In a living sponge a constant circulation of water is maintained by means of an aquiferous system (fig. 25), which is constituted by the oscula and pores - already alluded to - and by a system of canals excavated in the substance of the sponge, and uniting the two sets of apertures. The water passes in by the "pores" or inhalant apertures, and is conveyed by a series of canals - the "incurrent" or "afferent" canals - to a second series of tubes - the "excurrent" or "efferent" canals - by which it reaches the "oscula" and is finally expelled from the body. These processes are regularly performed, and their mechanism was long a subject of speculation. It is now known, however, that beneath the superficial layer or "dermal membrane" of the sponge there exist chambers lined with sponge-particles which are provided with vibratile filaments or flagella (fig. 25, c c). The pores open into these chambers, and from them proceed the incur-rent canals, each being dilated at its commencement into a sac, which is also lined with flagellate sponge-particles. By the vibratile action of these cilia, currents of water are caused to set in by the pores; and as out-going currents proceed from the oscula, a constant circulation of fresh water is maintained through the entire sponge. In this way each individual sponge-particle is enabled to obtain nutriment; the process being at the same time not improbably a rudimentary form of respiration. The chambers or sacs lined with flagellate sarcoids have been shown by Mr Carter to be, as previously pointed out, the essential element in the organisation of the fresh-water and marine sponges, and to be the fundamental expression of the alimentary system.
Fig. 25. - Diagrammatic section of Spongilla (after Huxley). a a Superficial layer or "dermal membrane;" b b Inhalant apertures or "pores;" c c Ciliated chambers; d An exhalant aperture or "osculum." The arrows indicate the direction of the currents.
In a few sponges (the Myxospongiae of Haeckel), as in the genus Halisarca, there is no skeleton, and the organism consists of an aggregate of masses of sarcode, permeated by branched canals, which are everywhere inflated into chambers lined by flagellate sponge-particles or sarcoids. As a general rule, however, the soft protoplasmic aggregate which constitutes the living animal of the sponge is supported by more or less extensively-developed hard structures, which collectively form the skeleton. The nature of the skeleton varies greatly in different forms, and the variations are of great importance in the identification and classification of the sponges. In the so-called "horny" sponges (the Keratosa of Bowerbank) the skeleton (fig. 26) is composed of numerous fibres of a horny substance ("keratode") interlaced to form a matted network. In the sponges of commerce (Spongia) the skeleton is simply composed of these reticulated horny fibres; but in most of the "horny" sponges (such as Halichondria, Spongilla, etc.), we find in addition numerous siliceous bodies which partly strengthen the horny fibres, and are partly scattered through the sarcode. These so-called "spicules" (figs. 26 and 27) are of very varied shapes, and of microscopic dimensions, and their form is often characteristic of the particular sponge in which they occur.
The horny fibre of the skeleton of the keratose sponges is hollow ; and in the sponges of commerce the axial hollow contains neither foreign bodies nor spicules. In other cases, the central tube of the horny fibres may be filled with sand-grains or with spicules, and in addition to the central core the surface of the fibre may be roughened by projecting spicules; while in some cases the fibre is almost entirely made up of simple spicules bound together by a small quantity of sarcode.
In the "so-called "calcareous" sponges (Calaspongiae) such as Grantia, Sycon, etc, the skeleton is composed of numerous calcareous spicules (fig. 24, A), which are in the form of simple fusiform rods or of three-rayed (rarely four-rayed) needles, and are arranged in different ways in different species. The three-rayed spicules are the form especially characteristic of the Calaspongiae, but two or all of the known forms of calcareous spicule may occur in a single sponge.
Fig. 26. - Fragment of the skeleton of a horny Sponge (after Bowerbank), greatly enlarged, showing interlacing horny fibres with spicula.
Fig. 27. - Different forms of the spicules of the horny, calcareous, and siliceous Sponges, greatly magnified.
In the so-called siliceous sponges (Silicispongiae) the skeleton is composed of siliceous spicules, of various forms and variously disposed. Very commonly the spicules which primitively compose the skeleton become in process of growth fused together by a secondary siliceous deposit, so that the skeleton becomes a continuous one. In other cases, though the spicules are permanently distinct, they are so interlocked with one another as to confer practical rigidity upon the skeleton. In other cases, the spicules are united with one another by sarcode only. Independently, also, of the true skeleton-spicules, the sarcode of the body contains scattered through it numerous "flesh-spicules" of various and often very characteristic forms. Modern investigations have brought to light a great number of siliceous sponges, both living and extinct, and the structure of these is in many cases of the highest interest. Nothing further can, however, be attempted here than to briefly characterise the two principal groups of the Silicispongiae - viz., the Hexactinellid and Lithistid sponges.
Fig. 28. - A, Dactylocalyx pumiceus, a Hexactinellid Sponge from the West Indies; B, A spicule of the Lithistid Sponge Discodermia, greatly enlarged, showing the branched ends of the spicule ; C, Part of the skeleton of the Hexactinellid Farrea occa, greatly enlarged, showing the continuous lattice-like framework, the component spicules of which are only recognisable by their six-rayed axial canals ; D, Plan of a single spicule of a Hexactinellid Sponge. (After Lutken, Sollas, and Carter.)