52 
Williamson on Volvox globator. 
protoplasms have become broken : those which remain have 
drawn their respective protoplasms towards the sides of the 
cells to which thej are attached. All these circumstances in- 
dicate a degree of ductility in the protoplasmic membrane 
such as would scarcely exist supposing it to consist of per- 
fectly organized cellulose. 
In the young gemmae, as already observed, the protoplasms 
are in close contact on all their sides ; but it is only at a few 
points that an actual junction is established corresponding 
with the extremities of the future threads. What has been 
the eclectic power leading to this result ? It is not mere acci- 
dent. Reference to fig. 1 1 will show, that in passing from one 
protoplasm to another these threads always traverse the sides 
of the hexagonal cells, and never their angles. It is also ob- 
vious that these points of adhesion are chosen prior to the de- 
velopment of the outer cell-membrane. This is indicated by the 
unvarying continuity of the threads when they are single ; but 
still more so when they are double and treble, as is frequently 
the case (figs. 10 and We, e'). Whatever the number pro- 
ceeding from a protoplasm to any one side of its cell, the same 
number proceeds to the proximate side of the adjoining cell : 
I have scarcely seen one exception to this rule. I think the 
explanation just given meets the case ; if so, it may be a ques- 
tion whether the cell-membrane is developed between the con- 
tiguous extremities of the two protoplasmic threads, or whether 
it is deficient there, admitting of an actual as well as an appa- 
rent continuity. I have already given one or two reasons for 
believing the former hypothesis ; but even should the .latter 
prove the true one, we shall only have recurring in Volvox a 
phenomenon that is common enough amongst the perforated 
cells and ducts of the higher plants. In but one instance have 
I seen a specimen countenancing the latter idea. In it the 
peripheral layer of cells was very thin and compressed ; many 
of the cells appeared to be wholly detached from each other, as 
represented in fig. 9 ; nevertheless the threads proceeded from 
protoplasm to protoplasm, apparently traversing the intercel- 
lular spaces. This specimen is so entirely exceptional as to leave 
little doubt on my mind that is capable of being explained. 
I have no doubt that, owing to the thinness of the peripheral 
cells, a section of it would resemble fig. 13. If we suppose that 
the circumstances which render the majority of Volvox cells 
hyaline and invisible, still continue to affect the portions ^f 
those in question that are external to the dotted line 3 3^*, 
the remainder being visible, we should have precisely such an 
appearance as is presented in fig. 9 ; the only visible por- 
