In the Geodidae and several Ancorinidae the pores may open each into a cortical funnel (/sops), or into narrow canals which unite, or into sub-dermal spaces, from which in either case cortical funnels lead into a system of anastomosing subcortical crypts, and these in their turn into the inhalent canals. The exhalent canals may be wide spaces, or slender canals, uniting together to form larger and larger spaces or canals, which open finally into an oscular tube. The length of the latter varies; when very short it is often termed cloaca. Isops has exhalent cortical funnels comparable in all respects to the inhalent2.

1See von Lendenfeld's figure of a section of Ascetta (Leucosolenia) Madeayi (op. cit. ante, ix. Pl. 62, Fig. 12, cf. p. 1147), where small pores open externally, and large into a pseudogaster: so too in Leucopsis (ibid. p. 1089). He sometimes speaks of the inhalent canal closed by a pore-membrane as the pore\ cf. Z. W. Z. xxxviii. p. 240-1. In Euplectella there are pores of large size (= oscula ?) in the body-wall which open direct into the central cavity of the sponge, and have no connection to the gastric ampullae.

2The cause of lipostomy is generally assumed to be the obliteration of the osculum in growth. It is possibly never formed in some instances. Euplectella is often quoted as an example, but the sieve plate at the summit of the body contains ampullae identical with those of the body-wall.

3Haeckel terms these apertures 'gastral ostia.' He supposed the cones to possess also distal or 'dermal ostia/ non-existent according to more recent researches.

1The subfamilies named have been established by von Lendenfeld, Proc. Lin. Soc. New South Wales, ix. pp. 1090-1110.

2The family Sylleibidae, with two genera, Vosmaera and Polejna, is due to the same authority (op. cit. p. 1110). Polejna corresponds to Polejaeffs genus Leucilla (Calcarea, Challenger Reports, viii. p. 51). Vosmaera is a peculiar type. The ampullae are disposed in a radial fashion round the oscular tube, but communicate with a branched inhalent and exhalent canal system. The genus corresponds to Polejaeff's Leucetta in part, i. e. to L. imperfecta and L. vera. L. Haeckeliana is included by von Lendenfeld through apparently a mistake, as may be seen on comparing his definition, p. IIIO, with PolejaefTs figures, op. cit. VIII.

3Darwinella and Ianthella are stated by Polejaeff to agree with Aplysilla (Keratosa, op. cit. xi. pp. 22, 23).

The subdermal spaces and subcortical crypts are not to be confounded. The former lie immediately beneath the outer surface, covered by a thin layer of sponge substance: they are large simple cavities usually communicating one with another3. The subcortical crypts form a set of anastomosing cavities between the cortex and medulla, i. e. at a deeper level than the subdermal spaces. The funnels or chonae (supra) are divided by a constriction into an outer part, the ectochone, long and cylindrical, and an inner part, the endochone, short and more or less hemispherical4. Other structures to be noted are the endogastric septa of Haeckel, cords or membranes extending into or across the oscular tube and anastomosing together. They are of secondary origin and occur in both divisions of Porifera5.

1Polejaeff states that pore-canals to the ampullae are wanting sometimes in Spongidae; Kera-tosa, op. cit. p. 17. Transitional or variable forms between types (3) and (4) are certain to occur, and probably not infrequently.

2 In the Auleninae, the sponge appears to have a very irregular mode of growth into processes or lamellae. The intervals between the lamellae, etc, are generally traversed or protected by perforated membranes, thus becoming lacuniform. The lacunae are common both to the pores and oscula. Von Lendenfeld, Proc. Lin. Soc. New South Wales, x. p. 283; cf. his table, p. 490.

3The spaces in question are traversed in Bajalus by anastomosing filaments, in Dendrilla by slender vertical filaments, in both cases composed of mesoglaea covered by epithelium.

4 For the subcortical crypts and funnels, see Sollas, A. N. H. (5), v. pp. 135, 140, 252; and on hops, pp. 396, 402.

5 See Haeckel, Kalkschwamme, i. p. 252; Vosmaer, Mitth. Zool. Stat. Naples, v. p. 489, and Porifera, Bronn's Thierreich, etc, ii. p. 128.

The skeletal structures are either inorganic or organic, separate or in union, or partially, very rarely completely, represented by foreign inorganic bodies. They are rarely absent entirely, as in Oscarella, Halisarca, Bajalus, and Chondrosia, and are always lodged in the mesoglaea. The proper inorganic skeleton is made up of spicules. A spicule consists of a very scanty organic basis or spiculin, laminated, hardened by Calcium carbonate as Calcite in the Calcarea, by colloidal silica in the Non-Calcarea, and inclosing an axial thread of organic matter, the so-called canal, said to be absent sometimes. It is formed in the first instance within a celll, beyond which it soon projects. It continues to increase in length and thickness. The calcareous spicule is covered by a sheath or cuticle, by the calcification of which the spicule grows, and outside the cuticle by a layer of mesoglaeal cells except when it projects freely, the covering cells in this case being derived from the ectoderm (or in the gastric cavity from the endoderm?). The siliceous spicule appears to remain in close connection with a cell until it is of full size. When it projects above the surface it is, so far as is known, naked.

It sometimes attains a great length, especially in the rooting spicules of Hyalospongiae, e. g. of Hyalo-netna, which reach even two feet or more. As to shape, four leading types are recognised in accordance with the number of axes traceable, (I) Monaxile; straight and then pointed, blunt, or knobbed, smooth or spinulose; curved, and then of very various shapes - hook, anchor, bow, etc.: (2) triaxile, characteristic of Hyalospongiae and only present in rudiment elsewhere; the three axes are those of an octohedron, equal or unequal in length, one even suppressed in some instances: (3) tetraxile; the four axes are lines drawn from the centre to the vertices of a tetrahedron; of equal or unequal length; becoming triaxile by the suppression of an axis, as appears to be the case with the triaxiles of Calcarea and some other Porifera: (4) polyaxile; as in the stellate, or the globular, spicule; the stellate globule, etc. The ends of the axes may be pointed, knobbed, or branched. Most sponges possess a variety of spicules when they are present: some Calcarea, Renieridae, and Suberitidae, have needle-shaped monaxiles alone.