OPHIURANS OF THE PHILIPPINE AND ADJACENT WATERS. 283 



cula the upper arm plates show the same characters as in O. elegans 

 (pi. 43, fig. 1). 



Furthermore, if the dorsal surface of the arms be followed almost 

 to their extremity, it will be seen in both these species that the break- 

 ing up of the dorsal plates ends a short distance from the tip ; the 

 plates from then on are entire, as may be seen on the photograph 

 which I give on plate 43 as figure 7. These plates are longer than 

 broad, with the proximal border narrower than the convex distal 

 border; I notice also that on the photograph of O. elegans published 

 by H. J. Clark in 1915 ('15, pi. 12, fig. 7) the breaking up of the 

 upper arm plates may be very well made out, and the limits of the 

 successive plates may be sometimes distinguished. 



I may add that in the young individuals the breaking up of the 

 plates is not yet evident ; this condition I find in the two small speci- 

 mens from station 5151, in which the disk is only 3 mm. in diameter. 

 I include a photograph of the dorsal surface of one of them (pi. 43, 

 fig. 3), and it may be seen that the upper arm plates show very 

 clearly with the contours very distinct from the base of the arms 

 outward, and that they are quite entire throughout the whole length 

 of the arm. 



The arms are always very long, and their length may reach from 

 10 to 15 times the diameter of the disk; they taper very gradually 

 and are very slender in their terminal quarter. They are always 

 more or less twisted, and seem to be able to bend as easily in the 

 vertical plane as in the horizontal. The arm spines number six, and 

 their length increases from the first to the fifth, which is almost 

 equal to two segments ; the sixth is a little shorter. The first ventral 

 spine transforms into a hook with three or four branches at a short 

 distance from the disk (pi. 103, fig. 7a) ; its insertion is rather far 

 from the border of the tentacular pore, which is without a scale. 

 These spines are opaque and, except for the dorsalmost, which is a 

 little more pointed, they maintain almost the same width throughout 

 their whole length ; their tip is rounded (fig. 7d) . Their denticula- 

 tions are slightly developed on the proximal three-quarters of their 

 length, but they become very stout in the terminal region. They 

 are somewhat unequal, thick, conical, not very sharp, and often 

 slightly recurved in the form of a hook. Generally speaking, the 

 denticulations are more developed on the proximal border of the 

 spine. 



Ophiogymna elegans has been recorded by Ljungman from Singa- 

 pore and from Hongkong. The Siboga met with it at Banda and at 

 the Kei Islands in from 9 to 36 meters (5 to 20 fathoms) of water. 



