214 THE AUSTRALIAN HYDROMEDUS-E, 



These rudimentary Medusae, produced, of course, in the 

 ordinary course of natural selection, show that it is in many cases 

 advantageous for the species not to have free swimming stages. 

 In consequence of this the Medusae may even become more 

 rudimentary, and finally remain attached to the Hydroid colony, 

 where they mature the sexual products within them (Tabularia.) 

 When ijhey have once become so far rudimentary, their parts, 

 particularly the locomotive organ, the Umbrella will be obliterated, 

 unless it is utilised for some other purpose (Weismann, Dohrn), 

 and we shall finally have a Medusoid Gonophor before us. 

 According to Weismann all Gonophores are rudimentary Medusae 

 as stated above. I, however, do not wholly agree with Weismann 

 on this point, although I gladly admit that this is quite possible. 



I place all Hydroid colonies which produce Medusas, and also 

 those which possess Gonophores of a doubtless medusoid origin in 

 this group, which I designate as the second Suborder of the 

 Hydromedusae, the Hydromedusinae. I accordingly place the 

 Anthomedusidae (Anthomedusse Haeckel), the Tubularidae (com- 

 prising Coryne, Myriothela, Stylactis, Tubularia, Pennaria and 

 related genera), the Leptomedusidae (Leptomedusae Haeckel), and 

 those Campanularidae which possess decidedly medusoid Gono- 

 phores as Gonothyrea and Halecium in this suborder. 



In the first of these families, the Medusae bud on all the Zoids 

 and become free, in the second they also bud on all Polypes of the 

 stock, but are more or less rudimentary. In the third and fourth 

 family the Medusae bud on differentiated Polypes, Blastostyls 

 becoming free in the one and rudimentary in the other. 



Some Hydroid colonies producing Medusae, which belong to 

 Haeckel's Margelidae, are so different from the others that I place 

 them in accordance with Glaus in a separate family the 

 Hydractinidae, which evidently connects the calcareous Hydro- 

 corollinae with the chitinous Hydromedusinae. 



The Hydrocorallinae themselves, which I reckon as the third 

 suborder, also produce Medusae, or at least some of them do, but 

 the calcareous skeleton seems to me to be so important a 

 characteristic, because it points to fundamental chemical differences 



