466 TRANSACTIONS OF THE CANADIAN INSTITUTE. [VoL. VI. 
reagent gives granules in the peripheral portions of the central body a 
deep reddish-violet stain and these correspond to the “red ” or red-violet 
granules of Biitschli. These may be, for the present, called the granules 
of the first type. The granules of the second type stain bright red with 
picrocarmine and correspond to those which Deinega found and to the 
granules which, according to Palla and Nadson, stain blue with haema- 
toxylin. 
The granules of the first type vary very much in size. They are 
usually situated in the outer sections of the central body although 
occasionally a single one may be found in the more central portion and in 
the inner zone of the peripheral layer. The granules, as Hieronymus, 
Palla and Biitschli found, are, at times, hollow bodies, while to Zacharias 
their peripheral substance appeared to have a greater density than their 
centre possessed and the author found that all the larger granules, at 
least, are formed of an outer zone or coat which takes the hamatoxylin 
deeply, with a central cavity containing a fluid substance which is indif- 
ferent tostaining reagents. Methylene blue deeply stains these granules 
and in consequence their hollow character in such preparations is not 
evident. Acetic-methyl green gives them in fresh preparations a stain 
much deeper than that taken by the central body, while it leaves un- 
affected the granules of the second type. 
The granules of the second type correspond to the “ cyanophycin” 
granules of Borzi and Palla and, as they have a special affinity for picro- 
carmine, their presence can always be readily demonstrated. In the 
fresh cells they stain deep blue with Ehrlich’s hematoxylin. They are 
in the Osczllarie more abundant adjacent to the transverse septa and 
they appear to be situated in the thin cytoplasmic layer which is applied 
to the septum. In picric acid preparations of Osccllaria Froehlichit 
stained with picrocarmine, views of the flat faces of the discs frequently 
show a somewhat radiate arrangement of the granules, the larger ones of 
which are found near the junction of the lateral membrane with trans- 
verse septa. The granules in such cases are not exactly round, as they 
appear slightly elongated in the radial direction. In Mzcrocoleus 
terrestris, Tolypothrix, Scytonema and Lyngbta, these granules are 
distributed throughout what corresponds to the peripheral zone in these 
forms. 
It is not in staining alone that I would base the classification of these 
granules. A far stronger distinction lies in the chemical reaction of the 
granules and in the action of digestive fluids upon them. 
The granules of the second type very quickly dissolve in hydro- 
