1898-99. | ON THE CYTOLOGY OF NON-NUCLEATED ORGANISMS. 469 
lower proteids. As they are abundant in the rapidly growing and 
assimilating forms it is possible that their substance represents the first 
assimilation stage of the proteids synthesized in these organisms. As 
the granules disappear in trichomes which are not assimilating freely, 
they would appear to represent the reserve material of the cell, as 
Nadson and others have claimed. 
In Cylindrospermum mazus only one kind of granules is present and 
these are usually very large and homogeneous. They resemble the 
granules of the first type in that they have given evidence of the presence 
of “masked” iron’ and also those of the second type in their capacity 
for absorbing and retaining picrocarmine. I am unable to say whether 
they contain organic phosphorus, since I have not had at my disposal, 
during the last two years, any specimens of the organism to determine 
this point. 
In Nostoc commune and Oscillaria tenerrima only one kind of granules 
is present. In the latter they are comparatively large and usually 
hollow in the centre. (Fig. 16). 
Vie HETEROCYST. 
The development of the heterocyst depends on degenerative changes 
in the cell which at the outset is free from granules. The first evidence 
of the change occurs in the disintegration of the central body, which 
appears to be constituted of a coarse network in which ill-defined 
granular swellings are seen. This network stains deeply in hematoxylin 
and as the change progresses it fills the whole of the cytoplasm (Fig. 17, 
two lowest cells). At the same time a large granule or mass develops 
at one pole of the cell which quickly acquires a marked affinity for 
picrocarmine (Fig. 15) like that exhibited in the granules of the second 
type. Very often the cell which is to form a heterocyst is incompletely 
separated from its neighbour on division, and in the connecting strand of ° 
cytoplasm which extends through an opening in the transverse septum, 
the mass in question develops as a pear-shaped body with its terminal 
points in thetwo cells. (Fig 17). This structure has been referred to by 
Borzi, who believes that.it is formed of cyanophycin. When the develop- 
ing heterocyst is intercalary there may be such a mass at each pole ot 
the cell, but there is only one when the cell is terminal. 
In the next stage the structures of the central body dissolve completely 
1 The extent of the reaction for iron obtained is indicated in Fig. 8, Plate 10, ‘Quart. Journ. Micro. 
Science,” Vol. XXXVIII. 
