88 
W. B. Crow 
be demonstrated in the living cell of many forms, that there are 
three parts, namely the central region, rich in central granules, the 
peripheral protoplasm containing the pigment and cyanophycin 
granules, and the cell-membrane. In both the filamentous and uni¬ 
cellular types some species having these parts clearly differentiated, 
and other species where they are not so distinct, occur. Chodat( 5 ) 
has recorded a more or less definite central region for cells of Chroo- 
coccus turgidus (Kuetz.) Naeg. Olive. ( 15 ) and Gardner ( 9 ) described 
such a region as occurring in species of Glceocapsa. Some of the Chroo- 
coccaceae have been more recently investigated by Miss Acton (l) who 
found that a definite central region was differentiated in the cells of 
Chroococcus macrococcus (Kuetz.) Rabenh., Merismopedia elegans A. 
Br. and Aphanothece prasina A. Br., whilst in those of Merismopedia 
glauca (Ehrb.) Naeg. and a species of Glceocapsa no such specialised 
portions of the protoplast appeared; Chroococcus turgidus (Kuetz.) 
Naeg. and others showed an intermediate condition. Finally, in the 
genus Myxobactron , according to West ( 18 ), there is no central region. 
The present writer, working with material fixed in formalin, has 
found a very definite central region in the protoplasts of Microcystis 
flos-aquce (Wittr.) Kirchn. and Merismopedia elegans A. Br., staining 
with the fuchsin or methylene blue methods described by Kohl(i2). 
The deeply staining material of this central region varied in arrange¬ 
ment in different individuals, but it is a significant fact that the 
differentiation of this region was found to be constant for a large 
number of cells examined in both species. It is possible therefore 
that the degree of development of the central body might afford 
some guide to the classification of the Chroococcaceae, although 
further work is of course very essential in this connection. 
Of the remaining cytological characters many are very dependent 
on external influences (12) and cannot be brought to bear on the 
present problem. Such are the presence or absence of cyanophycin 
granules, central granules, glucoprotein, glycogen, sugar and fatty 
oils. All these substances, too, are widespread in the Cyanophyceae 
and hence differential chemical characters are not easy to establish. 
The presence or absence of pseudovacuoles has been used by 
Lemmermann ( 14 ) to distinguish various species of Microcystis and 
Ccelospheerium. He finds that the species of the former genus fall 
into two biological groups. The first consists of those practically con¬ 
fined to the plankton, often forming “water-bloom” and containing 
pseudo vacuoles in their cells ( e.g . M. ceruginosa Kuetz., M. flos-aquce 
(Wittr.) Kirchner), the second of those which are without pseudo- 
