146 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



Albatross station 5660; Flores Sea; Cape Lassa bearing S. 88 W., 

 38.00 kilometers (20.5 miles) distant (lat. 5 36' 30" S., long. 120 

 49' 00" E.); 1,266 meters (692 fathoms); December 20, 1909; gy. 

 M., S. 



Two specimens (Cat. Nos. E. 28, E. 29, U.S.N.M.). 



Albatross station 5668. Five specimens (Cat. No. E. 269, U.S.N.M.) . 



Notes. As I wrote in 1904, there are wide variations in the devel- 

 opment of the spines on the dorsal surface of the disk. I give here 

 photographs of two specimens of which one, from station 5450. 

 shows very few spines (pi. 24, fig. 1), while the other, from station 

 5618, has the spines very numerous and elongated (fig. 3). I also 

 include photographs of these different spines (pi. 96, fig. 1). 



In comparing the Albatross specimens with those collected by the 

 Siboga I may remark that in my description in 1896, which was 

 drawn up from a single specimen, I did not lay sufficient emphasis 

 on the characters of the under arm plates. It is more correct to 

 say that these plates are pentagonal, with a proximal angle so 

 obtuse that it may almost completely disappear ; this angle is bounded 

 by two straight sides ; the lateral borders are notched by the tentacle 

 scale, and the gently convex distal side is extremely broad (pi. 24, 

 figs. 2, 4, 6). These plates are separated from the second or third 

 outward, and the space between them may be longer or shorter 

 according as the plates are themselves longer or shorter; but, gen- 

 erally speaking, they are very short in proportion to their width; 

 sometimes they may be three times as broad as long, and separated 

 by an interval which itself is rather long. In certain specimens the 

 plate is even not very much longer than this interval. 



The distal border of the upper arm plates is sometimes convex and 

 very regular and joins very exactly the adjacent side arm plates: 

 but sometimes it seems to be slightly elevated and it then appears 

 more or less jagged, showing very small and very short denticula- 

 tions, which are often rounded, and are arranged very regularly 

 along the border ; this feature is more marked in some specimens 

 than in others. I give here a photograph of the dorsal surface of one 

 arm of the specimen from Station 5444 which shows this character 

 very pronouncedly (fig. 5) ; it may also be seen, though less accentu- 

 ated, in the specimen represented in figure 1. 



As I stated in 1904, the tentacle scale is rather thick at the base 

 and pointed at the tip; it is sometimes somewhat lanceolate and 

 spiniferous (fig. 6). It often happens that the scale, at first broad- 

 ened at the base, tapers sharply and becomes from then on very 

 narrow in the outer two-thirds or even three-fourths of its total 

 length (fig. 4). 



The type of O. rudis was found by the Investigator in the Anda- 

 man Islands at a depth of 780 fathoms. The Siboga met with it at 



