HYDROIDA II j(.j 



enormous, and makes itself felt among the mednsfc to the same degree as among other planktonic 

 organisms. Rut the variation of single species under different conditions has been even less studied 

 in the case of the medusce than in that of the polyp generation, and our ignorance as to these and 

 similar phenomena has given rise to a medusa system in which phylogeny and biological con- 

 vergency or divergency have been intermingled to a tangled confusion. A very fine proof of this 

 is afforded by the fact that the Canipanuliuidcp have their medusae distributed throughout almost 

 all the families of LeptomedusK, and yet we cannot, from this, proceed to split up the polyp family 

 in question into a similar series of new families. The question as to a division into genera on the basis of 

 gonophore types should be decided by a reference to the genus Hydractiiiia. Following this principle 

 to its logical conclusion, with the Ccuiipaiiitlariidcr^ we should have to make generic distinction be- 

 tween the male and the female of LaoDiedea flexiiosa Hincks, where, as we know, the female gono- 

 phores are heteromedusoid, the males styloid. But if this is out of the question, what gonophore type 

 can we consider as being generically adequate.^ The heteromedusoids are in reality the only ones 

 which differ in principle from all remaining types, which, since the discovery of the reduced crypto- 

 medusoid in Gramviaria abietina (M. Sars) form a finite series. — In Caiiipannlaria integra (Macgilli- 

 vray) \Ca)iipaiiularia coii/pressa Clark | we find, during one part of the breeding season free medusce, 

 at another, sessile eumedusoids; there is thus no clearer limit between free medusae and sessile gono- 

 phores in Caiiipaunhin'idw than in several other families. The only type of gonophore which appears 

 to be of any value must then be that of Gouotliyrcca. This is a highly interesting biological pheno- 

 menon; phylogenetically speaking, it is a typical cryptomedusoid, (cf. Kiihn 1913 p. 1S7) differing but 

 little from the gonophores in Cladocoryuf. Its systematic \-alue really lies in the fact that in certain 

 forms, it is thrust out from the narrower gonotheca without breaking loose. From a biological point 

 of view, this is a most interesting phenomenon, and we cannot but call to mind the gonophores in 

 certain Tubnlaria species, where the larvae are likewise developed in the sessile, medusa-like gono- 

 phore; it shows how nature here, in another way, has arrived at the same result as in several of the 

 Dipliasia species, in protecting the larvte during their period of development. Such care of the }'oung 

 occurs, of course, elsewhere in tlie animal kingdom, but in all other cases, it is regarded as a biological 

 phenomenon, which from a phylogenetic — and thus also from a systematic — point of view is of quite 

 subordinate importance. 



It may also be worth while to glance at one of the most interesting genera, Orthopyxis L. Aga.s- 

 siz, which Nutting (1915 p. 63) again seeks to revive by the .side of Clytia and Campanularia, all 

 on the basis of the gonophores. In his generic diagnosis. Nutting refers to the Medusae of Ortho- 

 pyxis as "without tentacles or manubrium", and under the heading of Orfliopyxis coinprcssa (I.e. p. 66) 

 he states that Torrey (1902) gives for this species an illustration of "a medusa with 4 tentacles just 

 escaping from the gonangium" ; under Orthopyxis cvcrta, again (1. c. p. 68) he informs us that "the 

 female, at least, contain meduste which eject their ova into an acrocyst without liberating the medusae". 

 Nutting has here given us a striking illustration of the value both of the genus itself and of the 

 characters on which it is founded. The genus in question is really a connecting link, standing as it 

 does with one foot, so to speak, in each of the other biological groups which we find represented in 

 the family Canipiuinlariidcc. 



