Sponges are propagated twice in the year by minute ciliated globules of sarcode, detached from the interior of the aquiferous canals, which swim like zoospores to a distance, come to rest, and lay the foundation of new sponges. The little yellow eggs of Halichondria panicea are lodged in the interstices between the interior canals; when mature, they are oval and covered with cilia, and are carried out by the currents; and after swimming about for some days fix on a solid object, become covered with bristles, spread out into a transparent film, charged with contractile vesicles of different sizes in all degrees of dilatation and contraction, as well as with sponge ovules. Spicules are developed at the same time, and these films ultimately become young sponges, and if two happen to meet they unite and are soldered together.[15] Besides eggs, larger bodies covered with radiating spicules are produced, containing granular particles of sarcode, each of which when set free by the rupture of the envelope, becomes an Amœba-like creature, and ultimately a sponge. Fresh-water sponges are sometimes branched, and sometimes spread over stones, wood, and other substances; and one species covers an earthy mass some inches thick formed by its own decayed matter. The skeleton of such species as have one, consists of bundles of siliceous spicules, held together and mixed with groups of needles, the rods of which project through the surface of the sponge and render it spinous. The motions in the gelatinous sarcode mass are the most remarkable feature in the fresh-water sponges, which all belong to the genus Spongilla. Mr. Carter observed that portions of the surface of some individuals of the Spongilla fluviatilis in his aquarium had long cilia by means of which they rapidly changed their places during the spring, but when winter came they emitted processes from such parts of their surfaces as were free from cilia and retracted them again just like Amœbæ. These portions often had cells, and when the Amœba-like motions ceased, a nucleus and nucleolus appeared within them, and at last the whole gelatinous sarcode mass consisted of these cells or globules. Some had no nucleus, but were filled with green or colourless granules. At certain seasons of the year, whatever the form of the fresh-water sponges may be, a multitude of minute hard yellow bodies are produced in their deeper parts. They consist of a tough coat containing radiating spicules like a pair of spoked wheels united by an axle with a pore in its surface. Within this last there is a mass of motionless granular cells, and when put into water the cells come out at the pore and give rise to new sponges.