4 THE VOYAGE OF H.M.S. CHALLENGER 



introduced in its place the more recent name Heliactis, for Sagartidae with numerous 

 large papillae ; although Oken adduces Cereus hellis as the type form, which stands in 

 the same relation to the genus Heliactis. The papillate Sagartidse are of two kinds, the 

 one having a soft surface, while in the other the body-wall is covered as far as its upper 

 edge with a bark-like cuticle which recalls the Phellidae ; it is therefore advantageous to 

 confine to the former the name Heliactis, applied, though unjustifiably, by Andres, and 

 for the latter to restore Cereus, the old designation of Oken, a representative of the 

 newly characterised genus being Cereus spinosus. 



In discussing the families instituted by Andres, we next come to the Paractidse. 

 As I understand the diagnosis given for this family, — " margine tentaculato, non 

 rilevato e privo d' acroragi," — the tentacles spring at the edge where body-waU and 

 oral disc pass into one another, just as is the case both in the Corallimorphidse and 

 Antheomorphidee, which I have described in more detail, and, generally speaking, in 

 such Actinise as are devoid of a circular muscle. But this relation also holds good in 

 Actiniae with a weak sphincter, as, for example, in Anemonia cereus (to which Andres, 

 strange to say, ascribes a "margine rilevato") ; and, finally, in Actiniae, in which the 

 sphincter is developed at some distance outwards from the upper edge of the body-wall. 

 The facts adduced are sufiicicnt to prove that this characteristic is systematically 

 useless ; and in addition to this I insist that the few forms grouped in the family do 

 not appear to agree with the diagnosis. The tentacles of an Anemonia are, according 

 to Andres, formations placed more at the edge than are those of a Paranthus or a 

 Paractinia. On the contrary, the Paractis peruviana, which Andres adduces as the 

 type of the family, seems to me to have no tentacles which would be marginal. Indeed, 

 it agrees so entirely with a Challenger form, Paractis excavata, that I long doubted 

 Avhether it were not right to unite the two. In Paractis excavata, I am certain that a 

 strong mesodermal sphincter is present, and, corresjDonding to this fact, body-wall and 

 o]-al disc are sharply marked off' from each other, whence I conclude that the same holds 

 for Paractis peruviana. Since I have thus good ground for holding unsuitable the 

 mcuhods by which Andres has instituted his family Paractidse, and can, in addition, 

 claim the right of priority, I adhere to the definition which I previously published, 

 leaving only to future investigators to decide upon the advisability of erecting'Actiniae 

 with marginal spherules, sucking -papillae, and papillae into a family separate from the 

 Paractidse (sensu stricto) with smooth body-wall. 



The next family in the system of the Italian naturalist is formed by the Actinidas, 

 and corresponds to the Antheadae and Actinidae of Gosse. I formerly followed Gosse 

 in separating these two families, but had previously maintained that anatomically they 

 are closely related, and should perhaps on that account be united. I have therefore 

 nothing to adduce against this proceeding of Andres, though the detailed investigation 

 of the Actinidae, which I recommended, has not vet been carried out. It is also 



