36 SPONGES 
tion beneath there had been no new fiber formed, but the sponges still rcmained 
attached by the small remaining portion of the base. 
It is most evident that a new growth can take place only where the inner 
membrane of the fiber still remains in direet contact with the sarcoie It is also 
evident, as will be seen, that the membrane itself, is capable of producing new 
growths by division only at its extreme tip, and always before it has beeome 
covered with horny matter. After this horny material is deposited, the mem- 
brane appears to become dead. ‘This I have proved conclusively by examin- 
ing severaltube sponges which have been broken, while living in the water, 
within an inch or two of the base. Phesc bases were living, had deposited sar- 
coileover the broken surface, ( thus completly covering the ends of the severed 
portion of the fibers ), and had in two or three eases even constructed new clos- 
ing membranes at the entrance of the central tube, but upon carefully measur- 
ing one of these broken cylinders and examining others, I find that there is 
absolutely no new growth from the broken portion of the skeleton ; ner are 
the tubes of the broken ends of the fibers even closed with horny matter. Of 
course, | have no means of judging when the sponges were broken, excepting 
from their appearance. These specimens were collected at Allen's Harbor, in 
about thirty feet of water, in a place that was quite well protected from heavy 
seas during ordinary weather. While it is of course possible that some large 
marine animal, like a shark or turtle, might have broken these sponges, it is 
more probable that they were injured during some hurricane. Now these 
sponges were collected in May, 1393, and the last hurricane before this that 
