REPORT ON THE ACTINIARIA. 23 



vi., Communications, p. G9) what was corroborated by Klunzinger (Korallthiere des rotlicn 

 Meeres, Heft i. p. 82, 1877), that the numerous tentacles, which in Discosoma (Verrill 

 and Klunzinger) and in the allied Homactis and Stephanactis (Verrill) are always 

 united in a radial series or a group, are connected with the same radial chamber. The 

 circular muscle seems also wanting in Discosoma, as the animal is not able to draw 

 the wall over the oral disk. We might therefore incorporate the genus Corallimor- 

 phus with Verrill's sub-family the Discostominse, were it not for the difference that 

 the secondary tentacles in Corallimorphus are limited to a single corona, whilst in 

 the Discostomin^ they appear in larger and variable numbers. This greater regularity 

 indicates an essentially higher grade of organisation in the Corallimorphidae. 



We must likewise bear in mind an affinity between the Corallimorphidae and AUmann's 

 genus Corynactis (Ann. Mag. Nat, Hist., ser. i., vol. xvii. p. 417), as in the latter the 

 tentacles end in a roundish head and are partly intermediate, partly marginal. Many 

 might also consider as points of affijiity the facts that in both genera the nematocysts 

 attain an extraordinary size, that both genera recall the skeleton-forming Zoantharia, and 

 that the nature of the mesoderm is the same in both. The cardinal point only remains 

 open to discussion. Are the intermediate tentacles secondary tentacles, which share the 

 intraseptal parts with the marginal tentacles, or have they merely been forced by growth 

 from the periphery towards the centre ? This question cannot be settled by studying 

 either the drawings or the descriptions given by Allmann, Gosse, Klunzinger, and others. 

 Verrill also, who placed the genus Corynactis among the Discosomidse, considered it as 

 probable, but certainly not proved by actual observation that several tentacles are 

 evaginated from each radial chamber. 



Finally, it may not be superfluous to lay stress on this fact, that the double corona of 

 tentacles does not justify us in assuming any connection between the Corallimorphidse and 

 the Cerianthidge, which also have a cii'cle of accessory tentacles in the periphery of the 

 mouth ; for what turns the scale in the definition of the grade of relationship is that the 

 Cerianthidse have not yet attained to the characteristic paired arrangement of the septa. 



As at present there is only one genus in the whole family, it depends upon the 

 degree of importance assigned to the special characters, whether we consider them to 

 be characteristic of the genus merely or of the whole f;xmily. The most important 

 undeniably are the double corona of tentacles, the equal distribution of the reproductive 

 elements, and the absence of the circular muscle, and for this reason I have included 

 these points in the diagnosis of the family. 



Corallimorphus rigidus (PI. U. figs. 1 and 4-6 ; PI. IX. 11, 12 ; PL XII. 1-7). 

 Corallimorphus rigidus, Moseley, 1877, Trans. Liim. Soc, sev. ii., Zool., vol. i. p. 301. 



Number of the intermediate tentacles twenty-four, of the marginal forty-eight. 

 Origins of the septa, in the lower third of the wall and outer third of the pedal disk, 

 shown by swollen thickenings of the supporting ^ilate. 



