110 BULLETIN 84, UNITED STATES NATIONAL MUSEUM. 



lateral parts of the upper plates have preserved a brownish-pink color, and there is, 

 besides, a large purple-brown spot on the lateral plates on either side of the distal 

 edge of the upper plate; the whole must have had in the live animal the appearance 

 of a bicolored stripe extending all along the upper face of the arms. 



The first under brachial plate is small, quadrangular, longer than wide, narrow 

 between the two corresponding adoral plates; the narrow proximal side is divided 

 into two parts by a median furrow extending on the under face of tlie plate. Tlie 

 following plates are very large, quadrangular, having a convex distal side and lateral 

 edges, which are divergent and strongly excavated at the corresponding tentacular 

 scale; one can hardly speak of a proximal angle, except for the first plates, where 

 this angle is very obtuse, while it reaches 180° on the following plates, the sides of 

 this angle being continuous and transverse, forming a proximal side. The first 

 plate is already a httle wider than long, but the following ones grow very rapidly 

 wider until they become almost twice wider than long; these plates are almost 

 contiguous, the interval between them being extremely narrow. 



The lateral brachial plates are short, but their distal side is very wide and thick. 

 They each are provided at tlie basis of the arms with nine large and strong spines, 

 the first ventral being equal to one and a half articles, and the last dorsal exceeding 

 four articles; the rows are not approximate dorsally. These spines are provided 

 with fine denticulations loosely spread but not very conspicuous and which generally 

 are even completely lacking on the large upper spines. 



The smgle tentacular scale is extremely developed; it is wide, thick, and 

 lanceolate, ending in a blunt and scarcely rough point; the length of that scale 

 almost reaches that of the corresponding bracliial under plate. The first pore 

 carries always two scales, but the following ones have but one; however, I excep- 

 tionally find on one of the arms another pore provided with two scales. 



Connections and differences. — 0. robusta is aUied cliiefly to 0. omata VerriU and 

 0. spinea VerriU. It can easily be distinguished from 0. omata, the radial shields 

 of which are contiguous on their whole length, and the mouth shields, as well as 

 the upper and under brachial plates of wliich, have a different shape ; moreover, the 

 brachial spines amount to five only; the characters of this species clearly appear 

 in Verrill's figures. The second species has not been represented and is known only 

 by a rather short description by VerriU; the radial shields are externally contiguous; 

 the adoral plates are thick and crescent-shaped ; the brachial spines, wliich amoiuit 

 to nine, as in 0. rohusta, are provided with very strong denticulations; lastly, the 

 oral papillie are much more numerous. All these characters fail to apply to the 

 specimen gathered by the Albatross, but, unfortunately, VerrUl gave no information 

 regarding the shape of the under and upper brachial plates of 0. spinea, which 

 plates are especially characteristic in 0. rohusta. 



OPHIOMITRA VALIDA Lyman. 



See for bibliography : 

 VerriU (99a), p. 353. 



Fish Hawk station 7280. Feb. 14, 1902. Lat. 24° 17' 05" N.; long. 81° 58' 

 25" W.; 132 fathoms; s.; temp. 52° F. Five specimens. 



Fislt Hawk station 7281. Feb. 14, 1902. Lat. 24° 13' 45" N.; long. 81° 58' 

 15" W. ; 304 fathoms; s.; temp. 52° F. Two specimens. 



