OPHIURANS OF THE PHILIPPINE AND ADJACENT WATERS. 25 



side of the arm. Their external surface is very convex, and their 

 internal concave, and they fit exactly over the corresponding surface 

 of the vertebra ; on the ventral surface of the arms they^ are continued 

 in the form of a process which reaches the median line; each of these 

 processes is separated from the following by a space occupied by 

 soft parts which cover an interval almost as broad as the process 

 itself. The spines, five in number in the type species, are composed 

 of a quite transparent calcareous substance; they are rather small, 

 subequal, each with a row of four or five sharp and gently recurved 

 teeth; they thus assume the form of compound hooks. There is no 

 tentacle scale. The tentacular pores are very small and situated 

 beyond the ventralmost arm spine. 



Remarks. — In my opinion the genus Ophiocanops must be placed 

 in the family Ophiomyxidae rather than in the Ophiobyrsinae, if the 

 distinctions established by Matsumoto are accepted; the mouth 

 shields, which are very small, are widely separated from the first side 

 arm plate by the adoral plates, and the articulation of the vertebrae 

 is zygospondylian ; it may be seen from the photograph reproduced 

 as figure 5 on plate 2 that this articulation is more complicated than 

 in the Ophiobyrsinae. 



This new genus may be compared with the genus Ophiosciasma, 

 but it differs from it markedly in the absence of under arm plates 

 and in having the rather numerous arm spines transformed into com- 

 pound hooks from the arm bases outward. 



Matsumoto has recently described two genera of Ophiomyxidae in 

 which the arm spines are transformed into compound hooks similar 

 to those in the genus Ophiocanops; these two genera are Ophiostiba 

 and Ophiohyalus. In the genus Ophiostiba the upper arm plates are 

 lacking, the side arm plates are subventral, and under arm plates 

 are present ; but the disk, which is naked on its dorsal surface, shows 

 a bordering of plates about its periphery. The genus Ophiohyalus, 

 which shows similar features, is distinguished from the preceding by 

 the presence of thin, hyaline, rudimentary upper arm plates. These 

 two genera thus differ from the genus Ophiocanops in the presence 

 of plates on the circumference of the disk and in the presence of 

 under arm plates; furthermore, the arm spines are fewer, only two 

 or three on each segment, and the tentacular pores are very large. 

 I do not find among the Ophiomyxidae any other genera which need 

 be compared with the genus Ophiocanops. 



In the Ophiobyrsinae Matsumoto has described a new genus, Ophio- 

 smilax, also from Japan, in which the arm spines are converted into 

 compound hooks; but in this genus there are similarly only two or 

 three arm spines, the upper arm plates are lacking, but under arm 

 plates are present, and there is only a single mouth papilla on each 

 side. 



