XIV 
THE VOYAGE OE H.M.S. CHALLENGER. 
three layers of tissue, the fundamental ectoderm and endoderm, and the so-called 
mesoderm.^ 
The mesoderm is a very thin layer, consisting of collenchyma (see p. xxxviii), 
the ectoderm is an epithelium of pavement cells {'pinacocytes) , and the endoderm is partly 
composed of flagellated collared cells {choanocytes), and partly of pinacocytes. 
The hypophare consists of mesoderm lined by pinacocytal endoderm above, and 
ectoderm below ; the spongophare is characterised by evaginations of the endoderm, 
which forms pouch-like recesses in the mesoderm, and within these it consists of choano- 
cytes, elsewhere of pinacocytes. 
The recesses, known as flagellated chambers, communicate with the cavity of the 
sac {jparagaster) each by a single wide mouth {apopyle), and with the exterior by a small 
pore {prosopyle). 
A single comparatively large opening (oscule) at the summit places the paragaster in 
free communication with the surrounding water. The oscule is not derived from the 
blastopore, which in the larval stage immediately preceding that of the Rhagon is 
situated in the centre of the primitive hypophare. 
Pig. II. — Diagram of a young sponge, showing the folded spongophare. S, spongophare ; H, hypophare ; 0., oscule ; 
G., paragastral cavity ; e, excurrent canal ; i, incurrent canal ; p, pore. 
In the case of many spherical sponges which in their earliest known stages are not 
attached, but free, the hypophare is probably not developed at all (or possibly metamor- 
phosed at a very early stage), the Rhagon in all probability having a spherical form and 
consisting wholly of the spongophare. 
Doubtless the Rhagon is readily derivable from the Ascon ; but it never actually 
passes through an Ascon stage, differing in this respect from the Sycon, from which it 
is also distinguished by the more spherical form of its flagellated chambers. The whole 
course of embryological history of the Rhagon difiers from that of the Megamastictora 
in numerous details, and particularly in the absence of an amphigastrula stage. 
The successive stages by which the Rhagon passes into the adult sponge have not 
been traced, but Schulze has sufiiciently shown, at least in some instances, the probable 
course of events, which consists simply in a folding of the spongophare which proceeds 
1 These terms are used merely for convenience ; it is far from certain that the tissues they denote are homologous with 
those similarly named in the Metazoa. 
