342 P- Kramp. 



features in their reproduction. In such cases Levinsen and Broch 

 seem to me right in laying chief weiglit on the structure of the 

 nutritive individuals. Though it must seem of great importance, 

 whether a species produces free medusae or sessile gonophores, it is 

 yet impossible to use this feature for the separation of the larger 

 groups, as we thus separate species, which are obviouslj' nearly re- 

 lated and belong together naturally (cf. the Fam. Campamilaridae). 

 On the other hand, to exclude the reproductive features entirely, is 

 not an ideal procedure for a systematist who desires to lind a na- 

 tural system, and to justify doing so we must also Ьале more de- 

 tailed evidence from comparative anatomical inл'estigations on the 

 medusae or gonophores of nearly related forms. For the present, 

 however, it seems to me both necessary and defensible, to keep the 

 characteristics of reproduction in the background. We need not go 

 so far as Schneider (51), who deprives the sexual individuals of any 



importance for the systematist, chiefly because " die Geschlechtspro- 



ducte entstehen nicht in der Meduse, sondern im Trophosom; die Meduse 

 ist nichts weiter als ein Apparat, der sie auf eine grosse Fläche ver- 

 theilt" (1. с p. 491). But every "apparatus" must be regarded by the 

 light of descent and the natural relationship, and cannot therefore 

 be altogether deprived of systematic importance. It must be admit- 

 ted nevertheless that the nutritive individuals are of the first im- 

 portance. — A systematist cannot however be consistent. The same 

 characters have not the same importance everywhere; an excellent 

 example of this is given by the families Lafoèidae and Cainpaniilinidae. 

 Levinsen characterizes the last-named family as having an operculum 

 in contrast to the first; as we shall see later, this distinction cannot 

 be maintained. It is scarcely possible to set any definite limits be- 

 tween these two families based only on the trophosomes. And yet 

 there is something which marks olf the Lafoëidae as a naturally con- 

 nected group, and this is the characteristic aggregates of gonothecae 

 and transformed hydrothecae, which were earlier regarded as inde- 

 pendent hydroid colonies (Coppinia and Scapiis), and were first 

 rightly interpreted by Levinsen, 1892 (38). Here we see that the 

 gonosome helps to fix the ])oundaries between two related groups. 

 Whether we should call such groups families or subfamilies is and 

 will always be a matter of opinion. — The presence or absence of 

 an operculum must therefore in this case be kept in the background 

 as a systematic factor; but this by no means lessens the systematic 

 importance which Levinsen ascribes to the operculum in the Sevtii- 

 Uiridae. 



Broch rightly emphasizes the importance of investigations on 

 the variations of the sjjecies, but when he says, "an den anderen 



