REPOET ON THE TETRACTINELLIDA. 
61 
is a little smaller, viz., 18 mm. in diameter and 12 mm. in height. The oscule of one of 
the larger specimens is 3 mm, in diameter. The slender thread-like rootlets are about 
25 mm. long. There are about four or five to each sponge. The oscule, which is 
not protected by a spicular fringe, leads into a wide almost spherical cloaca lined by a 
fenestrated membrane, beneath which are seen the circular openings of several small 
excurrent canals. 
Horizontal transverse sections w^ere prepared to determine whether any trace of radiate 
symmetry occurred in the arrangement of the excurrent canals. The result was negative, 
but the sections confirmed a suspicion suggested by an examination of the entire sponge 
both in this species and others, namely, that the excurrent canals near the oscule are 
bounded on the outer side by the ectosome only, as shown at a, fig. 1 0, PI. VIII. It is 
possible that a membrane like that of which traces are shown at h may originally have 
existed, and subsequently been torn away or displaced in the preparation of the slices ; 
but I could, at all events, detect no signs of it in any of them. If the appearances 
represented in the drawing are to be relied upon, we may suppose that after the folding 
of the choanosome which gave rise to the canal system, an upward growth of the sponge 
took place, in which the choanosome only incompletely participated, leaving only 
longitudinal zones of ectosome to complete the external covering of the excurrent 
canals. 
In vertical transverse section (PL VI. figs. 13, 14) the regular radiate arrangement of 
the spicules is clearly shown (though not in the illustrations), as well as the interdigitation 
of the excurrent and incurrent canals, the latter starting chiefly from the poriferous zone 
as widely open channels which ramify within the substance of the sponge as they proceed. 
The spicules are remarkably few in number, and this is true of all alike, from the 
oxeas to the amphiasters. The small size of the somal anatrisene is worth noting, though 
a similar reduction, but not to the same extent, occurs in some other species of the genus, 
e.g., Thenea wyvillii. 
Still more interesting is the transformation of the radical anatrisenes; in most species 
of Thenea a few club-shaped spicules, which are reduced anatriasnes, occur along with 
the normal anatrisenes of the roots, but in this case all the spicules which compose the 
rootlets are of this form. Most frequently the axial fibre of the rhabdome after entering 
the tylus gives off three irregular fibres, either from the same point, not necessarily 
the end, or at different points along its course. These fibres are thinner than that of 
the rhabdome, swollen and constricted irregularly, and crooked in their course; they 
usually bifurcate, and frequently give off a number of small lateral branchlets which 
give them a ragged appearance. Frequently when they reach the surface of the tylus 
their termination is marked by a small rounded projection, suggestive of an aborted 
cladus. In other cases only two, or even one, branch may arise from the axial fibre, 
and sometimes branches are altogether absent. In this last case the spicule is in every 
