1848. | SIR P. EGERTON ON PTERICHTHYS. 309 
in at least the part of the figure below the wings, the dorsal surface 
of the Pterichthys. In his restoration of the Pterichthy ys in the 
same tab. fig. 1, there is a marginal strip marked /, /, outside the four 
plates, which certainly overlap, as you remark, the lozenge-shaped 
plate in the centre. Now this marginal strip is merely part of the sides 
thrown out by pressure. These originally struck upwards at nearly 
right angles with the flat ventral surface, as the sides of a travelling- 
trunk strike upwards at right angles with its bottom. ‘They were of 
considerable height,—fully equal, I should say, to half the breadth of 
the ventral superficies,—thus giving depth to the creature’s chest, 
which was rendered still more capacious from the circumstance that 
the back was arched—to hold by my illustration—like that of the 
travelling-trunk. No travelling-trunk, however, has a back so steeply 
arched ; and instead of terminating in the centre in a longitudinal 
rectilinear ridge, it terminated in a longitudinal pointed one—a single 
plate, formed somewhat like a limpet-shell, covering the apex. The 
arched central plate on the upper side was nearly opposite the flat 
central plate on the under one, but, as shown in the original restora- 
tions of the Pterichthys published seven years ago in my little work 
on the Old Red, it was surrounded by five, instead of four plates. The 
arched back must have been of very considerable strength ; a thick 
band of bone, fig. 1, 2, ran along the top of the sides, 7. e. the walls, 
from which the arch sprang, as the builder places on the top of his 
wall a strong wall-plate ; and each of the plates that rose from them to 
form the arch had also a thickening of the bone at its edges, which 
seems to have served the purpose of rafters springing from these wall- 
plates, or rather that of the ribs of a stone roof. And now for my 
restoration. The carapace consists of eleven plates in all, five on the 
ventral and six on the dorsal superficies. The strongly arched plate a, 
—as I have found, at the expense of some labour and an indifferent pen- 
knife,—overlapped the plate ¢ ina broad squamous suture, of an irre- 
gularly curved form ; but it was in turn overlapped by plate d, in a 
suture of nearly similar form. Again, in a comparatively narrow band. 
of suture, that narrowed still more towards the centre, it overlaid the 
anterior end of plate 6, while yet again this plate on its side overlaid 
with a broad edge plate d, thus locking it fast. The diagonal rib, 
fig. 3, which you describe as traversing plate d, was, I find, prolonged 
to the apex of plate a. By far the strongest portions of the carapace 
were the bands which laced it, ring-like, at both its anterior and pos- 
terior openings (fig. 1, 1); and the thickest and stoutest parts of these 
were in what an architect would term the groinings. The anterior 
ring of the cuirass ran along the edge of the ventral plates e, e and 
the dorso-lateral plates c, ec, but it did not continue along the an- 
terior edge of the plate a. The head was at this central pomt dove- 
tailed to the body by the occipital plate; and it was on the male 
dovetail attached to c¢ that the ring was continued. This insertion 
of the head was both dovetail and keystone,—dovetail to the head, 
and keystone to the ring ;— the dovetail in the plan, if I may so speak, 
being a keystone in the section. ‘The wall-plate fig. 1, 2, was re- 
stricted to plate d. It occurs in the line in which pressure applied 
VOL. IV.—PART I. ZA 
