[35] MEDUSA FROM THE GULF STREAM. 961 



ceutral region and peripheral part well marked, while iu all in which 

 the marginal zone is present the division between the central disk and 

 the peripheral zone is not difficult to trace. The convex protuberance 

 on the subumbral side of the disk is homologous with a slighter pro- 

 tuberance in several other JEquoridse and morphologically the same 

 structure exists in the genus Orclmtoma and other genera. In Aurelia 

 flavidula, as I have elsewhere pointed out, we have a similar although more 

 angular pyramidal projection from the lower side of the umbrella above 

 the mouth opening. It seems not improbable that this projection in 

 Polycanna, Aurelia and elsewhere is homologous with the gelatinous 

 peduncle of such genera as Liriope, Garmarifia, Geryonia and others. 



The width of the coronal zone is about equal to the radius of the 

 central disk. The coronal walls are thin, crossed by numerous radial 

 chymiferous tubes very closely set together. Margin of the corona very 

 thin. Tentacles not numerous, but long, flexible, with erdarged bases. 

 Otocysts probably exist, but the poor state of ijreservation of the speci- 

 mens does not admit of their examination. 



The sexual glands in a few specimens are still preserved. They are 

 sHall, resembling frilled sacs, each extending the entire length of the 

 radial tubes. There are no rows of tubercles between the chymiferous 

 tubes on the subumbral surface of the umbrella. 



Stomach walls are fastened to the x>eriphery of the central disk at 

 the line of separation of the same from the coronal zone. They consist 

 of a thin, wide, more or less folded membrane, with muscles fastened on 

 one edge to the lower floor of the umbrella, while on the other edges are 

 borne many small tentacular oral bodies arranged in a single row, sep- 

 arated from each other by a considerable space on the lips. These 

 bodies axe smaller than the same structures, the oral tentacles, in other 

 ^quorida; known from American waters. The embryonic nature of 

 these bodies arouses the suspicion that this species may be the young 

 of M. Bairdii, a description of which follows. 



The thickness of the central disk in the specimen, No. 9304, is so 

 great and out of all proportion to that of the corona, while the coronal 

 furrow is in others so well marked, that it may be found that this species 

 maj^ connect the Oraspedote and Acraspedote medusae. We are cer- 

 tainly reminded in the partial differentiation of the central disk and the 

 peripheral corona of the similar relationship in A tolla,Nauphautop sis, and 

 some others. The differences between these genera and the Polycannw, 

 however, are so great that their resemblances cannot be regarded as of 

 morphological value, lor what is known of the development of the Eu- 

 ropean species of Mesonema and others indicate their unquestioned 

 hydroid affinities. It, is, however, still an interesting thing to see in 

 this largest of hydroid gonophores an anatomical separation between 

 the central disk and a peripheral corona closely parallel to what we find 

 in undoubted Acraspeda. 

 H. Mis. GS 61 



