136 
POPULAR SCIENCE REVIEW. 
masses of coloured endochrome within the cellulose wall. But 
so long as retained within it, neither did the globular masses of 
protoplasm coalesce with the darker masses of endochrome, 
even when so forcibly squeezed together as to flatten each other 
to a certain extent ; nor did these two portions of the contents 
mix in the least degree with the water that had gained ingress 
to the interior of the chamber. Up to this stage, therefore, 
there seemed just as much reason for assuming that both por- 
tions had become, as it were, instantaneously invested with a 
firm pellicle, as that one or both had, previously to the rupture 
of the cellulose wall, possessed some such investiture. But the 
moment the two portions escaped through the ruptured wall, it 
was only the colourless protoplasmic globules that still retained 
their perfectly defined outline. The endochrome mass which 
protruded showed perceptibly (by the irregularity of the freed 
margin, which could be seen amalgamating slowly with such 
portions of the globules as happened to be in immediate contact 
with it) the point at which its investing membrane also had 
become ruptured. Lastly, on the pressure being increased, 
simultaneously with the further escape of the contents not only 
did the cellulose wall become folded and puckered, but portions 
of the yet retained investing membrane of the endochrome 
could be distinctly seen within it in a similar condition. 
It appears evident, therefore, that the colourless protoplasm 
exists independently of any special investing wall of its own, 
and behaves in precisely the same way as sarcode ; whereas the 
true endochrome is encased in an imperforate membranous 
covering of sufficient strength to resist a considerable degree of 
pressure and distension before it yields and becomes ruptured ; 
and, being more or less granular, is not so viscid. 
I shall now proceed to show how far the facts already adduced 
in reference to the cell formation in the Desmidiacese can be 
said to furnish a key to that which is to be met with in the 
most nearly allied family, namely, the Diatomaceae. 
The structure of these organisms is undoubtedly more com- 
plex in some respects than that of the Desmidiacese, notwith- 
standing that their general physiological characters, the manner 
in which the processes of multiplication by “ binary division,” 
and of reproduction by the fusion of the protoplasmic contents 
of two or four cells, the formation of a 66 sporangium,” and the 
evolution from this body of the germs of an entirely new 
generation, are virtually identical in the two families.* 
Thus (speaking generally) we find the endochrome in both 
* See a paper on u The Relation between the Development, Reproduction, 
and Markings of the Diatomacese,” by the author, “ Monthly Microscopic 
Journal,” Feb. 1877, p. 77. 
