with Ohservations on Australian Hydroids. 105 



sin^ all those liydroicls which exhibit traces of a meclusoid 

 structure in the sexual zooids, following to some extent the 

 classification of Weissman. The latter author, however, 

 finding that in some genera the medusoid structure is present 

 in the female gonozooids, but not in the male, and arguing 

 from the evident alliance between the latter and those of 

 many other hydroids, places these also in the medusoid group, 

 and finally includes in the same section all the Calyptoblastic 

 and Gymnoblastic genera, leaving Hydra as the only repre- 

 sentative of the group to which Dr. von Lendenfeld has 

 applied the name Hydropolypinpe. This is on the assump- 

 tion that all forms which exhibit medusoid affinities, or 

 which appear nearly allied to such forms, are descendants of 

 older types which were, in the adult stage, free medusae. 

 Dr. von Lendenfeld dissents from this view, pointing out 

 that as the polyp-form is a more primitive type than the 

 medusa, the genera which exhibit no trace of the latter 

 form, or some of them, are at least as likely to be direct 

 descendants of those primitive organisms as to have retro- 

 graded from intermediate medusoid ancestors. This argu- 

 ment cannot be gainsaid, and it seems to justify Dr. von 

 Lendenfeld in ranking -under Hydropolypina3 all those genera 

 in which no medusoid modification exists. But both Weiss- 

 man and von Lendenfeld agree that in genera where the 

 gonozooids are modified medusae, the species are to be con- 

 sidered as the retrograded descendants of true medusa- 

 bearing species. This appears to me to be an assumption for 

 which the evidence is inadequate, for it is quite possible that 

 some of these species may be in process of gradual develop- 

 ment toivards the medusoid form, or arrested at particular 

 stages of such a process, rather than retrograding from a 

 more specialised condition. 



Although it may be conceded that the sub-orders Hydro- 

 polypinge and Hydromedusinse are truly natural groups, the 

 fact that in some genera it is difficult to decide whether the 

 gonozooids have medusoid affinities or not is an obstacle to 

 the general use of such a classification, at least until our 

 knowledge of the histology of the various genera shall be 

 more complete. 



As already mentioned, a number of species have to be 

 added to Dr. von Lendenfeld's list, and a few have to be 

 deleted — for example, Halicoimopsis rostratum, which is the 

 same as H. avicularis, and Sdaginopsis onirahilis, which 

 has evidently been added to the list under a misapprehension. 



