ASTEROIDEA 125 



analogus have broader jaws. Considering the difference in size of the two specimens, the 

 adambulacral armature is remarkably aUke. In validus the second or outer subambu- 

 lacral tubercle (distinct from the marginal series of granules) is not developed, although 

 on the margin of plate there are i or 2 granules larger than the immediately adjacent 

 actinal granules. In analogus these outer marginal granules are subequal to the adjacent 

 actinals. 



The disparity in size of the peripheral granules of actinal plates occurs also in validus. 



In Cladaster rudis VerrilP the only example has R 25 mm., r 12 mm., and 4 distal 

 superomarginals in contact medially. The pedicellariae have narrower jaws of a different 

 form to those of validus and analogus. In rudis there is no specialized naked area on the 

 superomarginals, the granules being scattered. The dark spots shown in Verrill's fig. 2 

 (which might be interpreted as specialized areas) I was unable to observe with a strong 

 lens. The adambulacral armature of rudis lacks the second subambulacral spine, there 

 being, as in validus, 2 enlarged marginal granules just external to the subambulacral spine. 



In C. macrobrachius Clark- from Cape of Good Hope (R 40 mm., r 16 mm.) the 5 

 distal pairs of superomarginals meet medially, and in spite of its smaller size most of the 

 adambulacrals have 3 furrow spines, while the subambulacrals are generally in 3 longi- 

 series of 3 spines each, the aboral of the first series and the median of the outer being 

 really spines — the rest scarcely more than granules. 



Hippasteria imperialis Goto,^ a large species from 640 m. off Misaki, Japan, is possibly 

 a Cladaster but hardly a Hippasteria. 



Genus Hippasteria Gray 

 Doderlein* has shown that Hippasteria is not closely related to Anthenea, with which 

 it was formerly associated in a special family by Perrier, but is a typical Goniasterid near 

 to Calliaster, while Anthenea is an Oreasterid. Verrill^ segregated Hippasteria from 

 Anthenea in 1899 making of it and Cladaster a subfamily Hippasteriinae. Doderlein 

 could have strengthened his case by citing Cladaster, which is obviously a typical Goni- 

 asterid yet very closely related to Hippasteria. Likewise, Hippasteria imperialis Goto,^ 

 possibly a Cladaster or else a new genus, bridges completely any hiatus which may have 

 existed between Hippasteria and thoroughly typical Goniasteridae. It follows therefore 

 that the subfamily Hippasterinae is superfluous. 



Hippasteria falklandica sp.nov. 

 (Plate III, fig. 2; Plate IV, fig. 4) 

 Diagnosis. Resembling Hippasteria heathi Fisher of Alaska, but lacking the promi- 

 nent abactinal and marginal spines of that species and differing in minor details of 

 adambulacral armature and in granulation of marginal plates. Differing from H. 

 phrygiana (and presumably also H. hyadesi) in the different form of the abactinal plates 



1 VerriU, 1899, P- i?^, pl- 28, figs. 2-2C. ' Clark, 1923, p. 268, pi. 13, figs, i, 2. 



3 Goto, 1914, p. 338, pl. 12, figs. 178-193. * Doderlein, 1922, p. 47. 



5 Verrill, 1899, p. 174. * Goto, 1914, p. 338. 



