OPHIURANS OF THE PHILIPPINE AND ADJACENT WATERS. 247 



In the specimen from station 5640 the general coloration is a rather 

 dark greenish; the dorsal surface of the disk shows light purple 

 spots on the radial shields, and the small spines borne by the plates 

 in the interradial regions are purple ; these plates themselves show a 

 purple bordering ; the arm spines are grayish white banded with red. 

 A median longitudinal purple line extends the whole length of the 

 dorsal surface of the arms, but it is formed by a succession of little 

 spots very close together. 



The hook formed by the first ventral arm spine at some distance 

 from the disk is peculiarly elongated and slender, and exceeds 0.5 

 mm. in length in specimens of medium size (pi. 100, fig. 3&). On its 

 concave border it bears four or five rather short and thick spines, 

 and the terminal spine is not stouter than the following; it is even 

 sometimes shorter ; in the distal half or third of the arms the number 

 of arm spines falls to three. The tentacle scale is rather small, oval 

 and delicate, and ends in a very long point, sometimes accompanied 

 by a second almost equally developed (fig. 4Z>). 



I have already discussed (Kcehler '98, p. 102, the affinities of 0. 

 lepida with 0. purpurea. Aside from the characters based upon the 

 number of arm spines and the form of the mouth shields for the 

 two species resemble each other very greatly in these respects they 

 may be distinguished from each other especially by the form of the 

 radial shields, which are much elongated and very narrow in O. 

 lepida, where their proximal apex comes very near the center of the 

 disk (pi. 36, fig. 5), while in 0. purpurea these shields are very 

 much broader and triangular in form, and their proximal apex is 

 some distance from the center of the disk. The outlines of these 

 shields may be made out on the photograph published by Doderlein 

 ('99, pi. 17, fig. 23), as well as on that which I give here (pi. 58, fig. 

 3). The upper arm plates are also a little more elongated in O. 

 lepida, in which they have a lozenge shape form with the two ex- 

 ternal angles tuncated, their greatest breadth being found almost at 

 their middle, while in O. purpurea the greatest breadth is close to the 

 distal border. Doderlein says that these plates are a little broader 

 than long with the internal and external borders almost straight and 

 equal ; but this is inexact, and on the photograph which he has pub- 

 lished, although the outlines of these plates are not very distinct, 

 it may be shown that their form does not correspond with the de- 

 scription given in the text. I have thought it well to include two 

 photographs (pi. 58, figs. 3, 4) of an 0. purpurea from the Sunda 

 Archipelago which is in my collection, in which better than in Doder- 

 lein's photograph the outlines of the different plates may be made 

 out and which permits a more exact comparison with O. lepida. I 

 may add that the very robust tentacle scale of 0, purpurea is very 



