og SPONGES 
ward; while at A it is evident that the tips of two twigs came directly in con- 
tact, and upon anastomosing the growth of both was checked. ) 
Itis absolutely certain according to my most carefully made and often re- 
peated observations that no buds ever start from the side of a twig or branch. 
Hence the interior of the skeleton of the Tube Sponge presentsa very finished 
appearance, there being no protruding twigs or arrested growths which have not 
anastomosed. (See Fig. 17, K, where I have given a portion of the interior of 
the skeleton, life size. This is not, however, where the twigs have come in 
contact with the skin of the outer surface, or with the membrane lining the 
central tube, for here as I have previously remarked, we may find many short, 
protruding twigs. 
As both sarcode and skeleton become arrested as soon as they become covy- 
ered with the lining membrane of the central tube, it will at once be seen that 
there can be no increase of growth inward after this membrane is grown; and 
that the form of the sponge is determined by it. The tendency of the growth of 
the sponge is thus upward, and although, somewhat singularly, the upward 
growth sometimes considerably precedes the thickened lining membrane (See 
lig 7, A) the inclination of growth is very rarely, or never, inward. For 
proof of the assertion that the lining membrane arrests any inward growth, see 
Fig. 7, A and B, a, a, a, a, being the lining membrane of the tube, also in the 
young specimen, fig. 13, B, where the tendency of growth is outward so as to 
sradually form a wider tube. 
Returning again to arrested growths of the twigs of the skeleton, we find 
that those on the outer surface of the sponge, differ asa rule, from those on the 
