47 2 Smith . — Cytological Studies in the Protococcales. II. 
walls of the marginal cells are colourless. I have observed this red colora- 
tion of the walls of the interior cells in living material when using the lower 
magnifications (500-750 diameters), but have never found it with the higher 
ones (1,500-2,000 diameters) ; I consider this red coloration therefore an 
optical illusion. 
In preparations stained with Flemming’s triple stain the walls appear to 
be composed of two parts : a very thin outer layer which takes the gentian 
violet and a thick orange-staining portion (Fig. 8). The portion stained by 
the gentian violet is not always of uniform thickness, but is likely to be much 
wider at the angles of the cells. The inner orange-staining layer does not 
show unless the preparations are very heavily counterstained with the 
orange G, so that in most preparations this inner part does not stain, 
but appears as a clear space between the violet-stained portion of the wall 
and the cytoplasm (Figs. 4—6). At first glance, therefore, many cells appear 
to be slightly plasmolysed when they are really turgid. 
Petersen ( 10 ) has recently found in various species of P ediastrum , 
including P. Boryanum , that there are long tufts of bristles protruding from 
each cell in the colony. In a few instances he was able to see these 
bristles in living material, but by means of special staining methods he 
demonstrates the presence of these bristles in several different species where 
they cannot be found by examining the living material. I have tried 
Petersen’s staining methods on the colonies of my cultures with negative 
results. His material, however, came directly from the plankton, and it is 
quite possible that these bristles are not developed except under plankton 
conditions. 
In my cultures I have observed a certain diurnal periodicity in swarm- 
spore liberation. The alga was observed for the entire twenty-four hours of 
the day, and with very rare exceptions swarming occurred only at day- 
break. Swarming colonies first appear about an hour before sunrise, but the 
maximum amount takes place just at sunrise. These observations are not 
in accordance with those of Braun ( 2 ), who found swarming only in the late 
afternoon. The precise hour of swarming varies from month to month, the 
maximum occurring at 5.30 a.m. in May emd at 7.30 a.m in December. 
That the induction of swarming is very largely dependent upon daylight 
can be demonstrated readily by placing a portion of a culture in a dark 
room before the first signs of daybreak (2-3 a.m.) and leaving the remainder 
of the culture where it will be illuminated by the rising sun. At the hour 
of sunrise the illuminated culture shows active swarming, while the un- 
illuminated one shows none ; but if the latter culture be brought out into 
the light after the illuminated culture has ceased swarming, it too will show 
a large number of swarming colonies. This relationship between the 
swarming period and the illumination was first observed by Braun ( 2 ) in 
Hydrodictyon . 
> 
