REPORT ON THE ASTEROIDEA. 639 



its extreme delicacy, and for this reason I am unable to say whether the pedicellaripe 

 were present to the same extent on the spinelets of the ovarial regions as on those of the 

 disk. The plating does not terminate with the ovarial regions, but is continued along the 

 ray, the plates, however, becoming much smaller and more delicate. I have traced it 

 considerably beyond the middle of the ray, and isolated plates with single spines occur up 

 to the extremity. It is only on the outer part that I have been able to find any traces of 

 the groups of pedicellarise, comparable to those on the saddle-like sacculi, which are so 

 conspicuous in other species. Certainly none are present where the plating is continuous. 

 The groups here noted are circular in form, placed near the lateral margin of the ray, 

 and the pedicellarise are extremely small. 



The ambulacral furrow is broad, and has a singular appearance, simulating rather 

 suggestively a lineal series of confluent keyhole-shaped apertures, in consequence of the 

 formation of the adambulacral plates about to be noticed. The adambulacral plates are 

 elongate (2"5 mm. in length at 80 mm. from the disk), very like vertebra} in form, with 

 their furrow-margin deeply incurved. This scooping out of the margin does not commence 

 immediately at the adoral end of the plate, but rather abruptly a little way beyond ; it 

 is continued, however, up to the extreme aboral end, and the latter, in consequence of 

 the depth of the concavity, has the appearance of being produced at an angle horizontally 

 into the furrow ; a keyhole-shaped outline is thus produced, and the width of the furrow 

 is obviously much narrower when measured between the thick adoral portions of the 

 adambulacral plates than across the deepest part of the incurvature, the breadth in one 

 case beirjg less than 1 mm., and in the other fully 2 mm., at a part where the width of 

 the ray is 3*6 mm. The successive pairs of ambulacral tube-feet have thus a widely 

 spaced and isolated appearance. The armature of the adambulacral plates is very simple, 

 and consists of only one very delicate spinelet which stands perpendicularly, and an equally 

 delicate lateral spine, articulated on a rudimentary ankylosed infero-marginal plate, usually 

 on every alternate plate. The small inner spinelet at the aboral extremity of the furrow- 

 margin of the plate, which is present in most species, is here wanting altogether. The 

 perpendicular spinelet measures about 4 mm., but frequently less, and is articulated on a 

 small tubercle situated on the actinal surface of the plate at its narrowest part, and con- 

 sequently in a line with the pair of ambulacral tube-feet and nearer the aboral than the 

 adoral extremity of the plate. It is encased in a delicate membranous sheath crowded 

 with pedicellarise. The lateral spinelets are of very great delicacy, about 7 mm. in length 

 at 150 mm. from the base, and like the perpendicular spinelets are encased in a delicate 

 membranous sheath with crowded pedicellariai. So great is the delicacy of both perpen- 

 dicular and lateral spines alike that very few can be found undamaged. 



The actinostome is wide, its diameter beiug 675 mm. in a disk measuring 12'5 mm. 

 The buccal membrane is semitransparent and permits the plications of the digestive 

 cavity to be seen. The mouth orifice is small, and its margin is much crenulated, the 



