The mesoglaea is soft or firm, hyaline or granular; it contains cells of various kinds, and sometimes fibres. The cells are either naked or provided with a delicate membrane. They fall under the following heads: - (i) fusiform or stellate cells, the latter sometimes connected by their processes; (2) pigment cells; (3) vacuolated cells; (4) amoeboid cells of variable shape and wandering habit, sometimes becoming granular and then vacuolate, or converted into hyaline, retractile, lumpy bodies of irregular shape with fat-like contents (Chondrosia), or partially filled by a starch-vacuole1. Fibres or fibrils united in bundles or lamellae, and apparently "non-nucleated, occur in some instances, e. g. in Chondrosia, but never in Calcarea. Other forms of cells are the spongoblasts, the spicule-bearing cells, gland-cells, contractile fibre cells and nervous cells. The gland-cells are saccular or pear-shaped, disposed in a single layer below the ectoderm, towards which one or more granular processes extend.

They are found in many Non-Calcarea and some Leucone Calcarea, and have been most completely investigated in Aplysilla violacea and Dendrilla by von Lendenfeld (Z. W. Z. xxxviii. pp. 254-6; 278-80). He found that if the ectoderm were injured, the cells in question secreted a slimy substance (? spongin), hardening under water in about twenty-four hours into a cuticle, beneath which a new ectoderm and layer of gland cells were developed. The contractile fibre cells are fusiform, nucleated, and in Euspongia canaliculata (=anfrac-tuosa) showing traces of transverse striae, more or less regular2. They are found sparingly in Calcarea Heterocoela, but are very common in the Non-Calcarea, round the pores and oscula, accompanying the inhalent canals or the bundles of spicules, e. g. in Tethya, or in the cortex (p. 807), or surrounding the fibres of the skeleton in Aplysilla violacea and Dendrilla. Nervous elements have only been detected recently: their character and position leave little doubt as to their real nature. There are two forms of them, the palpocil and synocil. The former is a delicate free process, springing from a mesoglaeal cell with one or more basal outrunners, near to which cells apparently ganglionic have been detected occasionally.

A synocil, as yet observed only in a Sycandra, is a process of mesoglaea, .1 mm. long, containing a number of fine filaments derived from as many cells situate at its basel. Palpocils have hitherto not been found in Calcarea Homocoela: but in other sponges they occur round or near the pores of the inhalent canals (various Syconidae, Vosmaera gracilis); on the outer surface in groups or singly (various Leu-conidae); upon the vestibular membranes (Aulena villosd); upon the membranes of the lacunar inhalent system (Halme globosd); or bordering the margins of the lacunar extensions of the exhalent system (Euspongia canaliculatar=anfractuosa).

1 The granules in the granular cells appear to be albuminoid; cf. Keller, Z. W. Z. xxx. pp. 570-2; fat-like bodies are described in Chondrosia by Schulze, Z. W. Z. xxix. p. 104; starch-containing or amylum cells in various sponges by Keller, op. cit. pp. 572-6. Brandt considers that the starch is due to the presence of algae: see Mitth. Zool. Stat. Naples, iv. pp. 232, 296. For starch vacuoles and granules in Spongilla, see Ray Lankester, Q. J. M. xxii. p. 241. Carter states that starch is met with in the ova of marine sponges as well as in the gemmules of Spongilla (A. N. H. (5), xii. 1883). Krukenberg has found an ethereal oil in Chondrosia and some other sponges (Vergl. Physiol. Studien, (1), ii. p. 42 et seqq.).

2 These cells resemble in form the smooth or non-striped muscle-cells of many Metazoa. They are certainly endued with contractility, and the closure, whether complete or incomplete of the pores, etc, or movements of the surface of a sponge, are, beyond doubt, due to their action. They are called contractile fibre cells because they have no connection with a nervous system; but, if recent discoveries are correct, and there is connection between the sense, ganglion, and fibre-cells of the pores or canal system, as von Lendenfeld supposes in some instances, then they may be termed muscle-cells. It is difficult to say what their function may be when they surround the skeletal fibres as in Dendrilla.

There is in some Non- Calcarea a marked structural difference between the superficial and deep portions of the mesoglaea, constituting a cortex and medulla. This is especially the case in Chondrosidae, where the cortex is fibrillate; in Corticium, where it contains lacunae, each lodging a cell which fills it partially; in Weberella, Polymastia, etc, where it has a plentiful mesoglaea, imbedding stellate and fusiform cells and fibrils; in Tethyadae and Geodidae, where its structure is complex, due partly to layers of peculiar spicules, of cells, of contractile fibre cells. The mesoglaea round the ampullae is generally granular, but not in Spongelidae.