242 Carter . — Studies on the Chloroplasts of Desmids. I. 
the front view, and it would be very easy to confuse such a chloroplast with 
the really parietal form which is occasionally met with in this genus. 
In order to determine whether the chloroplast is axile or not, it is necessary 
to examine it from the side or end. The central axile plate will then 
be viewed along its edge, and will appear as a sharp line if it be present. 
Such broad central axes occur occasionally in other species of the genus 
(cf. Figs. 82, 93, and 94), but it is the rule rather than the exception 
in Eu. Didelta . 
In most specimens, pyrenoids are rarely found in the central axis, 
but there may be one or two small ones, especially when this part of 
the chloroplast is more massive than usual. If there is a twist in the axis, 
a tiny pyrenoid is also commonly found at the point of twisting.. Eu. 
sinuosum , Eit. ampidlaceum , and Eu. affine , however, often have pyrenoids 
even of the ordinary size in their central axis, especially in well-nourished 
specimens with large chloroplasts. 
In transverse sections the radiating plates are seen connecting up the 
parietal plates with the central axis, and their length naturally depends on 
the shape and disposition of the latter. They may be long and distinct 
(Figs. 73, 99, and ioi), or in rare cases they may appear to be entirely 
absent, the parietal part of the chloroplast apparently arising directly from 
the central axis; cf. Eu. Didelta (Fig. 65). Towards the cell-wall these 
radiating plates thicken out, and finally spread out over the wall to form the 
parietal plates (Figs. 83, 99, and 101). 
The latter form a thick layer lining nearly the whole of the cell-wall, 
and there are four of them in each semi-cell, two side by side on each front 
face (Figs. 62, 63, 72, 82, 93, 94, &c.). The general shape of the parietal 
plate necessarily depends on the outline of the cell. Its one edge approxi- 
mately follows the lateral margin of the semi-cell, and the opposite edge 
closely approaches the median line. The plates are sometimes coarsely 
lobed, and these lobes are further subdivided, forming a fringe of pro- 
jections, the shape of which varies according to the species. In Eu. 
oblongum they are rounded or finger-like (Fig. 72) ; in Eu. crassum they are 
more jagged (Fig. 100). The lobing, particularly of those edges near the 
median line, is frequently very deep, and as a consequence the plates are 
sometimes nearly cut horizontally into two or three separate parts (Figs. 72. 
and 100). 
The external surface of the parietal plates is rare 1 / quite smooth, but is 
usually covered with tiny outgrowths extending towards the cell-wall. The 
size and shape of these projections vary with the species, but they are 
usually very similar to the teeth or outgrowths round the edges of the 
parietal plates of the same species. In Eu. crasstim they are frequently 
very large and can be seen quite easily in living specimens (Fig. ico). In 
the smaller species they are not so evident, but they can usually be 
