42 BOTANICAL GAZETTE [JULY 
ments, showing a wonderful periodicity. The cell (usually a 
symmetric one such as Nitschia or Navicula) begins to swim 
with a gradually increasing velocity in some definite direction ; 
then the velocity diminishes, and finally the cell stops its move- 
ment to recommence it with the same rhythm in the opposite 
direction, or in a line differing by some small angle from its first 
path. 
These movements have thus all the characteristics of pen- 
dulum movements, 7. ¢., the periodic change of the direction (or 
of the sine, to speak mathematically), and of the velocity, going 
through the zero (the stopping of the cell) and attaining its 
maximum at the middle of the path. These movements, as 
known, have their origin in such a combination of forces, that 
the point towards which the acceleration is directed is placed 
at the middle between the two termini of the movement. 
Some observations suggested to me the supposition that this 
point, in microscopical observation, is determined by the apex 
of the cone of rays reflected from the mirror to the opening on 
the stage. 
If we consider the osmotic hypothesis as true, to which, as 
we have seen, there is no serious objection, there are two func- 
tions, each of which is most constant in the cell, which may be 
the source of those products that are the ground of osmotic 
currents between the cell and its surrounding. These are res- 
piration, which is continuous and the products of which are 
perpetually exosmosed by the cell, and assimilation, which 
takes place only in light, and which produces most probably 
some kind of sugar (as is known, the Diatomacee do not con- 
tain starch), that is, a substance of high osmotic activity and 
producing osmotic attractions to the interior of the cell.® 
Many times I have observed that the movements of the 
Diatomacee are very energetic when freshly collected and 
brought under the microscope in the fresh, cold water, espe 
® The osmotic tension in the cells of Pinnularia major and Surirella biseriata a8 
measured by plasmolysis is no less than 4 or § atmospheres (Cf. O. Miiller, ibid 7 : 169 
8 
1889). That shows how strong must be the predominance of endosmosis over €x05" 
mosis. 
