108 BULLETIN 84, UNITED STATES NATIONAL MUSEUM. 



also their distal side becomes more and more convex and even protruding. All 

 these plates are separated from one another from the first one upward. 



The fairly protruding lateral plates carry six spines each. The first ventral 

 one is very short and thin; the two following ones are about equal to the article, 

 then the length increases up to the last dorsal one which exceeds two articles at the 

 base of the arms; the rows formed by these spines are not dorsally approximated. 

 These spines are thin and pointed, beset with small denticulations which, as a rule, 

 are very thin, conical, and rather loosely distributed. These denticulations, which 

 are never much developed, appear chiefly on the ventral and lateral spines and 

 become less apparent on the last dorsal spine. 



The tentacular scale is large, strong, lanceolate, with an obtuse point, and its 

 surface is strongly rough; it is notably longer than half the corresponding brachial 

 plate. 



Connections and differences. 0. porrecta stands evidently at the limit between 

 the genus Ophiacantha and the genus Ophiomitrella, for the plates of the upper face 

 of the disk, although being not only apparent but very plain on the younger speci- 

 mens, become somewhat indistinct as they grow older. Besides, I note a like 

 peculiarity in other Ophiacanthidse and, especially, in Ophiacantha aristata Koehler, 

 where, however, the plates remain somewhat easier to perceive on the larger samples. 

 It is with this latter species that 0. porrecta is most closely allied. It differs from it 

 in having the plates of .the upper face of the disk bear actual spinulous spines 

 instead of those stumps provided at their ends with those spinulous expansions 

 which are so peculiar and which I have described and figured for 0. aristata (09, 

 pi. 26, fig. 6); in the upper brachial plates a little larger than in 0. aristata, 

 where they remain very small and compressed; in the lateral brachial plates which 

 are stronger and which carry, on their strongly thickened distal side, spines which 

 are more numerous, longer, and arranged in rows, which at the basis of the arms 

 are approximated dorsally. The tentacular scale, at least, is less developed and 

 less rough in 0. porrecta than in 0. aristata. 



I beg to call attention to the latitude of the station where 0. porrecta has been 

 discovered; this species very likely represents in the Southern Hemisphere the 

 0. aristata of the Northern Hemisphere. 



OPmOMITRA ROBUSTA, new species. 



Plate 10, figs. 4-5. 



Albatross station 2347. Jan. 20, 1885. Lat. 23 10' 39" N.; long. 82 20' 

 21 "W.; 21 6 fathoms; co. One specimen. 



Type.C&i. No. 32300, U.S.N.M. 



This specimen is very incomplete; two arms are completely lacking, the other 

 three are broken off close to their bases and only a few articles are preserved; yet 

 the disk and the remaining parts of the arms are in excellent condition and the 

 specimen can perfectly well be described. I consider it as belonging to a new 

 species. 



The disk is thick and slightly excavated at the level of the insertion of the arms, 

 while it is most protruding in the interradial spaces. Its diameter is 16 mm.; 

 the arms are wide and strong; the specimen must have been very robust. 



