THE BRITISH SECONDARY ROCKS. 41 



metra, while I do not think that Semper's Oj>7iiocrinus, with only five 

 arms, can be regarded as more than a subgenus of Antedon, with 

 which it is similar in all respects, excepting the absence of division 

 in its rays*. The same variation occurs in the ten-rayed Promacho- 

 crinuSy one species having twenty arms and others but ten, as the rays 

 do not divide. The characters of the centrodorsal piece and calyx of 

 Ophiocrinus are essentially those of an Antedon^ with which type it 

 also agrees in the central position of its mouth and the absence of a 

 terminal comb on the oral pinnules f. 



Actinometra differs from Antedon in several important characters. 

 The mouth is excentric, the ambulacra unequal, and the oral pinnules 

 provided with a terminal comb. These are not characters, however, 

 which could be of any palseontological value ; but many such are to 

 be found in the peculiarities of its calyx. I have already referred 

 to this subject elsewhere J, though not systematically enough for 

 palaeontological purposes. The following descriptions of the centro- 

 dorsal piece and calyx of Antedon and Actinometra are based upon a 

 personal examination of the external characters of over 200 species of 

 recent Comatulce^ and upon close comparison of the dissected calices of 

 eighteen species of Antedon and eleven oi Actinometra. I trust, there- 

 fore, that they may be regarded as having some systematic value. 



The centrodorsal piece of Antedon is extremely variable in its 

 appearance. It may have the shape of a shallow basin or a hemi- 

 sphere, either complete or flattened at the pole ; or it may be conical 

 either perfectly so or more or less truncated ; or, lastly, it maybe more 

 or less distinctly columnar. Occasionally it is a thick disk, almost 

 thick enough to be called columnar, with well-marked upright sides, 

 to which the cirrhus-sockets are limited, the whole (or nearly the 

 whole) of the dorsal surface being free from them. This is the case, 

 for example, in A. hrasilie^isis and A. macrocnema. 



As a rule, there are at least two and, generally, three or more 

 rows of cirrhus-sockets, alternating in some species (PL Y. figs. 4, 5), 

 but vertically above one another in others ; while though there may 

 be a large central space free from them, it nearly always shows 

 traces of partially obliterated sockets, which are rarely found in this 

 position in Actinometra (PL Y. figs. 6, 8). 



The outer faces of the radials of Antedon are always much inclined 

 to the vertical axis of the calyx (PL Y. fig. 4). They are usually 

 much wider at their dorsal than at their ventral ends, having a 



* Schluter (p. 40) seems to be sceptical about Ophiocrimis with its " quite un- 

 divided arms which are said to start directly from the centrodorsal, a fact that 

 rather wants a closer examination." As all the recent examples of this type 

 which are known to science (including Semper's original specimen) are in my 

 hands at present, I can speak positively as to its characters. De Loriol's dis- 

 covery of a fossil species in the Urgonian of Switzerland (Denkschr. d. allg. 

 schvreiz. Gesellsch. f. d. ges. Naturw. Bd. 23, Zurich, 1869, pp. 57-59) is of 

 considerable interest. 



t In one of the ' Challenger ' Antedons only four rays divide out of the five, 

 the fifth remaining simple and undivided, as is the case with all the rays of 

 Ophiocrinus. This is a strong argument against the separation of Ophiocrinus 

 from Antedon as a distinct generic type. 



\ ''Actinometra," op. cit. pp. 61, 76, 81-84. 



