234 Carter. — Studies on the Chloroplasts of Desmids. I. 
Cl. Lunula . 
The ridges in this species are about fourteen to seventeen in number, 
and the axis in transverse section is usually broad, but the relative size of 
the axis and ridges is liable to considerable variation; cf. Figs. 1-4. 
Sometimes the ridges are so low that they scarcely exist at all in the broader 
part of the cell ; towards the apices they are always more evident (Figs. 1 
and 3). When the axis is very broad the pyrenoids are scattered in a single 
layer in its peripheral parts (Figs. 1 and 2). Other specimens may have 
very distinct ridges, which in transverse section are seen to be quite thin 
and plate-like, and consequently the axis in these cases is much more 
slender than before, the pyrenoids being necessarily crowded together in the 
more central region of the ceil (Figs. 3 and 4). This variation in structure 
is quite evident even in the living condition. I11 certain individuals the 
cell-contents are much more densely green than in others. Such specimens 
have very distinct ridges, and the pyrenoids are crowded together in the 
middle of the cell, forming a large refractive mass. Other specimens are 
much paler in colour, and the cytoplasm is very coarsely reticulated, the 
reticulation being sometimes visible even with the low power. Here the 
pyrenoids are scattered in a more or less superficial layer, and the ridges are 
very low and often scarcely perceptible. Both these types of structure and 
various intermediate stages are to be found in a single collection. The dense 
green colour of the first type is probably due to the much closer reticulation 
of the protoplasmic foundation of the chloroplast. 
In individuals having rather extensive plate-like ridges these may 
undulate very strongly (Fig. 3). Sometimes undulation only occurs in the 
apical end of the chloroplast when the ridges in the middle of the cell are 
too low to allow of such undulation. In transverse section, however, no 
branching of the ridges is to be observed, except occasionally in sections 
taken from the apical end of the cell, which may show just a suggestion of 
branching (Fig. 5). On the whole, however, the undulation in this case is 
merely the result of the ridges travelling in slightly waving paths, and not 
due to actual anastomosis, or fusion of the tissue of one ridge with that of 
other ridges. 
The degree to which the edges of the ridges are lobed naturally depends 
to some extent qn the size of the ridges themselves. In cells having very 
low ridges, the lobing of the edges may be practically absent except at the 
apices of the cell, where the ridges are always larger (Fig. 1). In individuals 
having more definite plate-like ridges the edges are usually distinctly lobed 
throughout their whole length, though the depth to which they are hollowed 
out varies (Fig. 3). Occasionally in specimens having a very slender axis 
with crowded pyrenoids it may be said that ridges scarcely exist, since they 
are represented merely by innumerable long finger-like outgrowths radiating 
