ON DOEYPTERUS HOFFMANNI. 261 



The pectoral fin (c), though in disorder, occupies its proper 

 situation immediately below the angle of the operculum, near 

 the longitudinal centre ; and one of the ventral fins (d) is well 

 displayed at the ventral margin in connexion with the lower ex- 

 tremity of the clavicle. It is narrow, and its pointed extremity 

 inclines backwards as far as the posterior extremity of the se- 

 cond ventral or thoracic plate ; it is upwards of half an inch in 

 length. Little more than half an inch remains of the elongated 

 portion of the dorsal fin (a) ; and mere traces are perceptible of 

 the narrow posterior division. The narrow backward portion of 

 the anal fin is scarcely distinguishable ; but the large anterior 

 member (6), though not entire, is distinct about the middle of 

 the ventral margin and a little behind the ventral plates. The 

 tail-fin is almost perfect, exhibiting both the general outline 

 with the lobes spread and the surface-characters. The upper 

 lobe, which is a little longer than the under, is about an inch 

 long. 



The transverse series of plates and rods are much disturbed 

 over the whole surface, though in places they are partially in 

 order, and show the same surface-characters as are displayed in 

 specimens in which these parts are more perfect ; and, as in 

 them, they are covered with black enamel-like matter. The 

 ventral plates (m), however, are arranged in complete order 

 along the ventral margin, from the base of the ventral fin to a 

 little in front of the anal, overlying each other apparently for- 

 wards ; but on closer examination they are seen to imbricate 

 backwards, and to give off from their upper extremities the pe- 

 culiar compound columns (n) already described, which, extend- 

 ing upwards in parallel order, incline towards the head. The 

 anterior ones are the shortest, and appear to terminate near to 

 the margin of the clavicle. A few of the posterior pass above 

 the angle of the operculum, and can be seen to join with the 

 lateral plates of this region, notwithstanding that they are much 

 disturbed. The large posterior abdominal rods (o) occupy their 

 usual position, sloping backwards from the last ventral plate. 



The specimen (PL X., fig. 2) next to be described is highly 

 instructive, though it is in many parts much disarranged. The 



