Coelenterata of very varied and often inconstant shape; frequently massive; devoid of tentacles. Concrescence between parts of the same individual or of different individuals of the same species is very general. The body-wall is perforated by innumerable minute inhalent pores, and as a rule by one or more larger exhalent apertures - the oscula. The epithelia are uni-laminar, the endoderm cells typically collared and flagellate, or in certain regions only, flattened and either flagellate or non-flagellate. The gastric cavity is (1) simple and either tubular or vasiform; or (2) tubular with radial outgrowths or cones, a system of inhalent canals more or less complex leading to it, and the tubular central cavity sometimes partially replaced by a system of exhalent canals; or (3) it is represented by sac-like, pyriform, or spherical ampullae with inhalent and exhalent canal systems as in (2), the central cavity being as a rule much restricted in extent. Skeletal structures are rarely absent, and consist of either variously shaped spicules hardened by calcite or silica, or of horny spongin fibres with or without siliceous spicules or foreign inclosures. Unisexual or hermaphrodite.

The sexual cells are mesoglaeal': the ovum developes into a ciliated larva within the mesoglaea except in Clione (? Clionidae). Discontinuous gemmation occurs in two forms. Colonial (?). Marine with the exception of the freshwater Spon-gillidae.

The external outline of the body even when most constant and characteristic, e. g. in some Calcarea, many Hyalospongiae, is not uniform as it is in other Metazoa. Indeed, it is extremely variable in most instances, the variability amounting to Polymorphosis, or change of form without a corresponding physiological division of labour; see p. 238. The variability in question depends chiefly on two factors: - (I) an irregular mode of growth; and (2) on concrescence or a fusion of parts belonging to the same sponge, the same colony (?), or to sponges of the same species growing hard by one another. Such fusion frequently leads to the enclosure of spaces really external to the sponge-body, which form a false gastric cavity (pseudogaster) opening by a false osculum (pseudosculum s. pseudostome) and false pores (pseudopores). Anatomy, especially that of the gastric and skeletal systems, affords at present the only characters of classificatory value.

The primitive Poriferan, as seen in the simple form (Olynthus) of the Asconid Leucosolenia (Ascetta) primordialis, is a vasiform individual with an osculum or aperture at one end and a peduncle of attachment at the other; with thin body-walls perforated by numerous pores, and a gastric cavity of the same shape as the body, lined throughout by collared and flagellate endoderm cells. Such simple forms are few, and the external shapes assumed by Poriferans are very numerous and may differ at various stages of growth. A radial symmetry may be acquired by the development of conular outgrowths from the body or by the arrangement of the skeleton. But a sponge may be a shapeless or sub-spherical mass, tubular, cup or saucer-shaped, columnar, leaf or fan-like, branched or dendriform. It is very rarely, if ever naturally, free; but either adherent to some foreign object or rooted in sand or mud by basal processes or by spicules as in many Hyalospongiae. The Calcarean Homo-derma is an exceptional instance of the union of individuals by a creeping tubular stolon. The Clionidae bore into shells or stones1. The thickness of the body-wall undergoes great increase, and its consistence is variable, both depending largely on the character of the mesoglaea and the skeleton.

And the gastric cavity undergoes at the same time great changes.

The ecto- and endo-derm are unilaminar. The former consists as a rule of flattened, rarely cubical, polygonal cells, with clear contents except immediately round the nucleus. Each cell bears a flagellum in Oscarella (Halisarca) lobular is, Plakina monolopha, Aplysilla violacea, and Dendrilla) 1 and in Halisarca Dujardini the cells appear to change into a slimy layer. The pavement epithelium lining the inhalent cavities and canals (pp. 793-5) is probably always, in some cases certainly, of ecto-dermic origin. The endoderm cells are typically flagellate, the base of the flagellum surrounded by a clear collar - an extension of the exoplasm of the cell. The endoplasm is variably granular and pigmented: it generally contains vacuoles (? contractile). The cells are large in the Calcarea (•012 mm.), small in Non-Calcarea. Their shape is not the same in all sponges: they are as a rule closely apposed, but in Euplectella are planted apart in diagonal lines, connected, however, by slender cords of protoplasm.

This typical endoderm is restricted, except in Calcarea Homocoela, to limited portions of the gastric cavity, and is reduced in the oscular tube and exhalent cavities and canals (post) to a pavement epithelium, flagellate in Oscarella lobularis, Plakina monolopha, and Aplysilla violacea (at least in part), but usually non-flagellate2. Columnar cells are sometimes found where the two epithelia pass into one another. The mesoglaea is scanty in Calcarea, especially in Homocoela, and most particularly so in thd Hyalospongian Euplectella. It is soft or firm, hyaline or granular; it imbeds the skeleton, and contains cells of many kinds, some of which are contractile or muscular, whilst others have lately been regarded as sensory and nervous (pp. 806-7).

1Nassonow states that the larval Clione commences to bore into a shell as soon as it fixes itself, and before any skeletal structures are developed. It throws out slender processes, which perforate the shell-substance and cut out small semi-ellipsoidal pieces, which are taken up and then cast out by the sponge. The penetration of the shell-substance is effected beyond doubt by chemical means. See Z. W. Z. xxxix. pp. 297-300.

The gastral system is one of the most characteristic features of the Pori-fera; it opens externally by pores and oscula. The pores are very numerous, round or oval in shape; minute as a rule, e. g. in Asconidae, .oi - 02 mm. as a mean; sometimes large and small (macro-, micro-pores) in the same sponge; and in many instances capable of being closed either partially or completely. They are either scattered or aggregated in areae, and in the latter case generally perforate a thin pore-membrane stretched across a larger canal or aperture. False pores may be similarly aggregated. The whole pore-system is, properly speaking, inhalent, i.e. gives entrance to currents of water, but when there is no osculum certain pores are large and exhalent1. The osculum or proper exhalent vent is usually considered as the homologue of the mouth, a view beyond a doubt incorrect, (pp. 801-3). It is at first single, but during growth the number undergoes increase; and then the new oscula may be either scattered or grouped together.