56 HYDROIDA 



ing, in larger colonies of Perigonimus, individuals with pseudohydrothecae bulging even more loosely 

 round the shrunken and contracted polyps than that rendered in Kramp's drawing, while, at the 

 same time, in extended polyps in the same colony, a superficial view fails to notice the presence of 

 this formation (Broch 1911, fig. 12). The intermediate state represented by Jaderholm is the com- 

 monest observed in preserved colonies. Still one point of difference might seem to remain, the chiti- 

 nons cup in which the gonophore, in Garvcia groenlandica, is resting when it is about to empty its 

 ripe contents. This cup, or rather this remnant of the flayed-off external periderm cover of the gono- 

 phore, is, however, frequently observed also in the female gonophores of typical Perigonimus roseus, 

 in which hitherto no particular importance has been attached to this character. Accordingly, we have 

 to include Garveia groenlandica as a synonyme under Perigonimus roseus. 



The chief occurrence of the species is attached to the middle and the lower parts of the litoral 

 region of the boreal waters. It occurs rather frequently from Bohuslan as far as Lofoten, generally 

 attached to stalks of Tubularia indivisa. However the species also penetrates far into Arctic waters, 

 and has been recorded from the White Sea and as far to the north as Nova Zembla. It has previ- 

 ously been recorded from Greenland (Jaderholm 1909), where now also Lille Hellefiskebanke, Fiske- 

 nes (West Greenland), and Danmarks Havn (East Greenland) have to be added to its localities 

 (Text-fig. R). 



Family Eudendriidae. 



Hydroids forming colonies, the polyps of which are provided with a single whorl of filiform 

 tentacles. The proboscis is capitate, placed with a narrow base on the broad polyp body above the 

 tentacle whorl; the stinging cells of the tentacles are small and rodshaped. Also larger stinging cells, 

 narrowly oval, occur, particularly in the specific stinging organs of the polyp. The stinging cells of the 

 tentacles are arranged in very distinctive transverse belts. The polyps are wholly naked. The endoderm 

 of the polyp is differentiated into two portions, an oral portion, consisting of small-nucleated, indifferent 

 cells, among which occur some mucous gland cells, and the proper gastral portion. The limit is formed 

 by the tentacle whorl. The tentacles lack a central cavity. The colonies have no calcareous skeleton. 



Nearly all investigators of hydroids have distinguished the Eudendriidae as a family of their 

 own. To this Levinsen (1892) forms an exception, considering the arrangement of the tentacles in 

 a single whorl round the polyp as indicative of so near a relationship to the Bougainvilliidae that he 

 unites the two families into one. However, the Eudendriidae, in their structure of the polyps, show, 

 as compared to other athecate hydroids, such peculiarities that we are forced to distinguish them as 

 a family of their own. In the first place are obvious the broad structure of the polyp body and the 

 capitate proboscis, placed with a narrow base above the well defined tentacle portion. In this cha- 

 racter the polyps bear a strongly marked resemblance to the thecaphore Campanulariidae. This like- 

 ness is the more interesting because, in the more delicate structure of the polyp of the Eudendriidae, 

 several points of resemblance to the thecaphore hydroids are demonstrable. Thus the stinging cells of 

 the tentacles of Eudendriidae, as well as those of the thecaphore hydroids, are arranged in well defined 



