40 
Wallich^ on Siliceous Organisms. 
vicinity of a free-moving diatom^ and which are unquestion- 
ably associated with it as cause and effect^ we have no alter- 
native but to admit the existence^ although as yet undetected^ 
of highly attenuated prehensile filaments_, in virtue of which 
the diatom itself moves, and produces the movements we ob- 
serve in surrounding particles."^ The grounds upon which I 
rest my view are wholly based on the behaviour of the diato- 
maceous frustule with reference to such particles. This may 
probably be looked upon as assuming more than is warranted 
without ocular proof of the existence of the organs alluded 
to. I can only reply that the assumption is supported by 
stronger and more tangible proof than the theory of endos- 
motic and exosmotic action ; which is, after all, nothing more 
than an assumption. In favour of the former view, there 
exists strong presumptive proof. In favour of the latter, not 
only is presumptive proof deficient, but such other evidence 
as can be brought to bear upon the question, tends at once to 
disprove its correctness. This I will endeavour, as briefly as 
possible, to demonstrate. 
In general terms, endosmotic and exosmotic action may be 
defined as the eflPort whereby two fluids, of different densities, 
and which happen to be separated from each other by an 
animal or vegetable membrane, slowly commingle, and acquire 
the mean between their two original densities. Such an action 
is constantly going on in all living animal and vegetable struc- 
tures, and, without doubt, in those we are more immediately 
discussing. The various processes of assimilation, absorption, 
and secretion are most powerfully influenced by it. But all 
I contend is, that this kind of action is not that whereby the 
peculiar movements of the Diatomacese are effected. 
Were these organisms bicellular or multicellular, instead 
of being unicellular, or were the motions invariably exercised 
in only one given direction of the frustule, such action might 
possibly be taken to account for them. But there is no con- 
dition of the frustular contents which can impart the alter- 
nating character we see to the movements ; unless, indeed, we 
admit the primordial utricle to be a bicellular structure, or 
that the terminal apertures of the frustule have the faculty 
* The movements of particles of matter, backwards and forwards, along the 
edges of the frustule, are evidently produced by different organs to those 
which influence the movements of particles towed along at a distance from 
it, inasmuch as they are carried on during the progress of the diatom, and 
often in a direction opposite to that it is pursuing. Judging from the 
manner in wiiich these marginal movements of frustules take place, it M'ould 
seem probable that the organs consist of an extremely minute series of 
extensile tubular suckers, similar to those of the Echinidae. 
