Carter . — Studies on the Chloroplasts of Desmids. I. 237 
The latter species has a central axis occupying the centre of the 
semi-cell, and from this about eight plates radiate towards the periphery. 
As these approach the cell-wall, they begin to thicken rapidly, forking 
slightly so as to embrace a larger proportion of the cell-wall (Fig. 81), 
From the exterior the edges of these plates are seen as parietal bands 
running longitudinally (Fig. 79). The edges of the parietal bands are often 
quite straight, but they may sometimes be fringed with tiny projections or 
else undulate gently. 
The pyrenoids are confined to the central axis of the chloroplast, and 
are very variable in size and number. Sometimes there are only two or 
three long narrow pyrenoids occupying the whole length of a single chloro- 
plast, but more often there are about four to nine smaller spherical 
pyrenoids, either in a single median series, or when very numerous they 
may occur crowded together, two or more side by side in the distended axis 
of the chloroplasts ; cf. Fig. 80. The pyrenoids are often seen in an active 
state of division, the pyreno-crystal budding in various directions and 
being consequently very irregular in shape. 
In T. granulatus and T. laevis the axile chloroplasts are also provided 
with a central row of pyrenoids, which number about three to five in each 
chloroplast of the larger species, and one or two in T. laevis. In both cases 
the axis is provided with a number of radiating plates, but in neither species 
is there any attempt at the formation of definite thick parietal strands as in 
T. Brebissonii. T. granulatus has about eight to ten radiating plates, and 
in T. laevis there are about eight, but, in the former species particularly, 
these are often so complicated that it is scarcely possible to count them even 
in sections ; cf. Fig. 77. 
The edges of the plates are usually deeply cut, forming long attenuated 
outgrowths which stretch towards the periphery, some projecting on one 
side of the ridge, some on the other, and when this happens the individual 
identity of each ridge becomes very obscure (Fig. 76). In other specimens 
the toothing of the edges is not so deep, and the plates are much more dis- 
tinct (Fig. 78). The extent to which the ridges are toothed greatly influences 
the appearance of the individual. Where definite plates occur these are 
sometimes seen in the front view of the specimen to undulate in a manner 
similar to that occurring in certain species of Closterium , adjacent ridges 
alternately touching and becoming distant. 
VI. The Chloroplasts of the Genus Euastrum. 
The larger species of this genus are so very densely green in the living 
condition that it is impossible to get any idea of the form of their chloro- 
plasts, even after very careful study. For this reason they have never been 
accurately described or figured. 
The earliest investigators did not usually attempt to illustrate the 
