EEPORT ON THE TETEACTINELLIDA. 
xvii 
him to the formation of the body-cavity in the Ccelomata. Coincidently with the 
cleavage, the choanosome folds within the ectosome, the cleavage cavity enlarging into 
the incurrent sinuses (Fig. VI.), whUe the places where the choanosome and ectosome 
remain coherent mark the position of the ends of the excurrent lobes. 
I have been careful to say that it appears as though the subdermal cavities arise by 
fission, for, on the face of it, if this apparent method of formation be the true one, it 
introduces great difficulties in the interpretation of the structure of the sponge on 
the germ-layer theory ; for instance, in sponges where the supposed fission occurs, the 
epithelium would not be formed from the ectoderm as in Placina, but as an endothelium 
derived from the mesoderm, unless — and the supposition is improbable — an invagination 
of ectoderm occurs coincidently with the progress of cleavage. On the other hand a rapid 
invagination of ectoderm might readily be mistaken for cleavage, and the facts presented 
by a larval form of Stelletta phrissens seem to point to such an explanation. In this 
larva, not more than 0'65 mm. in diameter, the ectosome is already differentiated as a 
separate layer from the choanosome, but some of the subdermal cavities have an almost 
spherical form, no wider than high, and might very well result from invagination of the 
ectoderm; this given, the rest becomes clear; we have only to suppose that these cavities 
increase with the continued folding of the choanosome and the wide incurreut sinuses will 
be produced with of course an ectodermal lining. 
In sponges next higher in the scale to Tetilla pedifera and Thenea delicata, such 
for instance as Thenea muricata and Tetilla grandis, the choanosome^ is not only 
intricately folded again and again, but structurally modified in certain regions by the 
increased development and differentiation of the mesoderm and the suppression of the 
flagellated chambers. Thus the main excurrent sinuses or canals no longer directly com- 
municate with flagellated chambers situated in their walls, for these chambers have become 
restricted to the smaller secondary or tertiary sinuses of the smaller or secondary folds, 
and an abundant development of mesoderm about the main sinuses has converted them 
into thick-walled canals. 
From the walls of these canals, transverse outgrowths are developed at close but 
irregular intervals, and extend as thin diaphragms for a greater or less distance across 
the lumen ; sometimes converting a canal into a succession of vesicles. These velar 
diaphragms, or vela as we shall briefly term them, consist of an extension of the epithelium 
of the canal, ectodermal or endoderma] according as the canal is incurrent or excurrent, 
with a thin intervening layer of mesoderm. In the centre of the velum is an aperture 
(more rarely two) which can be enlarged or diminished by the action of fusiform muscle- 
cells (myocytes), which are arranged partly concentrically and partly radiately about it. 
The flagellated chambers still communicate directly with those excurrent canals about 
which they are situated ; so that the chamber- system is still eurypylous. 
^ It will be recollected that the spongophare is now differentiated into an ectosomal and choanosomal region. 
(zooL. CHALL. EXP, — PART Lxiii. — 1888.) Err c 
