1848. | SIR P. EGERTON ON PTERICHTHYS. 307 
the midwater. It is then fair to presume that Agassiz is correct in 
considering the members of this genus as ground feeders, living on the 
mud and sand at the bottom of the sea. If so, what an impediment 
to its motions, and what an obstacle to progression in any direction 
must this umbonated sternum have been! The most vigorous efforts 
of the tail, although aided by the cephalic oars, would have failed to 
effect the most limited amount of locomotion. But if we reverse the 
order of things, what form could be better adapted for such a mode 
of life? The level ventral surface would glide with the slightest im- 
petus over the slimy bettom; at the same time the vaulted carapace 
would afford a most effectual buckler of defence against injury from 
external violence. This view is fully corroborated by comparison with 
other fishes of similar habits. In modern times we see the Sturgeons 
and allied forms, which seek their food by grovelling at the bottom of 
rivers and estuaries, protected by strong bony plates covering their 
arched heads and bodies, those along the dorsal me very much re- 
sembling the central dorsal plate of Pterichthys. Among the con- 
temporaneous inhabitants of the primeval seas we find the nearest 
allied forms, Cephalaspis and Coccosteus, flattened below, arched 
above, and defended by a similar panoply of plate armour. This 
argument, founded on reason and analogy, for supposing Mr. Miller’s 
description of the economy of Pterichthys to be the correct one, 
receives remarkable confirmation from the knowledge of the mode in 
which the plates of the carapace are jointed together. I have before 
stated that the plates covermg the back and sides are united by 
simple apposition of their corresponding edges, like the masonry of 
an arch, the umbonated central plate representing in position and 
effect the key-stone, the plates on the flat surface being united to the 
central diamond-shaped plate by broad squamose sutures. As these 
plates overlap the edges of the central plate, a very slight vertical 
blow from a stone or other moving body would suffice to dislocate it 
were it supine, but situated on the belly it would be effectually re- 
moved from all danger from external violence. If this be its true 
position, we may assign the exceptional mechanism of its attachments 
to the discharge of a most important function in the economy of the 
animal, namely that of providing for the occasional distension of 
the internal area of the body when required for the lodgement either 
of food or ova. The position of the pectoral fins or oars must not be 
overlooked in arguing this question. The plates carrying the ar- 
ticulating surfaces for the attachment of these organs are situated on 
the anterior margin of the flat region of the body. Now if this be 
considered the back of the fish, these appendages would bear more 
resemblance in form and position to the jomted horns of Pliny’s 
Kthiopian bull than to the natatory organs of which they are 
affirmed to be the homologues; whereas on the opposite supposition 
of the relative positions of the parts, they would be serviceable either 
for progression or defence, for stirring up the mud in search of food, 
for scooping out hollows for concealment or deposition of spawn, or 
for clouding the water on the approach of an enemy. In search- 
ing for further evidence in support of my views, by comparing Pte- 
