allied to Rhabdopleura. 339 
and in connexion with the canals and cavities of the 
coencecium. 
The surface of the polyparium, moreover, is dotted, especi- 
ally at the bases of the spines, with large rounded apertures, 
which lead into the interior of the stem, the latter being honey- 
combed from end to end by an irregular system of wide canals 
and cavities intersected by bridles and arches, which thus 
provide for the constant ingress and egress of sea-water 
throughout the entire system. The inner surface of these 
canals and chambers is as smooth and glistening as the outer, 
the secretion being perfectly homogeneous and evidently the 
product of the inhabitants. It cuts with great readiness and 
as cleanly as a soft Fucus. Microscopically it consists of 
numerous layers of a translucent and very fine membranous 
secretion, the sheen or lustre being doubtless due to this 
arrangement. The whole disposition of the tissue points it out 
as the secretion of the polypides, just as much as the tube of 
an annelid or Phoronis, the more regular and less bulky ccene- 
cium of Rhabdopleura*, and, it may be, the shell of a mollusk. 
Like the annelidan tubes it most approaches, it is little affected 
at first either by nitric acid or caustic potash, though the 
former by and by somewhat softens and bleaches it. This 
secretion of Cephalodiscus is paralleled by the curious invest- 
ment or “house” of Appendicularta, and is therefore probably, 
like it, the homologue of the Ascidian test. The “house” has 
two funnel-shaped apertures, supported by a fibrous trellis- 
work, leading into the cavity containing the body. 
It is not a matter for surprise that creatures so minute 
should secrete so conspicuous a home for themselves, or that it 
should preserve the algoid or zoophytic outline, especially 
when the productions of sponges and other forms are remem- 
bered, or when we reflect that even’a transparent structureless 
fluid inside a smooth capsule (as in the Nemertean stylet- 
pouch) can produce, in countless examples of each species, 
precisely the same form of solid crystalline stylet. ‘The en- 
listment of numbers in the present case supplies any deficiency 
likely to arise from minute size. ‘The peculiar shape of the 
coencecium, moreover, has probably been found that best 
adapted for the preservation of the animals, by its resemblance 
to the seaweeds in its neighbourhood, on the one hand, and, 
on the other, by its affording complete aeration, abundant 
supply of food, and security to the little architects and their 
delicate plumes. 
In the interior of the cavities and canals of the semitrans- 
parent coencecium numerous opaque masses (the polypides) and 
* Rhabdopleura Norman, Allman, and R. mirabihs, Sars. 
23* 
