ARE THE DESMIDS AND DIATOMS SIMPLE CELLS? 139 
Desmid the endochrome of the parent cell constitutes a con- 
tinuous mass till division takes place, this being the only 
condition under which the protoplasmic material for the two 
new valves about to be formed could be exuded from* the divided 
surfaces; for, obviously, no exudation could take place were 
these surfaces already sealed by the closing’ in of the mem- 
branous investment. But whereas in the Desmid the develop- 
ment of the cellulose covering* proceeds simultaneously with the 
exudation of the young' segment, becoming; thicker and firmer as 
IflGi t. 
C . D 
A. A frond of a Desmid, namely Cosmarium, iu which division is about one-third 
completed, p s. p s, the two parent segments normally separated by a deep 
notch ; n s, n s, the two yoitng segments being budded forth from the narrow 
but open passage i i, by which the parent segments were held together 
like Siamese twins. 
B. A segment of a Closterium, in which genus the segments part company at a 
very early period ; n s being the incipient young segment, which is being budded 
forth as in the last case. 
C. A frustule of a Diatom, Biddu/phia, about half-way advanced in the process of 
division — p v, p v, the- parent valves ; n n, the endochrome masses being 
budded forth, precisely as in the Desmid, from the open surface of the divided 
cell contents ; c z, c z, the two connecting zones sliding apart, one within the 
other, telescope fashion', to accommodate themselves to the increasing bulk of 
the cell contents during division. 
D. The same, division being completed, and the two perfect and consolidated 
siliceous valves, n v, n v, shown within tb,e sti 1 persistent connecting zones 
which had protected them during their development. 
the process advances ; in the Diatom the secretion of the siliceous 
wall does not commence at all until the soft plastic protoplasm 
of which the new valve is to be formed, has already attained 
its full proportions and figure. This fact admits of a ready 
explanation, inasmuch as the unyielding nature of the siliceous 
coat, however thinly it may be deposited, would effectually bar 
all further development of the soft parts, were the siliceous 
deposit to begin at any stage anterior to that above described. 
(^See fig. 4, in which the process of development of the new 
