Smith . — Cy to logical Studies in the Frotococcales. //. 471 
portion of the pyrenoid stained and the other did not (Figs. 3, 6). These 
may possibly represent stages in a metamorphosis of the body of the 
pyrenoid into starch similar to that described by Timberlake (16) ; but 
since the unstained portions are always a part of the pyrenoid and never a 
segment that has been cut off, I am inclined to consider them as due 
to irregularities in staining and not as stages in starch formation. The 
starch plates which surround the pyrenoid are curved, and usually three or 
four are visible when it is viewed in median optical section. The space 
always perceptible between the starch plates and the pyrenoid appears 
empty. This space is not always of uniform width, as I have found in 
Characium (15) and Scenedesmus (13), but may be much wider at one side than 
another (Figs. 5, 10, 22). There is also considerable stroma starch scattered 
throughout the peripheral region of the cell, especially in colonies from older 
cultures in which it occurs in abundance. In preparations destained 
sufficiently to show the nuclei well, the ‘stroma’ starch plates are fre- 
quently so decolorised that the cytoplasm appears to consist of fine angular 
vacuoles (Fig. 4) ; but proper staining demonstrates that what seem to 
be vacuoles are really ‘ stroma ’ starch plates (Fig. 7). This accumulation 
of starch plates in the cytoplasm is not peculiar to Pediastrum , but is 
a constant characteristic of old cultures in all the Protococcales that 
I have had under cultivation. The accumulation of starch is accompanied 
by a loss of the chlorophyll in the plant, so that the general appearance of 
the culture is not a bright fresh green but a sickly yellow-green. The same 
loss of colour and accumulation of starch may be seen in the cells of 
Oedogonium , Spirogyra, and other filamentous forms collected in the field 
during autumn. Cells containing these excessive amounts of ‘stroma’ 
starch are not normal, as is shown by the fact that such cells in the fila- 
mentous species fail to divide, and similarly in cultures of Protococcales, 
where the cells contain large amounts of starch, this same pathological 
condition is evidenced by the small number of young, recently formed 
colonies. 
Askenasy (1) calls attention to the fact that there is a definite relation- 
ship between the size of the pyrenoid and the size of the cell, since increase 
in size of the cell is accompanied by an increase in size of the pyrenoid. 
I have observed this relationship in both the cells with the single pyrenoids 
and in those with more than one pyrenoid. In the latter there is the growth 
of the pyrenoids as the cell grows, but the rate of growth is much slower 
than in cells with but a single pyrenoid. 
The colonies of Pediastrum are not invested with a gelatinous sheath 
as are those of many Protococcales. This fact can be demonstrated by 
using Schroder’s (11) method of staining living material with vesuvin, 
or by Errera’s Indian ink method (4). Various systematic works figure and 
describe the walls between adjacent cells as being red, while the outside 
K k 
