Keller,[[268]] on the other hand, finds that the spherasters of the Tetractinellid Chondrilla originate as spheres (Fig. 119, B); and spheres have been observed in the gemmule of a Tethya; no spherasters were as yet present in the gemmule, and spheres were absent in the adult.[[269]]
In the genus Placospongia certain spicules are present which outwardly closely resemble the sterrasters so characteristic of certain Tetractinellidae. Their development, however, as will be seen from Fig. 120, shows that they are not polyaxon but spiny monaxon spicules. Placospongia is consequently transferred to the Monaxonida Spintharaphora.
Sterrasters originate within an oval cell as a number of hairlike fibres[[270]] (trichites), which are united at their inner ends. The outer ends become thickened and further modified. The position occupied by the nucleus of the scleroblast is marked in the adult spicule by a hilum.
Fig. 121.—Three stages in the development of an anisochela. al, Ala; al', lower ala; f, falx; f', lower falx; r, rostrum; r', lower rostrum. (After Vosmaer and Pekelharing.)
The anisochela has been shown repeatedly to originate from a C-shaped spicule.[[271]]
What little is known of the development of Hexactinellid spicules we owe to Ijima.[[272]] Numerous cells are concerned in certain later developmental stages of the hexaster; a hexaster passes through a hexactin stage, and—a fact "possibly of importance for the phylogeny of spicules in Hexactinellida"—in two species the first formed spicules are a kind of hexactin, known as a "stauractin," and possessing only four rays all in one plane (cf. Protospongia, p. [207]).
Physiology
Production of the Current.—It is not at first sight obvious that the lashing of flagella in chambers arranged as above described, between an inhalant and an exhalant system of canals, will necessarily produce a current passing inwards at the ostia and outwards at the osculum. And the difficulty seems to be increased when it is found[[273]] that the flagella in any one chamber do not vibrate in concert, but that each keeps its own time. This, however, is of less consequence than might seem to be the case. Two conditions are essential to produce the observed results: (1) in order that the water should escape at the mouth of the chamber there must be a pressure within the chamber higher than that in the exhalant passages; (2) in order that water may enter the chamber there must be within it a pressure less than that in the inhalant passages. But the pressure in the inhalant and exhalant passages is presumably the same, at any rate before the current is started, therefore there must be a difference of pressure within the chamber itself, and the less pressure must be round the periphery. Such a distribution of pressures would be set up if each flagellum caused a flow of water directed away from its own cell and towards the centre of the chamber; and this would be true whether the flagellum beats synchronously with its fellows or not.
The comparative study of the canal systems of sponges[[274]] acquires a greater interest in proportion as the hope of correlating modifications with increase of efficiency seems to be realised. In a few main issues this hope may be said to have been realised. The points, so to speak, of a good canal system are (1) high oscular velocity, which ensures rapid removal of waste products to a wholesome distance; (2) a slow current without eddies in the flagellated chambers, to allow of the choanocytes picking up food particles (see below), and moreover to prevent injury to the delicate collars of those cells; (3) a small area of choanocytes, and consequent small expenditure of energy in current production.