2 2 6 Carter. — Studies on the Chtor op lasts of Desmids . /. 
the other had a group of small ones (see Fig. i ii). In view of these facts 
it seems very unlikely that there is any definite relation between the number 
of pyrenoids contained in a cell and its size. 
The method of division observed in the pyrenoids was nearly always 
by constriction or pulling apart of one individual to form two. More rarely 
the pyrenoids divide by a clean cut into two parts, or the pyreno-crystal 
fragments to form many. In many cases when actively dividing the pyreno- 
crystal was observed in the process of budding off two or more new 
pyrenoids at the same time (Figs. 59, 60, and 80). 
Whilst examining the chloroplasts of Cosmarium ochthodes , Nordst., 
numerous darkly staining globules were noticed, sometimes in the peripheral 
parts of the radiating plates, but more often in the reticulated films of 
chloroplast stretching from these over the interior of the cell-wall. In 
stained preparations they have the general appearance of small naked 
pyrenoids, staining in exactly the same way as the pyreno-crystal of the 
larger pyrenoids of the cell. They do not appear to have a definite starch- 
sheath, but are always surrounded by a lighter space. In sections stained 
with iodine the globules become distinctly brown, and when tested with 
other reagents they always give the same reaction as the pyreno-crystals of 
the ordinary pyrenoids. They stain brilliantly with acid fuchsin and are 
undoubtedly of a proteid nature. 
Similar globules have been observed in several other species of Cos- 
marium and Closterimn (Fig. 44), in one species of Cylindrocystis , and they 
have also been found scattered amongst the larger pyrenoids in the chloro- 
plasts of Micrasterias denticidcita and Xanthidinm armatum . When tested 
in several of these other species they were found to behave in exactly the 
same way as before. 
Their position in those parts of the chloroplast nearest the cell-wall is 
a constant feature, and is very marked in every case in which they were 
observed with the exception of Micrasterias denticidata , in which species 
this distinction is scarcely possible. They vary considerably in number and 
are not found in every individual of such species in which they have been 
observed. 
They seem to have no relations with the ordinary large pyrenoids of 
the cell, and are probably not derived from them in any way, being most 
likely formed de novo when the conditions in the cell are favourable for 
the storage of large quantities of reserve food. In all the species of 
Cosmarium in which these small globules were observed, the points at which 
the ordinary pyrenoids occur are definite and fixed, usually in the interior 
of the cell, and even if these do divide the products of their division remain 
in the position of the original ones, whilst these small globules are only to 
be found near the cell-wall. Thus the possibility of their origin from the 
division of the ordinary large pyrenoids can be neglected. 
