24 SPONGES 
injured portion. Unquestionably in this case, and also in the case of the 
young sponge cited above, an attempt has been made to repair the injury to 
. the cylinder. This injury could not be remedied by a new growth from the brok~- 
en portion, for asI have demonstrated under the head of Skeleton, no growth 
of either sarcode orof horny matter ever takes place from a broken surface. 
Hence it is that the membrane is first thrown across above the injured portion 
as a preliminary to the growth of the sarcotde and accompanying skeleton, thus 
accomplishing a permanent repair of damages. The specimen in Fig. 11, B. 
shows an example of where a permanent partition has been formed above a 
broken cylinder. In this case thecavity, d, has been filled with sand and 
the partition a, has been formed above it, completely across the tube. So sol- 
idly was this partition constructed, that it formed’ anew base for the sponge 
and, as seen, a bud grew frem above it. 
I have found one other remarkable variation in the formation of this mem- 
brane in a tube, where, by some accident, the whole sponge was pressed fiat 
when living, probably temporarily closing the terminal orifice of the central 
tube. Then there wasa new orifiee formed on one side ; but the pressure after- 
wards being removed, and the original terminal orifice reopening, that in the 
side was closed by a membrane. | 
Sometimes, more especially in large single cylinders, the tendency to grow 
outward at the extremity isso great that the tube gradually widens, and be- 
comes trumpet-shaped. Such cylinders are inclined to bend downward, and I 
have found quite a number of these specimens among the sponges gathered at 
Inagua. These sponges being simple cylinders, with thin walls, have evident- 
ly thrown all of their vigor into an upward growth, but all stages of intergrades: 
eccur between this form and the types. 
