OPHIUEANS OF UNITED STATES NATIONAL MUSEUM. 123 



The upper brachial plates are very large, quadrangular, with a narrow proximal 

 side, a very wide distal side, meeting by sharp angles the lateral sides which are 

 divergent. The distal side is rather concave on the first plates, then it becomes 

 convex and it finally resolves itself into two small sides which are united by a very 

 obtuse and rounded angle, behind which the surface of the plate offers a very small 

 obtuse protuberance ; these plates are never much wider than long, and sometimes 

 they are as wide as long. The shape of the plates in the first half of the arms does 

 not appear very plainly, for the said plates are almost always more or less strongly 

 divided into fragments which, in certain rtegions, are so many that the limits be- 

 tween the successive plates are no longer distinct ; then the fragments become fewer 

 when the plates are only divided into two or three parts, and finally they remain 

 simple up to the end of the arms. This dividing up exists, besides, in very variable 

 degrees according to which arms are examined, and on one of them it is less con- 

 spicuous than on the others. The specimen being* singie, it is impossible to tell 

 whether this dividing constitutes a specific character or is accidental; none of the 

 arms show any trace of regeneration. 



The under brachial plates are absolutely invisible, and they are covered up by 

 the teguments. 



The brachial spines amount to six only, and they are not very much developed. 

 The first one is very short and thin; the second one is notably longer and stronger; 

 and then the length progressively increases up to the fifth one, which is hardly 

 larger than the corresponding article; the last dorsal spine is a little shorter than the 

 foregoing one. The relative length of these spines increases a little in the second 

 half of the arms. These spines are bare, flattened, fairly wide, with rounded ends, 

 and thjey are provided on their edges with pretty strong denticulations, which are 

 conical, somewhat unequal, lying close to one another on the ventral spine, but 

 more loosely arranged on the last two dorsal spines, where, besides, they are less 

 conspicuous. The first ventral spine changes into a hook near the fifteenth article. 



There is not the slightest trace of a tentacular scale, which is probably hidden 

 by the tegument. 



Connections and differences. 0. convoluta recalls certain forms of 0. fragilis, 

 but I thought I ought to separate it specifically, owing to the disk armature, to 

 the shape of the mouth shields, as well as to the peculiar characters displayed by the 

 upper brachial plates, and also owing to the shape of the arms, which are wrapped 

 around on themselves. I know among 0. fragilis no form in which the spines of the 

 upper face are so much developed and appear so thick on a disk having such a com- 

 paratively small diameter, while, on the contrary, the armature of the under face 

 of the disk is extremely reduced and almost nonexistent. The upper brachial plates, 

 instead of showing a more or less conspicuous distal angle, which is often upturned 

 in the shape of a protruding beak, are rather quadrangular with a slightly convex 

 distal side and they are not much wider than long. The shape of the mouth 

 shields is also rather peculiar, since they are almost lozenge-shaped and as wide as 

 long. As to the very curious dividing up displayed by the upper brachial plates 

 in the first part of the arms, it is not safe to dwell on it, for it is perhaps an individual 

 feature, but this seems to me most doubtful. 



