DoNKiN, on Marine Diatomace<e. 
27 
which a filament of its fmstules exhibits. The following are 
the facts which I have ascertained regarding this most re- 
markable phenomenon : 
1 . When the filament is in a quiescent state, the frustnles 
are all drawn up side by side^ their extremities being all in 
a line_, thus forming a group. 
2. When a filament, previously at rest, resumes its activity, 
the movement is commenced by the second or i7iner frustule^ 
at one end of the filament, gliding forward along the con- 
tiguous surface of the first or outer frustule, until their oppo- 
site extremities overlap each other. This is soon followed 
by a similar movement of the 3d, 4th, and 5th, &c., all 
moving forward in the same direction, and each frustule 
gliding along the surface of the one preceding it, until they 
have extended themselves into a lengthened filament or chain. 
In the course of two or three seconds after this has been 
accomplished, a retrograde movement, exactly of the same 
character, begins to take place, and continues until the fila- 
ment has retraced its course, and stretched itself out in a 
direction exactly opposite to the position it had previously 
occupied. This phenomenon is repeated again and again, 
and in this manner the whole group is kept in a state of 
activity for an indefinite period of time, and all the while, if 
no impediment produces irregularity, the outer or terminal 
frustule, next to which the movement commenced, maintains 
a stationary and fixed position. 
3. The rapidity with which each individual frustule moves 
is in direct ratio to its distance from the terminal stationary 
frustule, being most rapid at the opposite or moving extre- 
mity of the filament. On this account, most of the frustules, 
while the filament is moving to and fro, cross a line drawn 
at right angles to the middle of the long axis of the sta- 
tionary frustule, at the same instant of time, afterwards 
shooting past each other like horses on a racecourse. 
4. The force with which the filament moves is very great, 
so much SO; that I have observed it upset and shove aside a 
large frustule of A. arenaria, n. sp., at least six times its own 
bulk, obstructing its path. This force is, in a great measure, 
due to the rapidity with which the frustules move. The 
time which a filament, even of considerable length, occupies 
in crossing the field of the microscope being only a few 
seconds. 
5. Light appears to be a necessary stimulus for the main- 
tenance of this motion. When a filament, in active motion, 
is placed in the dark for a short period, and then examined, 
the movement is seen to have ceased ; but again commences 
VOL. VI. e 
