228 Carter . — Studies on the Chloroplasts of Desmids. I. 
the pyreno-crystal, and it is only very rarely that there is only one layer of 
starch grains round each pyrenoid. In most cases the starch layer is of 
some thickness, and consists of very numerous starch grains closely packed 
together, either in a compact even layer or projecting irregularly in all 
directions into the reticulum of the chromatophore (Figs. 14, 16, 17, 18, 
and 19). 
In some cases the pyreno-crystal is observed to fragment, forming 
several shorter, or very many small spherical or irregularly shaped globules 
in a longitudinal series embedded in the central cylinder of starch-grains 
(Fig. 17). The size of the small pieces of pyreno-crystal varies very much ; 
they are often very irregular in shape and are not necessarily arranged in 
a single series in the middle of the chloroplast, two or more often occurring 
side by side. 
Such complex pyrenoids have only been observed, outside the genus 
Netrium , in the chloroplasts of Ciosteriam Libellula, Focke, Penium spiro- 
striolatum , Barker (Fig. 20), P. polymorphum , Perty, P. mar gar i taceum , 
(Ehrenb.) Breb., and Cylindrocystis Brebissonii , Menegh. 
In Netrium interruptum one specimen was observed to have only one 
chloroplast in each half-cell, with a single pyrenoid stretching throughout 
its whole length (Fig. 19). There were, however, signs of the constriction 
of the chloroplast about half-way between the nucleus and the apex of the 
cell. This suggests that for a time after cell-division each individual has 
only one chloroplast with one pyrenoid, and that the division of this to 
produce the two usually present does not occur until some time afterwards. 
IV. The Chloroplasts of the Genus Closterium. 
In 1910 Lutman published a paper describing the chloroplasts of 
CL Ehrenbergii and Cl. moniliferum , figuring transverse sections of both 
species, and pointing out how the structure of the chloroplast in the genus 
Closterium , as he found it in these two species, differs from that figured by 
Nageli ( 1849 ) for Cl. parvidum and CL moniliferum . According to Lutman, 
the chloroplast of Closterium in transverse section is like a coarsely cogged 
wheel, whereas according to Nageli’s figures it is like a hub with radiating 
spokes. Lutman asserted that Nageli’s optical sections give quite an 
erroneous conception of the true state of affairs, the chloroplast in Closterium 
consisting of a curved cone-shaped structure with relatively low ridges on 
its surface, and not of a series of radiating plates arranged round a slender 
central core. 
With the object of ascertaining whether the examination of other 
species of Closterium would support Lutman’s statement, the chloroplasts 
of the following species have been investigated, both by means of whole 
mounts and of sections: — Cl. Lunula, (Mull.) Nitzsch., Cl. lanceolatum , Kiitz., 
