66 
W. B. Crow 
The closely crowded cells are characteristic and distinguish this 
species from those of Aphanocapsa with cells of the same size and 
somewhat similar colonies. 
Loc. Tank Andankulam, four miles from Trincomalie, Oct. 20th. 
MICROCYSTIS PULVEREA (Wood) Migula (see (5)) 
Colonies spherical or ellipsoidal, often lobed. Cells 2-3/x diam., 
subspherical, without pseudo vacuoles. 
Loc. Lake at Colombo, — ; tank at Dambulla, Sept. 9th. 
Var. Incerta (Lemm.) nob. = M. incerta Lemm. 
G. M. Smith ( 7 ) points out that M. incerta Lemm. differs mainly 
from M. pulverea (Wood) Migula in the size of its cells and that it 
might be better to consider M. incerta as a variety of M. pulverea. 
Lemmermann( 5 ) says that the colonies of M. incerta are always 
isolated, those of M. pulverea “oft zu vielen dicht nebeneinander 
Legend.” The latter condition was not frequent in our specimens. 
It therefore seems that the essential difference between the two forms 
is one of size and as it is not customary to establish species on this 
character alone we have regarded M. incerta as a variety of M. pulverea. 
As in M. flos-aquce the smaller colonies, apparently produced by 
the breaking up of the larger, have much more clearly defined margins 
than the larger, i.e. older colonies. 
Loc. Rock pool connected with Kalawewa-tank, Sept. 29th. 
Forma Elongata n.f. As var. incerta but colonies more greatly 
elongated, i.e. to about three or four times their length, thus ap¬ 
proaching M. stagnalis, although of course still very much less 
elongated than the latter. 
Loc. With var. incerta. 
RELATIONSHIP OF CCELOSPHAERIUM DUBIUM Grun. 
(see (5)) TO MICROCYSTIS 
This species, which also occurs in the freshwater plankton of 
Ceylon, shows relationship with certain members of the genus Micro¬ 
cystis in the following features: (a) the presence of pseudo vacuoles in 
the cell, ( b ) the not infrequent grouping of the colonies in clusters, 
and ( c ) the size of the cells which averaged 6 /x as in M. flos-aquce . 
As the latter species was associated with C. dubium in the lake at 
Candy and tank Tissawewa and as this association has been noted 
by Lemmermann, a genetic connection would at first sight seem 
probable. But the older colonies of C. dubium always show very 
clearly a single layer of cells and in these, as in some of the younger 
