A Critical Study of certain Unicellular Cyanophyceae 83 
variable and presumably depend to a large extent on external con¬ 
ditions. But in some Chroococcaceae the angular form of the cell 
cannot be explained so readily. Throughout the genus Tetrapedia the 
cells show various polyhedral shapes, in fact it is the remarkable 
cell-form that alone affords a satisfactory distinction from the allied 
Merismopedict. There is no evidence that the angular form is due to 
compression, since the individual cells are often free or very loosely 
associated. It is in fact an inherent quality of the cell and, under 
natural conditions at least, is retained from one generation to the 
next. Of course the angular character may have taken origin in 
relation to a former crowded type of colony. In some species of Meris- 
mopedia mutual pressure of the cells gives them an angular character 
just as in Chroococcus, except that they are quadrate in the former 
case, in correspondence with the mode of cell-division in Merismo- 
pedia. But in Tetrapedia there are other complications which are 
not explained on this view. For the cells are compressed in a plane 
at right angles to their division plane, and moreover they possess a 
very remarkable pair of constrictions dividing the cell at right angles 
into four parts. The first of these features is primarily due to the 
differential growth of the cell, which enlarges chiefly in directions at 
right angles to the directions of cell-division. The second feature is 
also related to the planes of fission. The position of the constrictions 
coincides with these planes, and in fact the partial segmentation of 
the cell-body in this way is little more than the precocious separation 
of daughter cells from the protoplast. The origin of the character 
may be traced in the genus Holopedium in which one species (H. 
Dieteli Richt.) differs from its allies inasmuch as the cells become 
more or less lengthened as seen from the surface of the colony before 
division and are recorded as being slightly constricted in the middle. 
The lack of this constriction in the other species and in the genus 
Merismopedia merely indicates that in these forms division is more 
direct. The constricted form of cell is also incipiently developed in 
Synechococcus, where, however, there is only one plane of division 
and consequently only two lobes. In 5 . ulcericola Rabenh. the hour¬ 
glass shape of the cells is a constant feature from the moment of their 
origin. 
A more common departure from the spherical cell-type is seen in 
those forms which have elongated cells. Where this is the case we 
may speak of a definite polarity, indicated by a preferential growth 
in one direction. The phenomenon is related to the process of cell- 
division, the axis of elongation being at right angles to the plane of 
6 —2 
