1898] THE MOVEMENT OF DIATOMS 409 
they remain separate give rise to the canals. Consequently, each of 
these intermediate pieces forms a niche-like depression between two 
projecting transverse canals and a corresponding piece of the 
longitudinal canal. Lobular processes of the chromatophores pro- 
ject into the transverse canals and are always enveloped by proto- 
plasm, which passes from their extremities towards the outer wall 
as a strand, usually undivided. These strands exhibit both longi- 
tudinal and transverse fibrillation during life, and, as a rule, appear 
to consist of five or six longitudinal rows of cellular compartments. 
The longitudinal canal was described by Flogel as closed exter- 
nally, but Lauterborn’s sections show that it is interrupted along its 
outer edge by a very narrow cleft, thus placing the interior of the 
cell in direct communication with the surrounding medium (Fig. 4). 
This cleft is not only visible in sections through the frustule, but 
can also be made out in a surface view when one of the alae is 
viewed vertically, parallel to the direction of the transverse canals. 
Yet another circumstance, not taken into account by Flogel, points 
to a breach of continuity in the cell-wall at this spot. In the living 
Surrirella, foreign bodies (eg. particles of Indian ink, sand, small 
diatoms,) are readily observable adhering along the edges of the 
alae, where they are moved briskly to and fro in the same manner as 
along the raphe of Pinnularia, Navicula, etc., a circumstance very 
difficult to explain if the cell-wall were in reality unbroken. 
Most of the Swrrirella were infested externally by a small 
alga belonging to the Cyanophyceae (Chroococcus sp.). The small 
blue-green spherical cells always occurred in a definite situa- 
tion, viz., in the niche-like intermediate pieces of the alae, only a 
single alga being present, as a rule, in each cavity (Fig. 5 pa.). 
More than once, examples of Swrrirella were noticed where all the 
intermediate pieces of the alae sheltered these little lodgers. It is 
pointed out that the advantage may well lie exclusively on the side 
of the Chroococcus, a possible explanation being that the alga, seated 
on the diatom in the very fine and easily disturbed mud, suffers less 
interruption of the assimilative processes than if it were free in the 
mud, because the diatoms, if buried by the movements of fishes or 
creeping molluscs, soon work their way up again to the surface 
and so to the light. In this connection it may be mentioned that 
Gastrotricha, of the genera Chaetonotus and Ichthydiuwm, constantly 
attach their eggs to the surface of large Surrirellae. 
The Movement of Diatoms.——Various hypotheses have been 
advanced to account for the characteristic movements more or less 
familiar to every student of the Diatomaceae. Amongst earlier in- 
vestigators, M. Schultze held that protoplasm, protruded through the 
slit-like raphe, served to set the cell in motion, and this view was 
maintained by Pfitzer and Engelmann. On the other hand Nageli, 
2 
