AEE THE DESMIDS AND DIATOMS SIMPLE CELLS? 
141 
unaltered in outline for a considerable period, would appear to 
have its own investing covering. But these masses, instead of 
occupying a position along the central region of the general 
chamber, are suspended, as it were, in the colourless protoplasm 
within, but not in immediate contact with the inner siu'face of 
the valve and connecting zone ; the middle of the chamber, 
together with the space intervening between the lamellae of 
endochrome, and the inner surface of the siliceous wall being 
occupied, just as in the Desmid, by colourless protoplasm. 
On cursorily looking at a Diatom frustule— -say of a Navicula — 
it might easily be supposed that the relative positions of the 
true endochrome and* the siliceous wall, both before and after 
division, differ from* those which obtain in the Desmid frond. 
Indeed, this mistake is a common one, due to the overlooked 
fact that, as a rule, division takes place in the Diatom in a plane 
which bisects the shortest axis of the organism ; whereas, in the 
Desmid, it takes place in a plane bisecting the longest axis. In 
other words, the tran verse or short axis of the Diatom frustule , 
coincides with a line passing across the centres of both its 
valves ; whereas the longitudinal axis of the Desmid frond cor- 
responds with a line passing through the centre of both its 
segments. Or, to state the case in still another way, the front 
view of the Diatom 'ucdve (see fig. 5, A, a) corresponds with the 
end view of the frond or segment of the Desmid (D, e v) ; whereas 
the front view of the Diatom frustule (A, 6, or C), and the front 
view of the Desmid frond (fig. 4, A, and fig. 5, D, d ), represent 
corresponding aspects in the two kinds of organisms. This last- 
named correspondence of parts will be best understood, however, 
by comparing the figures of Docidium (a Desmid) fig. 5,D, c/, 
with the front view of Melosira , a cylindrical but short Diatom, 
of which an outline representation is given at E. 
Lastly, I have to touch briefly on the subject of the extra- 
frustular appendages as these present themselves in the Dia- 
tomacese. They may consist merely of a gelatinous mass, more 
or less shapeless, in which a colony of frustules is imbedded, as 
in Dickieia; of a tubular sheath, of almost horny consistence, 
though perfectly hyaline,, in which the separate frustules move 
freely, as in Encyonema * T of stalk-like processes or pedicels 
which anchor the frustules to* the substance from which they 
spring, as in Cocconema , or which constitute a kind of stem to a 
broad, fan-like expansion of frustules, as in Licmophora ; of mere 
pads which connect the frustules of certain filamentous forms 
to each other, as in Buldulphia ; of a highly elastic and yet 
subtle film which completely invests some of the filamentous 
bacillarian forms, and at the same time either permits of their 
peculiar movements, or itself contributes to the performance of 
these movements, as in that most remarkable of all Diatoms, 
