REPORT ON THE ACTINIARIA. 3 



by Andres must be comprehended in the four divisions, Edwardsiaj, Ilcxactiniae, 

 Zoantheae, and Cerianthese, and accordingly hold to the systematic classification which 

 I have published. The groups of Paractinise and Monaulese are in all respects natural, 

 and would also certainly be retained by Andres had representatives of them been 

 known to him. 



Even greater discordance than that of which I have hitherto spoken, between the 

 classifications of Actiniae followed by Andres and myself, presents itself \^-hen the 

 determination and nomenclature of families and genera are regarded. Independently 

 of each other, and from different standpoints, we have taken in hand a systematic 

 revision of Actiniae : Andres starting with the advantage of a richer material, and 

 studying species with which earlier publications are especially concerned, and which he 

 could command in a living condition ; while my qualification for a systematic classifica- 

 tion was that afforded by close anatomical investigation, namely, that I relied for 

 systematic characteristics upon such weighty differences as the structure of the 

 sphincter, the arrangement of the mesenteries, the structure of the musculature and of 

 the oral disc, etc., points which Andres has, hitherto at any rate, entirely left out of 

 consideration. Thus it has resulted that in the determination of families and genera, 

 and also in the value assigned to existing names, we have in many cases taken up a 

 totally different attitude ; and as, in consequence of this, no inconsiderable confusion 

 has arisen in the method of diagnosis, I hold it advisable to inquire critically what 

 must be retained of the system of the Italian observer. 



Of least importance are our differences of opinion relating to those Actiniae which 

 possess acontia. Andres has here adopted the separation, instituted by VerriU, into 

 Sagartidae and Phellidae. Having regard to his wider acquaintance with the species, I 

 agree with him in accepting as a distinctive character the chitinous covering extending 

 over two-thirds of the body-wall ; and for clearer characterisation of both families, the 

 following marks not mentioned by Andres should be included in the diagnosis, — a 

 mesodermal sphincter, and a differentiation of the mesenteries into sterile complete 

 primary mesenteries, and incomplete secondary mesenteries provided with generative 

 organs. Of the Challenger Actiniae, there would belong to the Phellidae only Phdlia 

 IKctinata; to the Sagartidae, Sagartia sp., Cereus spinosus, Calliactis polypus, 

 Bunodes minuta. Of these, the two latter require an alteration of name ; Calliactis 

 polypus must be termed Adamsia polypus,^ and Bimodes minuta be known as Cylista 

 minuta, since it has been shown by Andres that the typical Bunodes possesses no 

 acontia, and therefore cannot belong to the Sagartidae. 



Andres has incorrectly allowed the generic name Cereus (Okcn) to drop, and has 



1 The specific name Rondeletii has been wantonly substituted by Andres for the older poh/pu.i, the former being 

 used for the first time by delle Chiaje in 1825, wl.ile the latter was already instituted by ForskSl in 1775. Milne- 

 Edwards is therefore correct iu calling the animal Adamsia polypus. 



