HYDROIDA II 



The diagnosis of the family as here given excludes the two bilaterally built genera Siliciilaria 

 and Eucopella ; these two genera should, partly on account of their being bilateral, partly also from their 

 peculiar, almost leathery hydrothecse, and the small size of the latter, which can only accomodate 

 the basal part of the contracted polyp, be separated off as a distinct family Stliculariidce, as I have 

 pointed out already in a former work (1909). Nutting, in his latest work on Campanulariidce (1915) 

 still treats the two families as one, though he gives no reason for so doing 1 . 



The Campanulariidce, as here defined, make up a well marked family which can hardly be the 

 subject of much dispute; the generic sub-division of the family, however, is a highly debatable question, 

 and a study of the literature thereupon affords an interesting example of the extent to which, even 

 at the present day, it is found "convenient" (i. e. less troublesome), or "advisable" to persevere in the 

 adherence to earlier tradition, or acceptance of biological phenomena, in dealing with the great group 

 of the Hydroids. This is in fact the obstacle which for years has barred our progress towards an 

 understanding of the group in question, its development, and the various degrees of relationship be- 

 tween its subdivisions. - - I have in a previous work (1909) explained at length why, in the northern 

 waters, it is only permissible to recognise two genera of Campanulariidce from a phylogenetic point 

 of view, to wit, Campanularia and Laomedea. The description of Laomedea sargassi (Broch 1913 p. 

 13) does not weaken this standpoint. Nutting, however, opines (1915 p. 24) that "while there may 

 be considerable argument on theoretical grounds for such a course, it undoubtedly leads to unneces- 

 sary and practically insurmountable difficulties when large numbers of species are to be handled and 

 described". The correctness of this view is not immediately obvious. In the first place, it is hard to 

 see why the same characters should be easier to handle as generic than as specific distinctions; fur- 

 thermore, it should be borne in mind that the pursuit of science is not an armchair occupation. In 

 comparison with large genera of other groups the pennatulid genus Pteroeides, for instance, or the 

 cirriped genus Balanus or even with the hydroid geim^PJumularia, the species with which we are here 

 concerned are neither more numerous nor more difficult of distinction; yet it has not hitherto been found 

 necessary to break up one genus into several merely because it comprised a large number of species, 

 or because these were difficult to distinguish one from another; any such subdivision should be based 

 upon grounds of far greater scientific importance. The "insurmountable" difficulties in connection 

 with the Campanulariidce arise from the enormous number of badly described species which have been 

 established often, moreover, on the basis of inadequate investigations with regard to the study of 

 variations. The result is a state of things which absolutely precludes a thorough biogeographical sur- 

 vey. And this is just where the recognition of phylogeny as fundamentum divisionis would help us 

 to clear matters up; from a biogeographical point of view, co-operation between phylogeny and biology 

 is of the highest interest, and by giving phylogeny the precedence, with biology in the second place, 

 we should obtain a sound basis and definite lines on which to set about a rational reconstruction of 

 the chaos at present existing, particularly with regard to the Campamilariidce. 



We have, as a matter of fact, to face the simple question, whether biology 

 or phylogeny is to be regarded'as the fundamental principle for systematics. The 

 medusa system, in its present state, cannot help us in dealing with the thecaphore hydroids. The 

 plasticity of the pelagic organism under the influence of physical factors in its environment is 



1 Nutting still writes "Campanularid<&'. 



