SPONGES oS 
be contrasted with the more sunken membrane of its somewhat smaller neigh- 
bor at A. Just what the function of these membranes is, it is difficult to say. 
unless it is to prevent the accumulation of sand and other debris, in the central 
tubes. As confirming this hypothesis, it may be remarked that the central 
tubes of all the Verongia that I have examined, are remarkably free of foreign 
material of any kind. 
When a cylinder becomes broken off anywhere, a membrane forms at its 
broken termination, but it is questionable whether the sponge ever increases in 
length again. (See remarks on this under Variation of the Skeleton. ) 
In one specimen which has been attacked by a parasitical sponge, and its 
structure greatly disturbed thereby, a partition has grown up within the main 
tube, near the termination, supported by the horny skeleton, thus forming a 
kind of double tube, each division of which is provided with its own membrane. 
On the side of the cylinder not far from its termination are two oscules, each 
with a membrane that stretches nearly across it, while other oscules are form- 
ing about it. This seems to me a clear case of reversion toward some ancestral 
branching form, brought about by the disturbance produced by the parasitical 
intruder ; but to this subject I shall refer again. 
An extiaordinary case of membrane formation may be scen in the young 
sponge given in Fig. 18, where the terminal end is completely closed by a mem- 
brane ( see ib., A, a, ) and the lower or basal extremity being broken, exposing 
the inner tube, a memorane is partly formed across the exposed portion, which 
us the normal opening was closed, became the mouth. (See ib. B, b, being 
the supplementary closing membrane. } 
In two or three other instances where I have seen holes broken in a cylin- 
der, I have also found a membrane stretched partly across the tube above the 
