OPHIUKANS OF THE PHILIPPINE AND ADJACENT WATERS. 393 



large specimens the radial papillae usually appear as small rounded 

 granules, forming a little more or less important clump on each 

 side of the base of the arms (figs. 7, 8). Sometimes these granules 

 elongate somewhat, especially on reaching the ventral surface, but 

 in the small specimens the papillae are very much longer, and they 

 even elongate in such a way as to form true spines (figs. 2, 3, )-. 



More important variations are found in the mouth shields. In my 

 type these shields are lanceolate, twice as long as broad, with a sharp 

 proximal angle. This is the form which they have in all the speci- 

 mens from station 5623; but in that from station 5624, which is of 

 large size, these shields are shorter and very much broader ; they are 

 not even half again as long as broad, and their proximal angle is 

 obtuse or even rounded (fig. 9). A similar form is seen in the small 

 specimens from stations 5127, 5428, and 5605 (fig. 1) ; but in these 

 specimens they are more elongated than in the large individual from 

 station 5624; the proximal angle is rather sharp, and they are at 

 least half again as long as broad. 



I find also some differences in the arm spines, which may be four 

 in number at the base of the arms. 



The tentacle scales are usually larger and more important than I 

 described them in 1896 ; I then indicated only one large tentacle scale, 

 but I have found on the Albatross specimens that in addition to a 

 large tentacle scale which occupies the proximal border of the pore 

 there may be present two or even three small scales on the distal 

 border. When the pores are contracted these distal scales may not be 

 apparent, and it is doubtless this which occurred on the Investigator 

 specimens; but I always find them on those from the Albatross col- 

 lection, even in small specimens like that in figure 1. In the large 

 specimens of which photographs are given in figures 6 and 9 it may 

 happen also that the proximal scale is divided into two or three small 

 scales. The tentacle pores are rather large, especially those of the 

 first pair, and their size decreases progressively to that of the third 

 pair. 



Because of the variations to which 0. elegans is subject I have 

 reexamined the specimens on which in 1896 I established O. famttiare, 

 a species which is rather close to O. elegans ; but the comparison 

 between these specimens and the O. elegans collected by the Albatross 

 has confirmed the validity of the first species, which is abundantly 

 distinct from the second. The arm spines of O. familiare are five in 

 number, and sometimes I find six of them at the base of the arms in 

 the large individuals ; the two first ventral spines are somewhat closer 

 together, and they form a small distinct group while four others 

 follow at regular intervals along the distal border of the side arm 

 plate. The largest number of arm spines which I find in the large 

 specimens of O. elegans is four, and the normal number is three. 



