1894. PROCEEDINGS OF THE NATIONAL MUSEUM. 257 



inconspicuous. The small, conical, marginal spine on the upper edge 

 of each of the sui»erior plates is relatively shorter and stouter in the large 

 specimens than in the smaller ones; sometimes there are two of them 

 side by side. Taken at many stations, in 705 to 1,407 fathoms. 



Family P E N T A a o N A s T E R i D ^, Perrier.* 



PAKAGONASTER FORMOSUS, Y e r r i 1 1 . 



Archaster formosKs, Verrill, Amer. Joiirn. Science, xxviii, p. 383, 1884. 

 fParagonaster ciiVndratus, Sladen, op. cit., p. 314, pi. 51, figs. 3, 4 ; pi. r>3, tigs. 3, 4, 

 1889. 



Our species appears to be very closely allied to the form well described 

 and figured by Sladen from off the Cape Verde Islands. Our species 

 appears to have the adambulacral j)lates more salient and angular on 

 the furrow-margin, and the notches between them deeper; the furrow- 

 spines appear to be more slender and form a more strongly curved or 

 angular group, which is continued by three to five shorter ones in a 

 fasciole-like row on the proximal and distal edges of the plates; there 

 are about five on the furrow-edge proper; the spines on the actinal 

 surface are more elongated and more regularly stellate, with a longer 

 one in tho middle of the group. 



The spinules of the lower marginal i)lates have the same arrangement 

 as in Sladen's species, but are slightly more slender and acute than 

 shown in his figure; of the larger median series there are usually two 

 or three irregular, indefinite rows in the larger specimens, instead of a 

 single definite row. These differences are, however, so slight that the 

 two forms may eventually prove to be the same species. 



Narrow, imperfect fascioles occur between the marginal and actinal 

 plates in our species. 



The distinction between Faragonaster and Fsendarchaster seems to 

 me very slight, depending almost entirely upon the narrow abactinal 

 area of the rays in the former. 



Tlws species was taken at several stations in 1,390 to 2,031 fathoms. 



I S A s T E R , new genus. 



It seems necessary to institute a new generic group for the elegant 

 starfish formerly described by me under the name of Archaster bairdii. 

 It cannot be i)laced in any of the numerous genera proposed by Sladen 

 without changing the definitions considerably. It appears to be most 

 nearly allied to Mediasfcr, P<(ragonaster, and Nymphaster. It might be 

 defined as a Nymphaster with broad rays having wide abactinal areas. 



The form is stellate, with a rather broad disk and tapering rays, having 

 rather wide abactinal areas. The margiiuil plates are well developed 

 in both series, but the upper ones are flattened above, or bevelled, and 

 do not form a wide margin on the disk; they are all paired, aud those 

 of the two series are nearly opposite each other. They are everywhere 



~ (roniasferidw is an earlier ami better name for thi.s group. 



I'roc. N. M. 94 17 



