CLARK: BRITTLE-STARS. 283 



lecting (by sifting the mud through a sieve) but a few perfect speci- 

 mens were secured. These uninjured specimens were very active 

 when placed in a basin of clean water, moving around the dish re- 

 peatedly and rapidly. Their method of progression was very striking 

 and is worthy of note; one of the arms was thrust straight ahead its 

 full length, the tip was sharply bent back, forming a hook, the arm 

 was then contracted into a wide S, and thus the disk was drawn up 

 near the hook, with the other four arms trailing inertly behind. The 

 process being rapidly repeated led to a speed not often attained by an 

 echinoderm. The excessive length of the arms, fully twenty times 

 the diameter of the disk, gives the animal an extraordinary appearance 

 when in motion. 



As no detailed description of this remarkable brittle-star has ever 

 been published, the following data will be of use. The disk is soft, 

 puffed, about 6-8 mm. in diameter, black, with the radial shields, 

 genital plates and genital scales white in marked contrast, as in 

 Liitken's figure. The arms are 140-150 mm. in length, more or less. 

 The upper arm-plates are broadly oval, about as long as wide, widest 

 proximal to middle, more or less fully in contact; they are thin and 

 their outlines are difficult to make out even in a dry specimen. The 

 whole upper surface of the arm is finely punctate. The color of this 

 upper surface is pale yellowish or nearly white, but on each side is a 

 broad, irregular and frequently interrupted longitudinal band of deep 

 purple. Orally the color is yellowish white but except on the basal 

 sixth or fifth of the arm, there are frequent splashes of faint purple, 

 covering two-four under arm-plates more or less imperfectly; although 

 these splashes are most irregular in their arrangement they give a 

 banded appearance to thf under side of the arms. The under arm- 

 plates are distinctly longer than wide, proximally rounded, distally 

 notched, fully in contact. There are no tentacle-scales, but the 

 tentacle-pores are large. Basally there are five arm-spines, slender, 

 acute, rather delicate; the first (loicest) is longest and exceeds the 

 segment; the next two are much smaller and about of a size; the 

 fourth and fifth are nearly or quite equal to the first. After about 

 30-40 joints, there are only four spines, and these are subequal and 

 exceed the segment. The madreporite is very large, rounded pen- 

 tangular, much wider than long; the other oral shields are much 

 smaller and have a rounded distal lobe; they might be described as 

 broadly triangular with all angles much rounded, the base of the 

 triangle proximal and the lateral sides sharply and rather deeply 

 concave. The adoral plates are small, somewhat triangular with 



