216 REPORT—1846. 
olfactory nerves groove instead of perforate the bones no.14._ But the trige- 
minal still determines the alisphenoid, whether it perforates or notches that 
neurapophysis in its escape: the relation of the alisphenoid to the division 
of the 5th, including the gustatory nerve, and that of the orbitosphenoid to 
the nerve of sight, are not more constant than is the relation of no. 14 to the 
nerve of smell. The differences of connection of no. 1a—‘les autres rap- 
ports’—are not specified by Cuvier, and I know none that affect its essential 
character. 
No. 14 is however the, most anterior of the neurapophysial or lateral 
bones of the true cranium, and is in relation with the anterior terminal divi- 
sion of the encephalon and with the first or anterior terminal pair of nerves. 
Like all extreme or peripheral parts, it is subject, as we should be prepared 
to find it, to a greater extent and variety of modifications than the more 
central neurapophyses. The difference between its connections in the fish 
and that of the cribriform plates and their sustaining basis in man may 
therefore be expected to reach the extremes of possible homology. It will 
be interesting to inquire whether there are intermediate modifications by 
which the nature of that difference may be appreciated, and how many of 
such links are permanently retained in the intervening species. 
We might anticipate the smallest amount of departure from the fun- 
damental vertebrate type, as respects form, size and connections of the bones 
in question, in that class where the principle of vegetative repetition most 
prevails and the archetypal plan is least obscured by teleological adaptatious. 
Adopting the name modified from the phrase applied to these bones by Cu- 
vier in those vertebrata in which they present their most typical characters, 
we find the ‘prefrontals’ in all bony fishes resting below upon the vomer (figs. 
4 and 5,13) and on part of the presphenoid (9), sustaining by their mesial and 
upper surfaces the nasal (15) and fore-part of the frontal (11), affording the 
whole or part of the surface of articulation for the palatine (20) or the palato- 
maxillary arch, and giving attachment exteriorly to the large suborbital or 
lacrymal bone (figs. 22 and 25, 73), when this exists. Besides their protec- 
tive functions, in relation to the olfactory ganglions and nerves, they close the 
cranial cavity and bound the orbits anteriorly. The most constant and cha- 
racteristic connections appear to be with the vomer, nasal, palatine and frontal. 
In the murenoid fishes, where confluence begins to prevail in the cranial bones, 
we find that the prefrontals coalesce with the vomer and nasal, not with the 
true frontal. This fact, though not of a class materially affecting relations 
of homology, is not devoid of significancy in regard to the real character of 
the bone usually described as one of the ‘ deux démembremens du frontal*.’ 
A elew not to be neglected in tracing the homologies of the prefrontals is 
their histological progress, although the value of such embryonic characters 
has been overrated and their application sometimes abused. The substramen 
of their ossification, like that of the exoccipitals, mastoids and post-frontals, 
is a cartilaginous mass, a part of that which M. Dugés has called ‘ cartilage 
cranio-faciale,’ and M. Vogt ‘ plaques protectrices latérales.’ The frontals 
and parietals, being ossified in supra-cranial fibrous membrane with so rapid 
and transitory a cartilaginous change as to have escaped general recognition, 
have been, on that account, rejected from the vertebral or endo-skeletal system 
of bones by Dr. Reichert, and with as little real ground as the rejection of the 
vomer and sphenoid from the same system, because they are ossified in mem- 
brane extended from the under and fore-part of the sheath of an evanescent 
subcranial ‘ chorda dorsalis,’ like the homologous basal ossification beneath 
the coalesced anterior abdominal vertebra of the siluroids. 
* Agassiz, op. cit. i. p. 123. 
