1870. | HANCOCK AND HOWSE—DORYPTERUS HOFFMANNI. 637 
apparently the skin. ‘The mandibles are not disturbed, and by their 
position inditate that the mouth was closed; but the bones of the 
upper jaws are not determinable, except the left maxilla, which is in 
a pretty perfect state; the premaxilla, however, has been torn away. 
The orbit (w) is pretty well preserved, but is removed a little below 
the natural position. ‘ 
The contour of the dorsal margin, in front of the dorsal fin, is 
entirely lost, this region having been forced downwards nearly as 
far as the centre of the body. Posteriorly, however, the whole form 
is well preserved, and exhibits the transverse plates and rods in a 
good state of preservation, the series of lateral (gq) and sigmoidal 
plates (7) being almost complete; the line of the central lozenge- 
shaped plates (h) is also quite distinct, with a plate here and there 
bearing a central ridge. In front this line is thrown suddenly 
upwards, in consequence of the ruptured condition of the anterior 
parts. Twenty-eight lateral plates can be counted, and fifteen or 
sixteen sigmoidal ones along the lower margin ; only ten or eleven 
can be made out of the dorsal row. 
One of the most interesting features in this specimen is the ven- 
tral plates (m), which are displayed in a remarkable manner. With- 
out the aid of this specimen it might have been difficult to determine 
the existence of a double series of these plates, one belonging to each 
side of the abdomen. Here, however, both sets are most satisfactorily 
seen, those belonging to the exposed, or left side, being arranged in 
their natural order, while those of the right side are turned down- 
wards below the ventral margin, so that their outer surface is placed 
upwards and in the same plane with the similar aspect of the other 
plates. The lower margins of the two sets are in close contact, 
the line of junction coinciding with the ventral margin. ‘The plates 
in both series incline forwards, their pointed extremities being 
turned backwards ; and the compound columns (7) in connexion with 
the upper extremities are distinctly displayed, those belonging to the 
left side taking their usual course upwards in the direction of the 
head and lozenge-shaped area, those on the right side, or those of 
the displaced series, passing outwards from the specimen, the 
columus, like the plates themselves, having been displaced and 
thrown downwards. And thus we have a satisfactory proof that 
both the ventral plates and columus are developed in two lateral and 
distinct series. And furthermore these columns are seen, in the 
specimen before us, to pass upwards and to become united to the 
lower anterior lateral plates, notwithstanding that they are much 
disturbed at this point. 
The dorsal fin(a) has been broken away, only three-quarters of 
an inch remaining. ‘The anterior basal extension, however, is well 
shown, as this portion is turned a little towards the observer and 
has been forced downwards. Consequently the short anterior basal 
plates, or fin-fulera (a''), are seen to be arranged in a bilateral series 
of about twelve. This bilateral arrangement of the fin-fulcra is not, 
we believe, usual among the ganoids, in which the two lateral halves 
are commonly united into one piece. Mere traces are observable 
