56 BULLETIN 84, UNITED STATES NATIONAL MUSEUM. 



As for the external mouth papilla, it is flattened, a little elongated, and erect ; 

 it is somewhat narrower at its basis than at its end, which forms a rounded edge. 

 Thus are explained the contradictions apparently existing between Lyman's 

 descriptions and his drawings. 



Lastly, the brachial spines amount to six on the first article and five afterwards. 

 The two spines which succeed the first ventral one terminate each in a little hyaline 

 point, which is conical and sharp, is at a right angle with the 'spine, and directed 

 toward the end of the arm; the fourth spine still shows a like arrangement, but 

 often less strongly marked. The spines are consequently not bihamuled, the 

 terminal hook being developed on one side only; on the other side the spines show 

 only a few rugosities. Under the terminal hook there may also be observed one 

 or two small denticulations, but these are very weak. The last dorsal spine may 

 show also a little terminal point, which is not so strong as on the preceding spine 

 and is sometimes directed in the axis of the spine and sometimes forms a certain 

 angle with it. The first ventral spine is cylindrical; the others are more or less 

 flattened; chiefly so is the last dorsal one. 



The mouth shields show a form slightly different from that figured by Lyman 

 in 1875. In the scheme published by him (75, pi. 5, fig. 68) he ascribes to these 

 shields an almost circular periphery, while on fig. 35, plate 3, the shield is oval. 

 This latter form is nearer to the one I observe on the two specimens which I have 

 in hand, the mouth shields of which are longer than wide and a little narrower in 

 their distal half, which terminates in a little truncated edge. That form recalls the 

 one I have observed and noted above in A. kinbergiensis. 



In the same bottle there was also a very small individual, the disk of which is 

 only 2.5 mm. in diameter; the arms are all broken at their beginning. I mention 

 this specimen because its primary plates are still visible on the upper face and the 

 plates of the under face are even more distinct than on the two larger specimens, 

 although of no greater size. Moreover, the mouth shields have an almost regular 

 oval shape and their form very much recalls that which Lyman represented in 1875 

 (pi. 3, fig. 35). 



The characters of A. palmeri are thus now defined and this species must take 

 place in the section of Amphiura s. str., in which the under face of the disk is 

 provided with a regular and uninterrupted covering of plates. 



AMPHKTRA FIBULATA, new species. 

 Plate 7, figs. 3-5. 



Fish Hawk station 7295. Feb. 24, 1902. Gulf of Mexico. Lat.2438' 40" N.; 

 long. 81 56' 26" W.; 5| fathoms; co.; temp. 19.5 C. A single specimen. 



Type.C&t. No. 32293, U.S.N.M. 



The disk is 8 mm. in diameter; the arms are incomplete, but some of them are 

 preserved to a length of 70 to 75 mm.; and they must have been much longer. 



The disk is rounded. The upper face is covered with imbricated plates, rather 

 uniform, somewhat smaller, however, in the central region and in the middle of the 

 interradial spaces; they become a little larger near the radial shields; on the con- 

 trary, near the margin of the disk and in the interradial spaces they rapidly become 



