Group of the * Physemaria.” | 11 
surrounding water. Taken in its most highly characteristic 
phase, this compound gemmule, or wrongly called “ ciliated 
larva” of the sponge, represents a spherical or ovate agere- 
gation of typical collar-bearing monads or spongozoa, con- 
nected laterally and by their bases with one another, and 
with their anterior flagellate and collar-bearing extremity 
directed outwards. In this condition they, in fact, present a 
wonderful resemblance to a social colony of one of the simpler 
Gymnozoidal types, and might be directly compared to a 
detached capitulum of the pedicelled genus Codosiga, Jas.-Clk., 
or to a spherical colony of the free-floating genus Astrosiga, 
S. Kent. It is only when certain of these sponge-monads 
withdraw their characteristic collars and commence to throw 
off and secrete around them the spiculiferous syncytium as a 
nidus for the further development of the colony that their true 
sponge nature becomes apparent. This brief chapter of their 
developmental history brings out in high relief the potential 
importance of the collar-bearing monads compared with the 
remaining elements of the sponge community, showing, indeed, 
as already intimated, that these latter are entirely subordinate 
to and derived from these essential sponge-units. 
The compound ciliated gemmules, whose true nature has 
just been discussed, are most abundantly encountered in asso- 
ciation with the calcareous division of the sponge tribe. With 
certain of these forms, but more especially among the siliceous 
group, other compound ciliated bodies occur, concerning which, 
although it has not been hitherto attempted, a very similar 
interpretation may be rendered. Reference is here made to 
those spherical ciliated chambers that first received from Mr. 
Carter the title of “ampullaceous sacs.” By Prof. Clark 
they have been denominated the “ monad-chambers ;”’ and in 
those sponge-types where they are present they seem, so far as 
has been observed, to embrace the whole and entire system of 
the collar-bearimg monads, each such chamber in its normal 
and fully developed state being completely lined, except at its 
point of communication with the general aquiferous system, 
with these essential spongozoa. The, examination by me of 
a species of Desmacidon revealed the presence of these ciliated 
chambers in great profusion and in every phase of their deve- 
lopment. It was further elicited, by a careful study of the 
earlier stages of the same, that they take their origin by a 
process of multiple fission in all ways identical with what has 
been followed by the detached compound gemmules. Both 
present in their initial stage a morula-like aspect, the subse- 
quent and essential point of departure being that, while in the 
latter case the free-swimming ciliated gemmule has the collar- 
