REPORT ON THE ANTIPATHARIA. 15 



described as branched Antipathidse having a spinose chitinous axis, presenting no 

 fusions between any parts of the corallum. 



No new species are recorded, and those previously described are allocated to one or 

 other of the above genera, as best could be done from the descriptions available. With 

 the exception of Duchassaing and Michelotti, subsequent investigators have not followed 

 Milne-Edwards in his subdivision of the genus Antipathes, partly on account of the 

 admittedly unsatisfactory state of our knowledge of the group, and partly owing to the 

 fact that it has been generally felt that genera founded on the structure and arrange- 

 ment of the axis alone could have little value in a natural system. Pourtales especially 

 has preferred, pending a fuller study of the polyps themselves, to regard the Antipathidse 

 as consisting of only a single genus, Antipathes, those species having a simple axis 

 being included in a subgenus, Cirrhipathes. 



As no new genera have been proposed since the publication of Milne-Edwards' 

 work, it may be well to consider here the systematic value of the genera therein defined, 

 in the light of the information brought forward in the present monograph. I do not, of 

 course, for a moment suppose that the information which has been obtained concerning 

 the structure of the zooids of some twenty species will prove sufficient for an adequate 

 classification of the group, particularly when it is remembered that we have absolutely no 

 information on the structure of the zooids in the majority of the species described. Still 

 I believe it to be sufficiently complete to be of service in the present instance. In the 

 first place it may be stated generally that although in those forms studied, a similarity 

 in the structure of the zooids is frequently associated with a similarity in the type of 

 branching of the corallum, this is by no means always the case. With regard to the 

 more or less frequent occurrence of fusions between different portions of the corallum, 

 a point which will be discussed in detail later, it may here be stated that the evidence 

 available at present appears to show that such fusions are not constant in all the species 

 of ti genus, and are therefore of no generic value. Indeed amongst the species referable 

 to the Rhipidipathes type of Milne-Edwards, there are at least three well-marked types 

 of polyp. In other cases where the fusions are slight, and more accurately defined as 

 adherences, the feature is more or less accidental, and probably not even of specific 

 value. 



Turning now to the genera adopted by Milne-Edwards, it will be well to consider 

 first those described by Blainville and Gray. 



Cirrhipathes, Blainville. — This genus was constituted by Blainville in consequence of 

 Ellis' observations on the polyp of Antipathes spiralis, Pallas, and for the reception of 

 those forms in which the sclerobasic axis remains simple. At the time Ellis' figures of 

 the polyps of Antipathes spiralis were the only reliable ones extant for any species of 

 Antipathes, and further investigation has shown that the form of polyp assigned to this 

 genus by Blainville is, so far as his definition goes, not confined to it, but is shared by 



