382 W. K. BROOKS ON THE LIFE HISTORY 



The observations here given show the correctness of Balfour's conjecture (65, p. 

 130) that " while the exact mode of formation of the gastro vascular canals of Geryonia 

 has never been worked out, the presence in the adult of hypoblastic lamella 1 , and the 

 mode of formation in medusa-buds, justify us in believing, with the Hertwigs, that 



they are the remnants of a once continuous gastric cavity." 



The metamorphosis of the young medusa has been well described by Fritz Muller and 

 Haeckel. The bell gradually becomes flattened, as shown in fig. 11 and in PI. 42, and 

 the gastrostyle gradually grows down from the apex of the sub-umbrella, carrying with 

 it the stomach and the oral ends of the radial canals, until, in the adult, the mouth and 

 stomach are far below the level of the veil. At various stages in its life the medusa 

 has three sets of tentacles, four in each, or twelve in all. Of these, one set is radial 

 and larval, soon disappearing with the growth of the medusa. The second set, figs. 

 9 and 10, h, next appear, and in some species persist throughout life, while they are 

 absent in the sexually mature medusa' of other species. They are interradial. The 

 third set, fig. 10, g, fig. 11, g, are radial; and are the long tentacles of the adult 

 medusa. 



The primary radial tentacles, figs. 9, 10, 11, are larval organs. They are the first to 

 make their appearance and they are present in very young medusae. They are solid, 

 consisting of a central axis of very large cartilage-like endoderm cells, and a surface 

 layer of ectoderm which is thickened at the tip of the tentacle, to form a knob or bulb, 

 which is crowded with lasso-cells. 



When these tentacles first appear in the larva they are situated at the edge of the 

 peristome, and when this becomes pushed in, to form the sub-umbrella, the tentacles 

 spring from the edge of the umbrella, just outside the velum, and their endoderm is 

 continuous with that of the circular edge of the digestive cavity; but as the medusae 

 grow, they are carried out on to the outer surface of the umbrella, some distance from 

 its edge, as shown in fig. 9. A string of degenerated endoderm cells persists for some 

 time between the base of the tentacle and the circular tube, and thus marks out the 

 line along which the tentacle has migrated. There is also an ectodermal ridge, 

 "Schirmspang," on the surface of the umbrella, running from its free edge to the base 

 of the tentacle. These tentacles drop off before the medusa attains to its full size, 

 and they arc entirely absent in the adult. 



The four primary interradial tentacles are the next to appear, figs. 9, 10, 11, h. They 

 also are solid, but they are distinguished from the primary radials by the fact that the 

 ectoderm of the axial side is thickened to form a number of ridges or incomplete rings, 

 each of which is filled with large, oval lasso-cells. The stiff interradials are usually, 

 carried turned up against the outer surface of the umbrella, with thin rings of lasso- 

 cells facing outwards. They are, at first, situated on the bell margin, but they migrate, 

 like the primary radials, and in the adult they are separated from the bell margin by 

 an interval which is somewhat greater than the diameter of the circular tube, with 

 which a row of degenerated endoderm cells connects the base of each tentacle, and 

 there is also an ectodermal ridge or " Mantelspang" with large lasso-cells running from 

 the axial surface of the base of the tentacle to the bell margin. According to Haeckel, 



