HYDROIDA. BALE. 259 



Considering, now, that notwithstanding differences of 

 opinion as to details, it is generally agreed that the opercular 

 characters must, to a greater or less extent, be utilised in 

 limiting the genera, it appears that there are only two courses 

 logically open to us. The first is to follow Levinsen in 

 abandoning entirely the colonial characters. Then Pasythea 

 quadridentata becomes a Sertularia, and P. hexodon a Thuiaria, 

 while the species ascribed to Selaginopsis and Didyodadium 

 must be divided between Sertularia, Sertularella, Thuiaria, 

 and a genus of Lafoeidae for the inoperculate forms. The 

 other course is to combine the colonial and the zooecial 

 characters, as Stechow has done, but in that case we cannot 

 follow him in retaining the old groups Pasythea, Selaginopsis, 

 and Dictyodadium, each as a single genus. P. hexodon must 

 be separated from P. quadridentata on exactly equal grounds 

 to those on which Thuiaria is separated from Sertularia. 

 Selaginopsis (with Didyodadium) must be split up into several 

 genera, according as the operculum resembles that of Sertu- 

 laria, Sertularella, Thuiaria, and possibly yet other types, 

 while in any case the inoperculate forms must be referred to 

 a separate genus. It is doubtless true that characters which 

 are of generic value in one group of species may be less 

 constant and of minor importance in another, but in the 

 particular case before us it is impossible to maintain that 

 where the opercular characters have a generic value in the 

 biserial forms, the identical characters should have less 

 importance in multiserial forms. 



I do not look forward to any early unanimity on these 

 points, but Levinsen's arguments, and the numerous instances 

 cited by him of transition-forms between the biserial and 

 the multiserial forms, furnish very strong support to his 

 views. Again, in such a form as Pasythea hexodon, how does 

 the arrangement of the hydrothecae in separate companies 

 differ from the arrangement in many species of Thuiaria 

 except that in the former the companies are separated by 

 longer intervals ? And in P. quadridentata we have in the 

 one species every gradation from the typical form to a variety 

 which differs in no respect from a true Sertularia. 



Of course there are in some instances special characters, 

 unconnected either with the hydrothecal arrangement or the 

 opercular structure, which may be regarded as of generic 

 importance, and here again unanimity of opinion is not likely 

 to be soon arrived at. For example, the old genus Didyoda- 

 dium seems to me, as to Levinsen, to consist merely of a 

 few species united by the trivial character of their union into 

 -a network by anastoming stolons, a feature often found in 



