THE CAMPANULARID^ AND THE BONNEVIELLID^. 17 



at all) suspended beneatli the radial canals of which they appear to be outgrowths or diverticula. 

 They are therefore endo dermal in origin. 



All medusae produced by hydroid colonies have a double nerve ring, one above and one 

 below the velum, and are thus in close connection with the otocysts and sense cells. The velum 

 being the principal organ of locomotion, these nerve rings have doubtless been developed largely 

 in connection with this function, and we therefore find a much more liighly developed nervous 

 system in the medusas than has thus far been found in the colonial forms of hydroida. In com- 

 parison with the higher metazoa, however, the nervous equipment of the medusae is exceedingly 

 primitive. 



The tentacles are confined to the margin, never being found in connection with the manu- 

 brium, as in many medusae produced by the gymnoblastic colonies. They are all of much the 

 same structure, being soUd, with a core of endoderm enveloped by a structureless stutzlamella, 

 and this latter, in turn, being inclosed in a cyUnder of ectodermal cells, among which are em- 

 bedded many cnidoblasts containing nematocysts. 



Muscle cells lie just outside of the stutzlamella or supporting layer. The tentacles vary 

 in number from 4 to 24, the former number being found in Clytia and the latter in several full- 

 grown medusse of the genus Ohelia. Torrey has described a species, Ohelia purpurea, which has 

 from 1.10 to 160 tentacles. As has been said before, the genus Orthopyxis is ^vithout tentacles. 



Sense bulbs appear as swellings on the tentacle bases in the genus Clytia. None of these 

 contain definite eye spots, so far as I know, although they sometimes do contain a rather 

 definite aggregation of pigment granules, the color being brown in C. johnstoni and green in 

 C. noliformis, according to Brooks. These pigment spots have usually been associated with a 

 rudimentary sense of sight, or rather of light detection. 



Nematocysts are found rather abundantly scattered over the surface of the bell of Clytia 

 johnstoni. 



The alternation of generation among the hydroid medusse has been discussed at length by 

 several writers, perhaps the most extended treatment of the subject being that of Allman.^ The 

 whole subject, however, needs a carefid reexamination, and the writer hopes to discuss the mat- 

 ter in connection with a futiu-e part of this work. It will suffice for the present to indicate the 

 belief that the fixed hydranth form antedated the medusa form in the phylogeny of the group. 

 The radial symmetry of the medusse indicates a fixed origin, and, wliile there are cases in which 

 actinides or hydranthlike forms develop into medusae {Cunoctantha octonaria), there seems to 

 be no well demonstrated case in which the medusa develops into a hydranth. 



So far as the Campaiiularidse are concerned, it seems evident that the medusse are function- 

 ally but motile gonophores or mechanical distributors of the sexual elements, and that their 

 use in the economy of the species is thus explained. How it comes, however, that such hydroid 

 colonies as Ohelia and Campanularia, practically identical as to their trophosomes, should 

 differ so remarkably in their gonosomes — the one producing typically free medusse and the other 

 producing gonophores which produce planulse direct — is more than our present knowledge 

 enables us to explain. 



If the gonophore is but a degraded medusa, as is claimed by many writers (and so ex- 

 plained in the present work), why is it that two genera living under practically identical condi- 

 tions and having practically identical trophosomes in some cases should present in the one case 

 typical functional medusse and in the other the functionless degraded .medusse known as gono- 

 phores ? 



We become still more bewildered when we find it reported that a single species {Orthopyxis 

 caliculata) has been found under certain conditions to produce free medusse {Agastra) while 

 under other conditions it produces sessile gonophores.^ 



Is it possible that genera arise in such a way ? 



' Gymnoblastic Hydroids, 1871, pp. 101-110. 



2 Giard, Compt. Rend. Soc. Biol., Paris, 1898, p. 17. 



