XXX 
THE VOYAGE OF H.M.S. CHALLENGEK. 
equatorially situated canals the currents setting equatorially are strengthened, and thus 
the sponge as a whole is benefited, since water lying on the surface of the sea-floor is 
skimmed off over a greater area than would otherwise be the case. 
Canal- Systems possessing Radiate Symmetry. — In some few sponges the larger canal- 
folds are formed on a symmetrical plan, which is constant for the species. Thus in the 
youngest known examples of Disyringa dissimilis, we meet with four S 5 nnmetrically 
arranged longitudinal excurrent canals, and four incurrent canals regularly alternating 
with them (p. 170, PI. XLI. fig. 4). This points to a folding of the choanosome in the 
manner shown in the diagram (Fig. VIII. 1). Subsequent to the folding we may suppose 
that the sinuses of the folds became converted into canals by concrescence at their roots 
■ (Fig. VIII. 2). 
It would appear as though the folding in this case were brought about by four 
invaginations of the ectoderm, commencing in one situated at the antoscular pole, 
whence they proceeded, burrowing beneath the ectosome, in a meridional direction to- 
folding of the choanosome, produced by the invagination of four incurrent canals (I), and leading to the formation of 
four primitive excurrent sinuses (E). 2. Stage in which the sinuses have been converted into canals (E,) (1 and 2 are 
imaginary sections taken through the body of the sponge). 3. Stage in which secondary excurrent canals (Ej) occur as 
outgrowths of the primary. 4. Stage in which tertiary excurrent canals (E 3 ) are budded off from the secondary 
(3 and 4 are imaginary sections taken across the cloacal tube). 
wards the oscule. With increased growth additional excurrent canals appear, and these 
are symmetrically arranged with respect to the four primitive canals, lying in the cloacal 
tube in the angles between them ; how they are directed in the body of the sponge we 
do not know, since of this stage only the cloacal tube is known. In Siphonia we have 
already had reason to suppose that additional excurrent canals arise by a kind of 
budding from the endoderm of the cloaca, and the origin of the additional canals in the 
sponge under consideration may be similarly explained (Fig. VIII. 3). 
Finally, eight additional excurrent canals appear, and these may be explained as 
paired outgrowths of the secondary ones (Fig. VIII. 4). 
The symmetrical arrangement of the canals in Disyringa, or rather in the closely 
related sponge, Agiardiella schultzn, has been adduced by Marshall in support of the 
hypothesis that the sponges are degraded Coelentera. These sponges are, however, 
somewhat highly specialised members of the Stellettidae, a family by no means primitive 
itself and one which in no other instance displays any tendency towards special 
