes. It gives a feeble but uniform 
reaction for “masked” iron, a slightly more marked reaction for phos- 
phorus and it is not affected during prolonged action of artificial gastric 
juice, which also does not dissolve the mass at the pole of the cell. The 
latter gives a distinct reaction for “masked” iron. This result indicates 
that the substance forming the mass is not related to the substance, 
cyanophycin, forming the granules of the second type, and that, therefore, 
Borzi’s supposition as to the composition of the mass is incorrect. The 
heterocyst is, therefore, a degenerated cell. It may also possibly 
represent more than this. Its formation next to the spore in Cylindro- 
Spermum majus and other forms, as well as its development beside the 
cell. out of which arises the lateral trichome branches in Tolypothrix 
would appear to suggest that the heterocyst may be the result of some 
rudimentary sexual process. 
VII.—CELL -DIVISION. 
Ordinarily the first sign of cell division in the Cyanophycez is the 
growth inward from the lateral wall of a septum which appears thus to 
separate, not only the cytoplasm as a whole, but also the central body, 
into two equal parts. It is, however, only in such genera as Tolypothrix 
and Scytonema, in which the cells are, in comparison to their length, very 
long, that one is able to ascertain what are the earliest phenomena of 
cell division. In preparations of trichomes from quickly growing 
cultures one may find a cell in which the deeply stained central body has 
a constriction which gives it a slight hour-glass appearance. In some 
other cells, perhaps in the same trichome, the constriction may be greater 
and at the same time the ingrowth from the lateral membrane to form 
the transverse septum may be found. The constriction of the central 
body, before the appearance of a trace of a septum, must be held to in- 
dicate that the central body initiates division. 
The mode of the formation of the transverse septa explains the central 
perforation or opening in the septa which Borzi saw. In old septa I 
have never observed these passages or perforations. They are really due 
to incomplete formation of the septa. 
In the division the granules are apportioned to the daughter cells 
according to their distribution. There is no rearrangement, no grouping 
