1910] LUTMAN—CELL STRUCTURE OF CLOSTERIUM 251 
by a series of transition stages to the violet color characteristic of 
the starch, but no such series of transition stages could be found in 
Closterium, although the segments cut off frequently appear as if 
stained a lighter red, owing to their smaller size. It is a question 
as to what becomes of these pieces of pyrenoids. Some figures 
seem to suggest that they round up and form smaller pyrenoids. 
It is very difficult to get any series of stages showing the migration 
of the parts away from each other, if such is the case. The hypothesis 
that they form other pyrenoids is supported by the fact that the starch 
layer is frequently of varying thickness, and may be quite thin on 
one side, as would happen if a segment had been cut off that side and 
the layer of starch had not been redeposited to any great degree 
(fig. 9). 
The great variation in size of the pyrenoids is a conspicuous 
fact. Scumurz, who first observed pyrenoid division, says distinctly 
that they may divide equally or very unequally. It is certainly true 
that they are capable of splitting up into segments, whether these 
segments become independent pyrenoids or develop into starch grains. 
In this connection the observation of BUTSCHLI that the paramylon 
bodies of Euglena split up into similar shaped disks is very interesting. 
He regards it as due to their internal structure. The similarity of 
his figures to those of the pyrenoids in Closterium may indicate that 
paramylon bodies and pyrenoids are alike in their essential structure, 
even though not in chemical composition. 
It should also be stated that these clefts in the body of the pyrenoid 
appear also in fixed and stained specimens that have not been sectioned 
_ but are examined whole. 
I have observed many of the typical figures of the division of one 
Pyrenoid to form two in material fixed at night. These figures are 
in all essentials like those observed by Scumirz in Hyalotheca and 
by a number of investigators since his day. It seems to consist 
simply in a pinching in two of the pyrenoid, and it is a curious fact 
that some of these same pyrenoids show the composite structure 
just described (fig. 26). 
n Hydrodictyon, according to TIMBERLAKE, the pyrenoids show 
no such numerous clefts as are present in Closterium, and this fact 
makes it still more difficult to believe that the segments in Closterium 
