40 
THE VOYAGE OF H.M.S. CHALLENGER. 
naturally shed from the cortex have not been forthcoming ; a non-spicular cortex through 
which the young sponges can easily make their way to the exterior has been the result. 
It is true that in a specimen of Craniella simillima, Bwk., also containing large 
embryos with fully developed spicules, the cortex retains its usual structure. The spicules 
of the embryos of this species, however, are not of so massive and rigid a character as 
those of Craniella schmidtii (?), so that the embryos may in this case possibly be 
squeezed through the canals of the parent sponge, which is, however, not a very likely 
supposition, and it is still possible that even in this sponge expulsion may take place 
through the cortex, which at a later stage may become modified for the purpose. 
Development. — Three stages in the embryonal history of the sponge are represented. 
In the earliest, the embryo is a solid, more or less oval body, about 0‘64 to 0’7 mm. 
long, by 0'4 mm. broad. It appears to consist entirely of rounded or polygonal cells, 
from 0'025 to 0'035 mm. in diameter; these cells are in turn composed of granular 
spherules, from 0'002 to 0'004 mm. in diameter; in their midst is one larger than the 
rest, and somewhat more deeply stained, varying from 0‘006 to 0'012 mm. in diameter; 
it must, I suppose, be regarded as the nucleus, though very different in appearance 
from ordinary nuclei. On the exterior of the embryo is a thin layer of structureless or 
finely granular stained material, which extends inwards between the cells, filling the 
interstices left by them. This material is similar in character and distribution to the 
structureless blastema which I have described as investing the blastomeres of Oscarella 
lohularis, 0. Sch. 
The existence of an outer layer of columnar cells could not be demonstrated in 
embryos at this stage though traces of such a layer appear in those of the succeeding 
stage. The subdivision of the large granular cells of the embryo appears to proceed very 
irregularly ; in one embryo, the other cells remaining unchanged, one in the centre has 
disappeared, its place being taken by a multitude of small (O'OOS to O'Ol mm. in diameter) 
more or less polygonal, finely granular, not deeply stained cells, in which a comparatively 
large spherical, vesicular nucleus, 0’004 to 0’006 mm. in diameter, with a small spherical 
deeply stained nucleolus, 0‘002 mm. in diameter, can be distinguished. 
In embryos of the next stage a structureless gelatinous matrix makes its appearance 
between the cells, many of which have lost their individuality and appear to be resolved 
into a crowd of granules ; others remain unchanged and are sharply defined against the 
gelatinous matrix. Numerous finer fibrils traverse the matrix and can be traced into 
connection with isolated granules and minute fusiform and polygonal cells. Traces of a 
columnar epithelium are now observable at the poles of the embryo, and spicules make 
their first appearance. These earliest spicules are oxeas; they are of almost hair-like 
fineness, and are arranged more or less radiately, their distal ends pointing at right angles 
to the surface of the embryo, their proximal ends lying tangentially about the 
centre. 
