146 Fritsch.—The Morphology and Ecology of an Extreme 
towards the pigment-cells, whilst their distal ends were rounded or almost 
truncate (Fig. 2, A), and it will be noticed that these akinetes were provided 
with an especially thick and stratified membrane. 
Since this mode of akinete formation was observed mainly in winter, 
it is possible that the extrusion of purple sap into the pigment-cell, by 
reducing the water-content of the protoplast of the akinete, renders it more 
resistant to frosts, to which in the exposed habitat it must be very liable. 
This explanation is, however, only advanced tentatively, since the pheno- 
menon may be due to other causes, but attention should be drawn to the 
fact that it was observed in the bulk 
of the filaments over a considerable 
area. On a much smaller scale it 
has also been encountered at other 
times of the year, but only with re- 
ference to occasional akinetes. 
As above indicated, the division 
leading to the production of akinete 
and pigment-cell is an unequal one 
and due to the development of 
a septum, generally of the asym- 
metrical type (p. 14 1), towards one 
end of the cell. In fact, the asym- 
metry of the septum is responsible for 
the usual pointed shape of the akinete 
at the end adjacent to the pigment- 
cell. When renewed growth takes 
place, the contents of the latter 
gradually disintegrate and disappear, 
so that sooner or later the pigment- 
cell is quite empty (Fig. 2, g). 
These empty cells constitute 
so many weak points in the filaments, 
at which rupture readily takes place, 
and therefore this mode of akinete 
formation probably brings with it the advantages of extensive vegetative 
propagation (cf. Fig. 2, G). As a matter of fact, many of the filaments of 
the Hindhead form terminate in short empty fragments of cell-membrane 
appearing just like the ruptured remnants of pigment-cells. In some cases, 
however, the adjacent akinetes, on resuming growth, protrude from either 
side into the empty pigment-cell, so that the cavity of the latter becomes 
more or less obliterated, and ultimately appears only as a narrow slit in the 
thick membrane between the two (Fig. 2, c). 
The dying away of occasional cells of a filament is not confined to 
Fig. 3. a, swollen akinete of terrestrial form 
from Wales, b-e, terrestrial form from Hindhead. 
B, rhizoid-formation. C-E, cells from microtomed 
material of the Hindhead form ; c-E in longi- 
tudinal section ; c' in transverse section, c, a 
mature cell. I) and e, stages in division of the 
chloroplast. The nucleus is shown black. (A and 
B, X400; c-E, X650.) 
