400 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



merous; in the specimen from station 5378 the primary plates are 

 not recognizable, but in others there occurs a more or less distinct 

 rosette of six contiguous and equal plates. Among the other plates 

 there may be distinguished in the radial spaces a series of three plates 

 separating the two shields of each pair; the first is elongated and 

 the third short and trapezoidal. In the interradial spaces there are 

 two tandem elongated plates, the first a little the shorter and pen- 

 tagonal, the second trapezoidal with the distal border broadened 

 and in contact with the dorsal border of the large interradial plate 

 which by itself covers the ventral surface of the disk outside of the 

 radial shields; a small part of this plate is visible when the animal 

 is viewed from the dorsal surface. The radial shields are triangu- 

 lar, with the apex rounded, slightly divergent, and separated for 

 their entire length; they are half again as long as broad, and their 

 length is equal to half the radius of the disk. All these plates of 

 the dorsal surface are covered with rounded granules, which are 

 slightly flattened, in contact, and always very evident ; these granules 

 also occur on the distal portion of the ventral interradial plate, but 

 they disappear on the ventral surface, properly speaking, of this 

 plate. 



This large ventral plate is trapezoidal, at least twice as broad as 

 long, with a rather narrow straight proximal border, sinuous and 

 strongly divergent lateral borders, and a slightly rounded distal side 

 which reaches the dorsal surface of the disk; this plate, which is 

 extremely large, covers the whole ventral surface of the disk beyond 

 the mouth shield, and there are no plates intercalated between this 

 and the genital plate. This large plate is smooth over the greater 

 part of its surface, and the granules appear abruptly near the pe- 

 riphery, forming on the ventral surface only three or four succes- 

 sive rows; this narrow border represents the only portion of the 

 ventral surface of the disk which is granulated, and over all the 

 rest of the ventral surface of the disk and of the arms the plates 

 are perfectly smooth. In figure 5 I give a photograph of this plate 

 and of the neighboring regions, taken obliquely from the ventral 

 side in order to show not only the disposition of these granules, but 

 also the entire extent of this plate in the interradial area. The 

 genital plates are elongated and broadened at the level of the point 

 of union between the mouth shields and the large interradial plate, 

 where their outlines follow a very obtuse angle. 



Some variation may be found in the development of the granules 

 which cover the plates of the dorsal surface of the disk. Some- 

 times these granules are much flattened and consequently only 

 slightly evident. This occurs, for instance, in the largest specimen 

 from station 5259, and the granules along the border of the large 

 ventral interradial plate are particularly very slightly evident. 



