IIYDROIDA II 



group may be supposed to originate from Hebellina; there is, however, much which would seem to 

 suggest closer relationship with Eudendriidoe. I have in a previous work (1909 p. 132) drawn atten- 

 tion to several features pointing in this direction; Kiihn, (1913 p. 246) endeavours to disprove the 

 close resemblance, while on the other hand Nutting, (1915 p. 20) notes further similarities, and, like 

 Stechow, (1913 p. 22) comes to the conclusion that a close relationship between Proboscoida and 

 Eudendriidoe is not to be denied. I would here merely mention one point, which Kiihn regards as of 

 great importance, but which Nutting has not subjected to closer consideration. Kiihn finds that there 

 is a difference of principle in the gonosomes. In this connection, several of the gonangia of the 

 Canipaiiulanidcr will be found of considerable interest. In the athecates, we repeatedly find that single 

 gonophores, as in Bougainvillea, Hydractinia, and Eudendrium, collect on the stalks of some few 

 hydranths; in several species moreover, we find a reduction of the terminal polyp, so that the whole 

 of the gouophoriferous complex is here transformed into a blastostyle. As a matter of fact, we have 

 in such cases to deal with gonangia aggregates, differing only in gonotheca formation from the gono- 

 some in species such as for instance Laomedea _flexuosa. The development of the gonangia here 

 shows us a reduced terminal polyp, which has now formed an "Endplatte", on the stalks of which 

 gonophores appear. If at the same time we imagine that the gonophores do not penetrate the peri- 

 sarc, but that the latter expands instead into a protecting sheath, we have then the gonotheca in 

 its typical form for Camfanulariidce. It may also be imagined as arising by formation of gonophores 

 on the basal parts of the hydranth, within the hydrotheca, when, on the one hand, the hydranth will 

 thereby be reduced to an "Eudplatte", while on the other hand, hypertrophy of the hydrotheca will 

 set in. Either of these alternatives may be considered as the possible starting point. I do not insist 

 that this explanation of the origin of the gonosome in the group is the correct one; there is, however, 

 nothing to disprove it in the results of investigations made up to date, and it will, if confirmed, alto- 

 gether disprove the existence of a difference of principle in the gonosomes, as maintained by Kiihn. 

 This proof is thus likewise inadmissible as finally disposing of the supposition of a closer relationship 

 between Proboscoida and Eudendriidoe. 



If we now endeavour, as in the case of the athecates, to summarise the elucidated features 

 and previous views in a key for determination, we must at the outset point out that in this instance 

 such a method cannot so easily be employed without reserve, as the transitions in the thecaphores 

 are more gradual than in the athecates. An endeavour may nevertheless be made to draw up a scheme 

 of the nature indicated. 



I. Polyps with conical, pointed oral part. 



A. The gastral endoderm uniformly developed; polyps and hydrotheca: constructed according to a 

 radially symmetric ground plan. (Family series Hebellina). 

 I. Hydrothecse without opercula or with a primitive closing apparatus formed from the thinner 



distal part of one side of the hydrotheca. Family Lafoeidcc. 

 II. Hvdrothecse with roof-shaped or conical opercula of composite structure. Family Campaimlinid® 

 B. Gastral endoderm differentiated in heterogeneous parts. 



I. Polyps with fore-stomach and digestive basal stomach parts. (Family series Haleciinai 



