REPORT ON THE TETRACTINELLIDA. 
Ixxxvii 
anatrisenes wliicli arise within the cortex and proceed from it for a short distance beyond 
the surface precisely after the manner of the cladoxeas in Proteleia. 
Methods of Attachment. — Some of the Tetractinellida are without attachment or free, 
but the greater number are in some way or other attached to the sea-floor ; in many 
cases the sponge is incrusting {Placina, Astropeplus), in others it rises from an in- 
crusting base [Pmcillastra, many Lithistids), in some it is attached by strong fibrous bands 
produced by an outgrowth from the cortex {Pilochrota), and even in the case of sponges 
which are practically free, similar processes are present and attach to the sponge numerous 
small stones and other foreign bodies. In most deep-sea Sponges the attachment is by 
means of radical spicules, variously adapted to this purpose as already described (Fig. XIII.). 
Some species are excavating [Pilochrota? lactea, Carter), and some are parasitic (an 
undescribed species of Geodia which occupies the oscular tubes of a species of Ectyon).^ 
Migration, Protrusion, and Extrusion of Spicules. — Observations on various sponges 
[c.g., Tetilla grandis, p. 12 ; Anthastra communis, p. 144) show that the young trisene 
spicules originate within the choanosome at some little distance from the cortex, from 
which the cladome is the more remote as it is less advanced in growth. The cladome of 
the fully grown trieene on the other hand lies as a rule either within the cortex or beneath 
the floor of the subcortical crypts. Thus with the growth of the spicule, the cladome is 
carried from the interior towards the exterior of the sponge. This is no doubt partly 
due to the fact that the growth of the rhabdome is chiefly in the long direction ; and the 
absence of triaenes from all parts of the spicular fibres except their distal terminatious 
may be explained as resulting from the close connection into which the cladome is 
brought with the cortex, so that the latter in its growth carries the former along with it. 
It is, however, possible that in addition to the movement of the spicule by growth, 
another takes place by which it is gradually but bodily translated from the interior 
towards the exterior of the sponge ; and only by some such process does it appear 
possible to explain the presence of sterrasters within the choanosome of the Sterrastrosa 
in all stages of development, while in the cortex none but those fully adult are met with. 
Supposing this outward migration to occur, its continuance would lead to the pro- 
trusion of the distal end of the spicules, and thus the occurrence of hispidating rhabdi 
and trisenes is to be explained. 
With a further progress outwards the hispidating spicules would at length lose all 
connection with the sponge and fall out as deciduous spicules. 
Since there is nothing to distinguish the deciduous spicules of a living sponge from 
those left behind by a dead one, it is not surprising that this process has hitherto been 
overlooked and indeed unsuspected ; there are cases, however, in which, while the 
organic connection of the spicules with the sponge is dissolved, yet a more or less close 
association persists ; thus in Chrotclla maccllata (p. 20) the toxaspires, which become 
^ Sollas, Ann. and Mag. Nat. Hist., ser. 5, vol. v. p. 408. 
