OPHIURANS OF THE PHILIPPINE AND ADJACENT WATERS. 223 



The tentacle scale in all the specimens is large and very evident; 

 it is rounded or slightly oval (p. 98, fig. 2>d). 



The color of the specimens in alcohol is gray or j^ellowish gray. 



Affinities and distinctive features. — Ophiothrix cumulata naturally 

 approaches O. crassispina because of the thickening of the dorsal 

 spines observable in certain specimens. The individuals in which 

 the dorsal spines are thickened indeed strongly recall O. crassispina^ 

 the more so since also in this latter the upper arm plates at the base 

 of the arms are without spinules on their distal border, these spinules 

 not appearing for some distance beyond the disk; the armature of 

 the dorsal surface of the disk also shows a very strong similarity in 

 the two species. But the other characters are very different. Only 

 a small number of individuals of O. cumulata show the dorsal spines 

 thickened, while all the known specimens of O. crassispina have this 

 character, and it is already manifest in individuals of very small size. 

 That which constitutes an anomaly in the first species becomes the 

 rule in the second. In 0. cumulata the arms are relatively longer, and 

 the dorsal spines are longer, more slender, and more finely denticulate ; 

 the specimens have a more robust appearance and may reach a larger 

 size. In O. crassispina the individuals do not seem to exceed a rather 

 small size, and in the largest which was dredged by the Siboga the 

 diameter of the disk is only 9 mm. To appreciate fully the differ- 

 ences between the two species, it is well to compare individuals of 

 the same size, and there fortunately exists in the Albatross collec- 

 tion a specimen of 0. cumulata in which the diameter of the disk is 

 7 mm. I give a photograph of this specimen on plate 51 as figure 2» 

 By comparing it with the specimen of O. crassispina shown on plate 

 49 as figure 4, in which the diameter of the disk is also 7 mm., it is 

 made clear that the two can not be referred to the same species, 

 even without invoking the thickening of the spines which occurs only 

 in the specimen of O. crassispina. The form of the upper arm plates 

 is entirely different in the two species; in 0. crassispina they are 

 less broad and more elongated, their surface is more convex, their 

 lateral angles are less sharp and their distal border is convex. The 

 number of arm spines is larger in the first species, but especially the 

 increase in their length takes place in a very much more gradual 

 manner. The comparison of the two photographs which I give here 

 (pi. 52, figs. 2, 5) is very interesting from this point of view; it may 

 be seen that in O. crassispina the three first spines are very short,, 

 and that the third does not reach even the length of the correspond- 

 ing under arm plate, while in O. cumulata the second spine is very 

 much longer than the first, and attains the length of that plate. 

 The hook formed by the first ventral spine is more elongated in O. 

 cumulata, where it shows typically four branches, the two outer well 

 developed, while in O. crassispina this hook has only three branches, 



