EASTERN TROPICAL PACIFIC ASTEROIDEA. 101 



E.' Armature of actinolatcral plates with a central, slender spine longer than its fellows. 

 F. Fourth actinolateral scries of plates present; suiieromarginal spines not much more 



slender than carinal spines; spinclets scarcely fleshy ftiUjnis} 



FF. Fourth actinolateral series rudimentary; superomarginal spines decidedly more slender 



than those on carinals; spinclets fleshier aclinodes.' 



EE.' Armature of actinolateral plates with a trio of subequal spines longer than the others. 



trispinosus. 

 DD.' Carinal plates as long as wide or longer; superomarginal spines usually well-developed and 



often forming conspicuous series diomedeae. 



BB. Superomarginal plates unarmed. 

 C Carinal plates unarmed. 



D.^ Covering of skeletal plates a close capillary spinulation. 

 E.^ Intraambulacral ridge (of alternate adambulacrals) with two spines, the outer with a 



huge pedicellaria at tip barathri. 



EE.- Intraambulacral ridge with three or four spines and no huge pedicellaria. . . .ackleyi juv. 

 DD.* Covering of skeletal plates made up of fine granules, little elongated; intraambulacral 



ridge with one spine and no huge pedicellaria adarni. 



CC Carinal plates armed with a more or less conspicuous central spine. 

 D.' Actinolateral plates with a conspicuous central spine-bearing tubercle. 



E.' Each adambulacral plate with two bunches of pedicellariae high up in furrow .... alfredi. 

 EE.' No such bunches of pedicellariae. 



F.' Rays strongly carinate; innermost adambulacral spine with a terminal tuft of about 



ten pedicellariae ophiaclis. 



FF.' Rays not strongly carinate; innermost adambulacral spine with four to eight pedicel- 

 lariae near tip. 

 G. Disk-plates unarmed; rays very long, slightly swollen at base and then tapering 



steadily to a slender tip; R = 18-20 r magnificim. 



GO. Primary plates of disk armed with a central spine. 



H. DLsk and arm-bases more or less depressed; non-projecting adambulacral plates 



with only two spines planus. 



HH. Disk and arm-bases not depressed; non-projecting adambulacral plates with 



about four spines hirsvius. 



DD.^ No conspicuous tubercle near center of actinolateral plates though one tubercle may be 

 somewhat larger than others. 

 E.'' R = 16-18r; arms slender and tapering; no huge pedicellariae on adambulacral plates. 



longicaiidus. 

 EE.* R = 10-15r; arms stouter and less tapering; huge pedicellariae present on adambulacral 



plates spiyiulosus. 



AA. Adradial series of plates more or less reduced or wanting, their place often occupied by a con- 

 spicuous groove (sulcus) on each side of carinal series of plates. 

 B.' Disk large, low, very distinctly delimited from rays by a ring of massive, semiglobular plates; 

 primarj' plates all massive and semiglobular; all plates covered with capillary spinelets with a 



few coarser spinelets interspersed gilesii. 



BB.' Disk not as above. 



C- Superomarginal plates conspicuously armed with a central spine macracanlhus. 



CC- Superomarginal plates practically unarmed. 



D.* Adambulacral plates with two transverse series of actinal spinelets. 



E.' Carinal plates armed, deeply notched on both sides angulatus. 



EE.' Carinal plates unarmed, squarish microporus. 



' The figures given by Sladen (Challenger Ast., pi. 66) purporting to hefulgens are almost certainly 

 diomedeae, for Sladen called all the western Atlantic specimens f^d gens; his description too seems to have 

 been made from the American form. This unfortunate state of affairs is misleading and may be why 

 Koehler considered his Bay of Biscay specimen a new species (trispinosus) instead of referring them to 

 fulgens. The differences between trispinosus and Sladen's fulgens are evident, but comparison of Koeh- 

 ler's descriptions and figiires with a true fulgens leads one to suspect they are identical. 



' As Fisher's description is brief and comparative, I am not sure that my summary of the distin- 

 guishing features will prove reliable. 



