CONGDON. — THE IIYDROIDS OF BERMUDA. 483 



hydrotheca wall to form four deep symmetrical scallops. The edges 

 meet in two deep intersecting ridges, which terminate at the four 

 points between the scallops (Figure 34). The narrow diaphragm con- 

 sists of a fold which encloses a considerable space. 



Colonies arise from creeping stolons, which are as large as a pedicel 

 and not annulated (Figure 33). They grow parallel to each other along 

 the blades of eel grass, sending frequent connecting branches across to 

 neighboring stolons. 



The few well-preserved hydranths are contracted into the lower half 

 of the hydrotheca. They are attached to the diaphragm only, possibly 

 because of unsatisfactory killing. There are about twenty tentacles 

 and the hypostome is rounded. 



The characters of the trophosome are allied to both Campanular- 

 idae and Sertularidae. The occurrence of diaphragm, pedicel, and an 

 ovato-cylindrical annulated hydrotheca suggest the former. Yet we 

 do not find a four-part operculum in that family except in the genus 

 Thyroscyphus, erected by Allman in 1877 for the single species T. 

 ramosm, which is of doubtful affinity (Nutting, :04, p. 10). The 

 jointed stem, shortness of pedicel, and four-part operculum suggest its 

 close relation with the genus Sertularella. If the gonosome is of the 

 Sertularelia type, the existence of a pedicel would be the only character 

 separating it from that genus. 



The genus Sertularella as revised by Hartlaub contains species with 

 a hydrotheca and operculum closely similar to the Bermuda hydroid. 

 The one character in which it differs from the genus is the presence of 

 a pedicel. This difference Allman judged so important in the case 

 of T. ramosus as to demand a new genus. I have accepted his view of 

 the matter. 



The chief differences between the two species, as far as we now 

 know, are these : T. ramosus is larger, has a bordered hydrotheca, and 

 a jointed stem bearing many hydrothecae instead of a short, annulated, 

 unjointed one bearing at most three hydrothecae. 



Genus AGLAOPHENIA Lamouroux, 1816. 

 Aglaophenia minuta Fewkes. Figure 37. 



This form is rather widely distributed. It occurs on the Sargassum, 

 which drifts to Bermuda from the south. Though I did not find the 

 gonosome, the complex trophosome was sufficient for identification. 

 A small nematophore is always present on each side of the axil of a 

 branch. The fact is not mentioned by Fewkes nor in the fuller de- 

 scription by Nutting. 



