CLARK: BRITTLE-STARS. 333 



which must be referred to this species, though it differs somewhat 

 from Koehler's specimens from Ubian, Sulu, and Damar in the East 

 Indies. The disk is nearly 5 mm. across, but the arms, now broken 

 at the tip, could not have been much more than 12 mm. Koehler 

 gives the arms as 15 mm. long, in a specimen with the disk 4 mm. in 

 diameter. There are 3 tentacle-scales on the first, and sometimes 

 on the second pore, of each arm, and beyond that there are two scales 

 to the end of the arm. In Koehler's specimens, there were 2 tentacle- 

 scales, only on the basal part of the arm. The fine striations on the 

 arm-plates to which Koehler refers are not very well-marked in the 

 present specimen. Finally, Koehler says nothing about color, while 

 the specimen at hand has a very distinct coloration; the disk and 

 the whole lower surface are yellowish white, while the arms are banded 

 with grayish brown and dirty white. 



Ophioderma phoenium. 



<j>oivLos = blood-red, in reference to the color of the disk. 



Plate 6, fig. 1, 2. 



Rolotype.— M. C. Z. 4,165 and 50 Paratypes, M. C. Z. 4,166, 4,207. 

 British West Indies: Tobago, Buccoo Reef, April, 1916. Carnegie 

 Expedition. H. L. Clark coll. 



Disk, 20 mm. in diameter; arms 70-80 mm. long. Disk covered 

 with a uniform coat of granules, about 50 to a square millimeter. 

 Radial shields very small, more than a millimeter long, less than a 

 millimeter wide and over .hree millimeters apart; two are entirely 

 concealed by the granules and one other is partially covered. At 

 the side of the base of each arm is a considerable area of overlapping 

 scales extending out from the disk as far as the eighth or ninth arm- 

 segment (second or third beyond disk-margin) and running back orally 

 to the third segment. Upper arm-plates quadrilateral with the outer 

 corners much rounded; they are 3^ times as broad as long, are very 

 rarely broken or cracked as in 0. cinercum and some other Ophio- 

 dermas, and are in contact for their full width. Interbrachial areas 

 below covered by granules like those of disk. Genital slits 4 in each 

 area, very small, widely separated; distal pair much the larger and 

 exceeding an arm -segment in length. Oral shields triangular with 

 rounded angles, rather wider than long. Adoral and oral plates 

 completely concealed beneath a uniform coat of granules like thos^ 



