XXXll 
THE VOYAGE OF H.M.S. CHALLENGER. 
’which persists in such adult spherical sponges as Myriastra clavosa, and in 
comparatively large but not fully grown specimens of Geodia barretti, that one 
can scarcely refrain from regarding them as comparable. That other observers 
have taken a different view is due to the remarkable structural similarity of the 
wall of the cloaca and the cortex ; this similarity has been taken for identity, and the 
cloaca consequently regarded as of secondary origin produced by an invagination of the 
cortex. The similarity does not however amount to identity, and another explanation is 
possible ; for if we refer to the young examples of Stelletta phrissens just mentioned we 
shall see how by a very natural process a cloacal wall similar to that of a cortex will arise 
by modification of the roots of the choanosomal folds, for these like the rest of the primi- 
tive folds will with the growth of the sponge suffer a metamorphosis, by which the 
flagellated chambers will disappear from them, and the mesoderm, increasing in volume, 
become converted partly into collenchyma and partly into fibrous tissue. Vela which 
appear generally throughout the excurrent canals will also arise about their terminations 
in the undivided remains of the paragastric cavity, which thus will become surrounded 
by a distinct wall, the complexity of which may be increased by the addition of a second 
series of vela behind those first formed (PI. XII. figs. 23, 40); in this way a structure 
remarkably like that of a cortex will result. The cloaca in this case will represent 
the remains of the original paragastric cavity, and thus the subsequent formation of 
meridional excurrent canals, as in Siphonia, by an outgrowth from the cloaca becomes 
intelligible. 
In fully grown specimens of Geodia barretti what appear to be the oscules of the 
sponge form a group over a scarcely depressed area, which is surrounded by a very distinct 
raised marginal ring ; the appearance is such as to naturally lead to the suggestion, 
which has been made, that we have here a case of commencing invagination of the cortex, 
the marginal ring representing the future margin of the commencing cloaca ; the history 
of the sponge shows, however, that so far from this being the case the exact opposite is 
true, the marginal ring is directly descended from the margin of the oscule of a previous 
stage and the apparent oscules are the openings of excurrent canals which originally dis- 
charged into a tubular cloaca ; so far from being a cloaca in process of formation this 
sfructure is a disappearing cloaca, — a cloaca which by lateral expansion has become con- 
verted into a shallow depression. 
In conclusion, it would appear that in most cases the oscules of such sponges as possess 
only one of these openings are directly descended from that of the Ehagon, as the cloaca 
is the direct descendant of the original paragastric cavity. As instances I may cite the 
following : — Tetilla sandalina (PI. I.), all the species of Thenea and of Myriastra, 
Theonella swinhoei (PL XXIX.), Siphonia, and N eosiphonia superstes (PI. XXXI. fig. 7). 
In the case of sponges like Jerea, which differs from Siphonia in not possessing a cloaca, 
we must suppose that the original cloaca has disappeared, or is only represented by the 
