PART 5 A MONOGRAPH OF THE EXISTING CKINOIDS 49 



genera, in which P3 is enlarged and has to a greater or lesser degree assumed the normal 

 functions of Pj. Of the three remaining genera, Eumetra shows the least modification of 

 the oral pinnules, while in Dorometra there is a progressive reduction of Pj which in such 

 species as D. nana has become, like P,, apparently functionless in con-elation with a 

 high degree of specialization in the cirri and a very small size. 



The most highly specialized genus in the subfamily appears to be Toxometra, in 

 which cirri of the Antedon bifida type, with the characteristic features exaggerated 

 reappear. 



[Note by A.M.C] A re-examination of the unique holotype of Hybometra senta, 

 formerly included in the Antedoninae, has shown that the large conical centrodorsal 

 and arrangement of the cirrus sockets ally it with Poliometra and Caryometra of the 

 subfamily Zenometrinae, rather than with any genus of the Antedoninae. Its curi and 

 Pi are unfortunately unknown. 



In 1938 Dr. H. L. Clark described the genus Monilimetra with four species from 

 north-west Australia. Mr. Austin Clark included aU these in Toxometra in the t3-pe- 

 script of this monograph but left no comparison between MonUimetra and Toxometra or 

 reasons for this move. 



Gislfc, 1955, has pointed out that Antedon bifida and to a lesser extent A. hupferi 

 may have the distal ends of the pinnule segments more or less spinous, which character 

 has, untU now, been thought to be characteristic of the genus Compsometra. He does 

 not beheve, and I concur, that there is enough difference in this or any other character 

 for two genera to be distinguished and also comments that the inclusion by Mr. Clark of 

 the Atlantic species nuttingi in Compsometra breaks down the geographical separation 

 of the Indo-Pacific species of Compsometra from the Atlantic ones of Antedon (though 

 nuttingi appears to be rather aberrant, having different proportions of the first two 

 pinnules — a character which Gislen supports Mr. Clark in beUeving is important — 

 from the species of both Antedon and Compsometra). The name Compsometra therefore 

 becomes relegated to the synonymy of Antedon. 



It seems to me that the generic distinctions in this subfamily are often too artificial. 

 In some cases differential proportions of the pinnules are regarded as diagnostic, else- 

 where they are set aside. In checking the following key against a table of numerical 

 characters derived from the descriptions of the species, some anomaUes were found 

 which necessitated considerable alterations to the key and which suggested that cer- 

 tain of the species could be better placed in other genera. Eumetra aphrodite could 

 well be transferred to Dorometra, as discussed on p. 77, leaving Eumetra monotypic, 

 E. chamberlaini being marked off from the rest of the subfamily by the relatively long 

 cirri equal to nearly half the arm length, with up to 33 segments, the highest numbei 

 recorded in the Antedoninae. Annametra I beheve should also be a monotypic genus. 

 The type species, A. occidentalis from South Africa, is distinguished from all the other 

 species of the subfamily by the very short segments of Pi, which are not longer than 

 wide (see fig. 6, p. 95). Annametra minuta from Japan does approach it in having 

 these segments less than twice as long as wide and relatively short cirrus segments but 

 some other species of Antedoninae are also similar with regard to these characters. 



As for the genera, Argyrometra and Euantedon seem to me to he very poorly 

 differentiated from one another. In 1917 and in this typescript, Mr. A. H. Clark dis- 

 tinguished them by the number of segments in Pj (18 to 21 in Euantedon and 12 or 13 in 

 Argyrometra), the total size (not diagnostic) and the shape of the centrodorsal (conical 



