The formation of the sexual Medusa form takes place by a kind of strobilization of the body of the fixed Scyphistoma. A series of transverse constrictions becomes formed round the body below the mouth, dividing it up into corresponding rings, each of which eventually gives rise to a Medusa known as an Ephyra (fig. 85). In each of these rings is a dilation of the stomach, and a section of each of the four rudimentary mesenteries described in connection with the development of the Scyphistoma. As the constrictions become deeper the segments of the body between them become disc-like, and their edges are produced into eight lobes containing prolongations of the gastric cavity (fig. 85 C). The lower surface of each disc, which forms the future aboral surface of the Medusa, becomes convex, in part owing to the development of gelatinous tissue. On the opposite surface a muscular layer becomes developed. During the above process the body of the Scyphistoma gradually grows in length and continues to be segmented, so that a series of Ephyræ are uninterruptedly formed, of which those near the base are the youngest. The original terminal ring of tentacles of the Scyphistoma gradually atrophies. In the further development of the Ephyræ each of their eight lobes becomes bifid at its extremity. As the Ephyræ successively reach this condition they become detached, and by a series of remarkable changes, amounting almost to a metamorphosis, and accompanied by an enormous growth in size, reach the adult condition. The alternation of generations in the Acraspeda cannot be quite so simply explained as in the Hydromedusæ, though the principle is probably the same in the two cases. Actinozoa. Amongst the Actinozoa there occurs in Fungia a peculiar process which is, as shewn by Semper (171), in many ways analogous to alternations of generations[88]. From the larva a nurse-stock is developed, at the end of which a cup-like coral resembling the adult is formed as a bud. The bud becomes detached and then gives rise to a permanent sexual Fungia. From the nurse-stock there is formed however a fresh bud at the centre of the scar left on the detachment of the old one. The fresh bud eventually becomes separated from the nurse-stock leaving a small portion of its stem behind; each succeeding bud similarly leaves a small portion of its stem, so that the nurse-stock eventually acquires a jointed appearance. In the above process we clearly have, as in the Hydromedusæ, a non-sexual form--the nurse-stock--produced directly from the larva, giving rise by budding to a sexual form; all the conditions of an alternation of generations are therefore fulfilled. It seems however possible that the nurse-stock itself may eventually become sexual. [88] Vide also Moseley. _Notes by a Naturalist of the Challenger_, pp. 524 and 525. BIBLIOGRAPHY. _Coelenterata. General._