SHARK-TAILED HYPOSTOMA. 143 



The above notes accompanied Mr. Schomburgk's 

 specimen and drawing, and we do not find any de- 

 scription by Valenciennes which will accord with 

 them. The specimen before us is in length thirteen 

 and a quarter inches to the extremity of the lower 

 or longest caudal ray, to the end of the upper twelve 

 inches, and is considerably depressed in form, par- 

 ticularly in the shape of the head ; the upper part 

 of the body is rounded, slight angles appearing, and 

 forming lines where the centre of the plates are 

 marked ; the lower row of plates running from the 

 opercula, are however, much angulated, show a 

 distinct line, and towards the tail, lie at nearly 

 right angles, causing the lower surface of the body 

 to appear much flattened. The head is much de- 

 pressed, between the eyes it is nearly flat, but in the 

 centre of the occipital plate rises into a narrow 

 ridge before joining the first dorsal plate. The 

 whole is rough, but not spined ; but before each 

 opening of the opercula there are a few short spines 

 placed in a tuft, and indicating the approach to the 

 structure we shall see in the next fish. The plates 

 upon the anterior part of the body are rounded on 

 their edges, but as they reach the tail the outline 

 becomes angular ; they are rough on the edges with 

 short spines, and towards the tail the angle becomes 

 marked with a longer spine, the others on each 

 side gTadually shortening. The lower parts of the 

 body anterior to the anal, and within the insertions 

 of the pectoral and anal fins, is thickly studded over 

 with 'minute points, separated into groups, and 



