10 



AMEBICAN HYDROIDS. 



and cites the branching forms Campanularia gracilis and Ohelia dicJiotoma, in both of which he 

 finds the cliai^hragm distinctly of two parts. 



Mr. Paarmann's manuscript has nevei' been published, and he has placed it in my hands to 

 quote or use as I see fit. From it I quote the following: 



At a little distance from the base of the hydrotheca is a cliitinous shelf upon which the hydranth rests. This is 

 the diaphragm. It may be either a disk-shaped or funnel-shaped membrane of extreme tenuity, or a complex structure 

 composed of a ring-like portion adjoining the hydrothecal wall and a funnel-shaped structui-e attached to the edge 

 of this "ring." The first I propose to call a simple diaphragm and the second a complex diaphragm. The simple 

 diaphragm can with difficulty be seen in optical section of the hydrotheca, while the complex diaphragm is plainly 

 distinguishable Avithout sectioning (figs. 24-44). 



Of 12 species of Campanularidse sectioned and found to have a simple diaphragm, 10 were 

 branched^ and 2 {Clytia hicopJiora and Ohelia longissima) were simple or fascicled. Of 7 

 species showmg a complex diaphragm, 2 {Ohelia geniculata and 0. Jlahellata) were branched 

 and 5 were simple or fascicled.^ We fiad, then, so far as these species are concerned, that 

 Levinsen's rule does not work iti 4 out of 19 cases. In other words it fails in about one-fifth 

 of the cases iavestigated. Tliis criterion can, therefore, not be regarded as a satisfactory one 

 for dividiag the Campanularidse into generic groups. 



ILLUSTEATIONS OF COMPLEX DIAPHRAGMS OF THE C AMPANXJL AEID.E . (After Paarmann, MS.) 



Fig. 37. — Clytia johnstoni. 



Fig. 38. — Campanularia vertidllata. 



Fig. 39. — Campanularia hincksii. 

 Fig. 40. — Obelia geniculata. 



r, "ring"; 



Fig. il.— Obelia flabellata. 

 Fig. 42. — Clytia noliformis. 

 "membrane". 



Fig. 43.—Campanulariaverticillata. • 

 Fig. ii.—Clytii 



One very striking failure of this criterion was found by the present writer in trying to 

 differentiate the species Clytia hicofhora Agassiz (fig. 24) and C. johnstoni (fig. 37). 



No one, I think, would consider these two species as generically distinct. Indeed most 

 recent writers and earlier, Hincks, consider them identical. They certainly agree in many 

 details of both trophosome and gonosome. But m sectioning the hydrothecse it was found 

 that C. hicophora has a simple diaphragm, while C. johnstoni has a diaphragm that is distinctly 

 complex. A structure that varies in species so nearly identical in nearly every other character 

 seems to the writer to be an entirely insufficient character upon which to base generic distmctions. 



The hydranth.— The hydranth shows great uniformity of structure throughout the Cam- 

 panularidse. It alone of all the calypteroblastic forms shows the so-called "trumpet-shaped" 

 proboscis formed by an apparent deep constriction of the hydranth wall above the insertion of the 



' Campanularia calceolifera, C. angulata, C. flexuosa, C. neglecta, Obelia amphora, 0. commissuralis, 0. dichotoma,' 

 0. hyalina, Gonothyrxa loveni, and gelatinosa obelaria. 



' Campanularia vertidllata, C. hincksii, Clytia johnstoni, C. noliformis, and C. sargassicola. 



