dS 
ON THE VERTEBRATE SKELETON. 923 
points of resemblance between the prefrontals of the crocodile and those of 
the fish are the mesial approximation and junction of their descending (neu- 
ropophysial or rhinencephalic) plates—the most constant and important parts 
of the bones in question. 
If the frontal of the ruminant or other mammal were expanded only at 
the parts corresponding with the detached bones called “frontaux anté- 
rieures” in the crocodile, there might then be a primd facie probability that 
such expansions were connate parts, dismembered in the crocodile’s skull. 
But the vastly increased lateral as well as anteroposterior development, and 
the more or less vertical convex expansion of the frontal in the highest 
vertebrate class, naturally indicate, in the first place, an inquiry into the 
concomitant modification of the nervous centres by which the development 
of that bone is mainly governed; and if such modification should then be 
found to exist, in the cerebrum, for example, which, from the ascertained 
correlative progress of the frontal in other classes, ought to cause or be 
associated with such a general development of that bone as characterises the 
skull in the mammalian class, it must surely be superfluous and gratuitous 
to explain that development by the hypothesis of a coalescence of another 
essentially distinct element of the cranial parietes: especially if that element 
be proved by a similar tracing of its relations to the progressive development 
of the cerebral centres, to have as essential and exclusive a dependence 
upon the rhinencephalon as the frontal bone has upon the prosencephalon. 
_ The position of the upper peripheral part of the prefrontal in the situation 
in which it is seen in the crocodile, is, in fact, the least constant and import- 
ant of the characters of that bone. in the bull-frog, for example, the ex- 
posed part of the prefrontal is mesiad of the conjoined parts of the nasals 
and frontals instead of being lateral: in the sword-fish the prefrontals barely 
appear, and in the python they do not appear at all upon the upper surface 
of the skull; but they retain in each their more typical neurapophysial po- 
sition, with all their more constant and essential characters. The enormously 
developed frontal of the mammal masks these characters, and usurps the 
less constant and least important one, viz. superficial position, on which alone 
Cuvier insists as proving the prefrontal of the crocodile, with its complex 
functions and connections, to be such a dismemberment of the true frontals 
if the ruminant, as may be marked off with the pen on the upper surface of 
the skull. 
The descending [rhinencephalic] plates of the prefrontal in the crocodile 
(fig. 9, 14) are subcompressed in the axis of the skull, and expanded laterally, 
especially at their upper part ; where, in the alligator, I find them forming a 
shallow cup, concave forwards for the lodgment of the cartilaginous olfactory 
capsule,—of that part, namely, which is ossified in mammalia, and there de- 
veloped into the great labyrinth of the superior turbinals and ethmoidal cells. 
The vertical plates, continued forwards from the prefrontals, which extend 
above to the nasal suture and descend into the vomerine groove below, to aid 
in forming the ‘septum narium,’ are cartilaginous in the crocodile: they are 
more or less ossified, and form the ‘lamina perpendicularis ethmoidei’ in 
mammals. The median plate, dividing the olfactory nerves at their exit, and 
developed backwards as a partial septum of the rhinencephalic chamber of 
the cranium, and continued into the simple interorbital septum of the croco- 
dile, also remains cartilaginous: when ossified in mammals, it forms the 
‘crista galli.’ Now not one of these cartilaginous representatives of the parts 
of the compound bone called ‘ ethmoid’ in anthropotomy, is united or con- 
nected with the portions of the frontal in mammals which Cuvier has assumed 
to be the homologues of the prefrontals in the crocodile ; those bones being 
