304 PROCEEDINGS OF THE GEOLOGICAL SociETY. ([Apr. 19, 
compose the ribbed and pointed roof of a Gothic cathedral, the arched 
ridge or hump of the back requiring in a somewhat similar way a 
peculiar form and arrangement of plates. The apex of the ridge is 
covered by a strong hexagonal plate fitted upon it like a cap or 
helmet, and which nearly corresponds im place to the flat central 
plate of the under side.”” (The Old Red Sandstone, by Hugh Miller, 
page 49, plates 1 and 2.) It will be seen from the above that Mr. 
Miller distinctly describes the back of the fish as arched, the belly 
as flat, each region being characterized by bony plates peculiar both 
in form and adjustment. Agassiz describes the arrangement of the 
plates on the flat side*, but he considers this the dorsal surface, in 
contradiction to Mr. Miller, who views it as the abdomen. The 
other or dome-shaped surface he does not allude to directly as a 
determinate character ; he nevertheless mentions it incidentally in 
the description of Pterichthys testudinarius as the “ face inférieure”’ 
of the fish, with some hesitation adding, ‘‘It will be necessary to 
ascertain from more perfect specimens whether the longitudinal ele- 
vation traceable along the median line be really a specific character, 
or whether it results from the bad state of the specimen.” This 
feature is more pointedly acknowledged in the description of Pterich- 
thys Milleri, and is truly considered as the “‘face supérieure.’ He 
says, ‘“‘ The carapace rises in the middle into a longitudinal keel, assez 
saillante, which gives to the superior face the form of a roof.” In 
the detailed account of the anatomy of this species, the central plate 
of the inferior face of Pterichthys testudinarius is stated to be 
enormously large, and the corresponding plate in the superior face 
of Pterichthys Milleri to have nearly the same proportional dimen- 
sions. Here then we see two prominent features, common to two 
species of the same genus, assigned to opposite positions im each. 
Whether Agassiz be correct or not in this particular will be con- 
sidered in the sequel; in the meantime it is clear that the pas- 
sages above-quoted afford presumptive evidence that in two species 
of the genus one surface of the fish was more elevated than the 
other, and that so far Mr. Miiler’s description is correct. These 
parts are generally greatly fractured and displaced ; the plane surface, 
on the contrary, is frequently preserved as perfect as when the fish 
was endowed with life. This is not to be wondered at; for in the 
process of decomposition, the destruction of the ternal soft parts, 
and the interstitial animal matter of the sutures would cause the 
dome to collapse, a result not so likely to befall the more even sur- 
face, especially if reposimg on a muddy or sandy bottom. Even 
should the parts retain their natural position until potted in their 
now hardened matrix, the irregularity of their contour would baffle the 
efforts of the geologist’s hammer to disengage them, a plane of far 
less resistance being offered by the more even surface. The hardness 
of the stone and brittle condition of the bones preclude the more deli- 
cate operations of the chisel. The plates however are not unfrequently 
found perfect when disjomted, and, from their characteristic forms, 
* Monographie des Poissons Fossiles du vieux Grés rouge, p. 7, pl. 6. fig. 1. 
