The mode in which nutrition is effected has given rise to much debate. A living sponge is traversed by currents of water passing in at the pores and out of the oscula, or, if oscula are absent, out of other pores (see p. 793, and note 1). The currents may be suspended, but probably not reversed: they may be swift (? nutritive) or slow (? respiratory). Nutritive substances, whether animal or vegetable, alive or dead, are carried into the sponge suspended in the currents; and they have been observed in the mesoglaeal cells; see Metschnikoff, Z. W. Z. xxxii. pp. 372, 374; von Lendenfeld, Z. W. Z. xxxviii. p. 254. Particles of carmine suspended in water are taken up solely by the collared cells in the young Spongilla (Carter) and Oscarella (Heider). Metschnikoff found (loc. cit.) that in Ascetta (Leu-cosolenid) primordialis and Spongilla carmine particles were taken up by the collared and mesoglaeal cells; so too in Halisarca Dujardini and H. pontica. Overfeeding the latter caused obliteration of the canals. On the other hand, no carmine was discoverable in the ampullae of Reniera aquaeductus and Siphono-chalina coriacea, though present in the mesoglaeal cells. Von Lendenfeld experimented on Aplysilla violacea.

Carmine particles were taken up by the epithelia of all parts if the sponge were kept for a sufficiently long time in water containing them in suspension; but the particles absorbed by the ectoderm of the subdermal cavities soon passed into the amoeboid wandering cells which were present in numbers in that part of the sponge; by them they were conveyed to the ampullae, and then excreted by the collared cells in a few days' time. Their angles were rounded off. Particels absorbed by the collared cells themselves were expelled unchanged. Whilst the epithelia of the subdermal cavities and ampullae, and the mesoglaeal cells were thus freed from carmine, this was not the case with the epithelia of other parts, even after the lapse of two months. The obvious conclusion is that absorption takes place by the ectoderm of the subdermal cavities, digestion by the mesoglaeal cells, excretion by the collared cells. Polejaeff thinks (Calcarea, p. 15) that a nutritive function must be assigned also to the endodermic pavement cells.

He finds (op. cit. p. 16) that the pavement cells here and there in the inhalent canals, in the exhalent, or in both alike, may be swollen and more granular than usual, facts very well seen in Leucetta vera1. The mechanical objections he urges (op. cit. p. 15) against the view that the collared cells are the chief agents in the absorption of food, do not seem well founded, especially if the flagella may be supposed to act in a manner similar to that in which they act in the Choanoflagellata. It may be noted that Krukenberg has extracted a diastatic and peptic ferment from many sponges: the peptic ferment is replaced by a tryptic in very few instances (Sycon = Sycandra, Reniera porosa, &c). A piece of raw fibrin laid upon the surface of Suberites massa, S. domuncula, Chondrosia reniformis, underwent resorption: so too in the osculum of S. domuncula. Surface digestion did not take place, however, in Hircinia variabilis, Spongelia elegans, and Euspongia adriatica. Fibrin was also digested when inserted into the mesoglaea of Suberites massa, but not of S. domuncula.

See Krukenberg, Vergleich. Physiol. Studien (1), i. 1881, pp. 65-75; and on reserve material in sponges, ibid. ii. p. 42 et seqq. p. 57; cf. ibid. iii. p. 113.

1 The synocils probably shrink at death; hence their cells may become separated.

Vosmaer and Heider regard the Porifera as a type in value coordinate with Coelenterata, Polejaeff, Schulze, and others as a sub-type, agreeing with a subtype represented by the three other Coelenterate classes as far as the gastrula-stage, and then diverging. The group is retained here among Coelenterata for reasons given pp. 715-16; but in the present uncertain state of opinion the value of a class assigned to it must be deemed provisional only. Both Vosmaer and Polejaeff term the Calcarea and Non-Calcarea classes, but it may be doubted whether the essential distinction between them, viz. the hardening of the spicules by calcite or colloidal silica respectively, is a sufficient basis for so great a classificatory distinction. The differences that obtain between some of the orders of Insecta are quite as great. As to the subdivisions of Calcarea and Non Calcarea, they constitute in the former a well-defined series, but in the latter their mutual relations, and in the Spiculispongiae and Cornacuspongiae their limits, are by no means established.

The points to be borne in mind in determining the position of the Porifera seem to be as follows: - (1) There are various forms of larvae, the one known as amphigastrula being probably a modified and not a primitive one; (2) the gastrula is formed either by invagination, or by delamination from cells the common rudiment of both endoderm and mesoglaeal cells; (3) the invaginate gastrula is fixed by its mouth, and the osculum is a secondary formation; but the gastrula cavity is a true archenteron, and there is no reason for withholding the term 'gastric cavity,' used in a morphological sense, from the various forms assumed by the archenteron in the adults; (4) the inhalent system of canals is primitively formed by foldings of the surface of the body, and the pores are probably structural adaptations to the mode of fixation; (5) the increase of the mesoglaea, and the consequent irregularity of form are possibly dependent on the presence of pores; (6) the skeleton is formed entirely by the mesoglaeal cells as in some Anthozoa Alcyonaria, and is much specialised; (7) the typical collared character of the endoderm cells is a unique feature; but it must not be forgotten that in the planula or larva of Oscarella Nodularis the ectoderm cells are also collared; (8) the group is one of extreme antiquity.

1 It is possible that there are sponges in which there are no ampullae. At least Vosmaer has been unable to detect them in the thin fan-shaped Phakelliae: see Bijdrag tot de Dierkunde, pt. 3, No. 12, Sponges of the Willem Barents Expedition, 1885, p. 24, Fig. 12. Canals pass from one side of the body to the other in these instances.