198 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



broader than long, and assuming an almost circular form; they are 

 all in contact. 



The first under arm plate has its distal portion much narrowed be- 

 tween the neighboring adoral plates, which are very close together 

 in the radial line, while the proximal portion broadens and takes the 

 form of a triangle which is broader than long. The following plates 

 are hexagonal, with three short proximal sides; the lateral borders 

 are much elongated, slightly convex and almost straight; the distal 

 border is narrow and slightly excavated in the middle. All these 

 plates are narrow and very much longer than broad; they are all in 

 contact. 



The rather strongly projecting side arm plates bear each seven 

 rather strong, subequal spines equaling the segment in length ; some- 

 times the central spines are a little more developed than the others. 

 These spines are cylindrical, and they bear at their tip some trans- 

 parent asperities, of which one is especially developed on the second, 

 third, and fourth spines, forming a little hook which is directed 

 toward the extremity of the arm (pi. 96, fig. 8). 



The tentacle pores are extremely large, but tentacle scales are com- 

 pletely lacking. 



The color of the specimen in alcohol is a very light brownish gray. 



If we compare the preceding description with Ljungman's short 

 diagnosis, we find the following differences : 



First, Ljungman in speaking of the radial shields says: "longa, 

 cuneiforma, . . . parte minuta supra insertionem brachiorum nuda 

 excepta, cute tecta." But the radial shields in my specimen are not 

 cuneiform ; they are naked throughout their whole length, and ad- 

 joining them there is found, instead of a naked integument, a band 

 of small plates broadening progressively toward the proximal ex- 

 tremity of these shields where they form several rows, an arrange- 

 ment not mentioned by Ljungman. Furthermore, Ljungman says 

 that the mouth shields are much longer than broad ; I find that they 

 are almost as long as broad. He also says that the under arm plates 

 are " parum longiora quam latiora " ; in reality I notice that these 

 plates are very much longer than broad. 



The two last differences are not of much importance; as for the 

 first, it perhaps is due to the fact that my specimen is larger than 

 Ljungman's. In addition, the Swedish author says, in the diagnosis 

 of the genus Ophiocentrus, that the dorsal surface of the disk is 

 provided with small erect spines, but he does not specify whether he 

 has distinguished the plates which bear these spines, which, however, 

 may well have been the case. 



The other characters mentioned by Ljungman conform so well with 

 those which I observe in my specimen that I have not had any hesita- 



