50 Williamson on Volvox glohator. 
see the true cells ; the protoplasmic membrane being in such 
close apposition to its granular and mucilaginous contents as 
to prevent its being identified as a separate tissue. But 
when the specimens have been mounted some time, we oftrn 
find that a change takes place. The protoplasmic matter 
shrinks up into a small irregular mass (7 h\ and thus becomes 
detached from the protoplasmic membrane (7 c, c') which forms 
a ring round it. When we succeed in compressing the object 
so as to force the cells into an oblique position (as is done in 
those to the left of fig. 7), we see that these circles are really iden- 
tical with figs. 14 and 15 c, c' in the sections. The external 
cells (7 a) remain in mutual contact, excepting at their angles. 
There is a little discrepancy between this description and 
that of fig. 1 1 in my former memoir, and the explanation of 
its cause will do much to diminish the real difference between 
myself and Mr. Busk, whilst it tends to confirm my ideas 
respecting the cellular structure of Volvox. In many of my 
mounted specimens, the outer or ceZZ-membranes have either 
failed to become visible, or have disappeared again. On the 
other hand, in these examples, the protoplasmic membranes 
have separated from their protoplasms and become very con- 
spicuous. I formerly confounded the two, and imagined that in 
the latter examples the incipient separation of the cells at the 
angles (seen in figs. 11 and 16) had been subsequently carried 
much further, causing a complete isolation of the cells, as is 
apparently the case in fig. 17. My error was corrected by the 
specimen delineated in fig. 7, in one part of which hoih these 
tissues are seen as there represented, thus enabling me to iden- 
tify the inner protoplasmic membranes (7 c) of the one with the 
only membranes seen (17 c) in the other. Numerous similar 
specimens have since confirmed the correctness of this explana- 
tion, which clears up many obscure points. I now find no 
difficulty in recognising the two structures ; the protoplasmic 
membrane, whether seen in front or in profile, is always more 
granular, from the adhesion to its inner surface of some 
of the granular elements of the protoplasm, than is the case 
with the cell-membrane, the outlines of which are invariably 
clear and fine. Fig. 8 represents three cells from the same 
specimen as fig. 7, in which the protoplasmic membranes 
nearly fill the respective cells ; such specimens, seen in 
section, exhibit the appearance of the centre cell of fig. 15. 
I conclude that if we could bring all these structures into 
view in a section of a living Volvox, they would present the 
appearance of fig. 12, where a represents the cell-membranes, 
b the protoplasm and its contiguous membranes, d the 
common pellicle, and e the prolonged threads of the pro to- 
