of the cranium, that it is still perforated by the third (ib. tr) and second divisions of the fifth or trigeminal nerve.

In tracing the alisphenoid downwards through the mammalian series, we cannot but be impressed with the conviction of its true character and importance as an essential part of the cranium, from its constancy in the formation of its walls, and by observing that, whilst the share which the squamosal takes in them progressively decreases,-until in the sheep, for example, it is quite excluded

Vertical longitudinal section of the cranium of a sheep (Ovis Aries).

from the cranial cavity, that of the alisphenoid (fig. 7, 6) increases as the cavity itself diminishes in size; and, further, that this increase is not accompanied with any material change in the relative size of the alisphenoid to the basisphenoid. The share which the alisphenoid takes in forming the anterior boundary of the otocrane increases; as does also the extent of its superior connections, especially of that with the parietal (7). It is important, in tracing these modifications, to note, also, the change in the relative position of the foramen ovale in the mammalian series. In Man the foramen ovale (fig. 6, tr) is close to the hinder border of the alisphenoid; and in some quadrumanes the third division of the fifth escapes through a notch in the same border. This position of the foramen ovale relates to the alisphenoid being pushed forward by the intrusion not only of a large ossified petrosal (16), but of a still larger squamosal (27). In the sheep, however, the foramen ovale is no longer at the posterior margin; but, the alisphenoid, having retrograded by the recession of the squamosal towards its more normal exterior position in the vertebrate series, the third division of the trigeminal now perforates its middle part (fig. 7, tr). It may be observed that, concomitantly with this retrogradation of the alisphenoid, the orbito-sphenoid (ib. 10) acquires larger proportional dimensions than in Man (fig. 6, 10).

In the bird the alisphenoid (fig. 8, 6) is recognizable by the repetition of the connections which it presented in the sheep; the squamosal being quite excluded from the cranial parietes, and, indeed, never again presenting itself in the capacity of a cranial bone in any of the oviparous vertebrates. The alisphenoid (fig. 23, 6) is in contact posteriorly with the petrosal (ib. 16), which soon becomes anchylosed with it, as well as with the exoccipital (2), mastoid (8), and other bones forming the cavity for its reception, in all birds. The alisphenoid further manifests its true homology in the bird by its other constant character of transmitting the third and also the second or maxillary division of the trigeminal nerve; which divisions, in the young ostrich, I

Partly disarticulated cranium of a young ostrich (Struthio camelus), natural size. found distinctly perforating the middle of its lower border (fig. 8, 6, tr). The alisphenoid is deeply impressed by the chief ganglions of the mesencephalon, viz. the optic lobes. The prosencephalon or hemispheres are still defended principally by expanded parietals (ib. 7) and frontals (ib. 11)*.

In the crocodile these spinal elements of the cranium are much restricted in their development, and a larger proportion of the hemispheres is defended by the orbitosphenoid (fig. 9, 10), which here surpasses the alisphenoid (ib. e) in size. This, however, still performs its essential and characteristic func

Vertical longitudinal section of the cranium of a crocodile (Crocodilus acutus).

tions of protecting the sides of the mesencephalon, and giving issue to the chief part of the trigeminal nerve. Owing to the diminution in size of the

*The right frontal has been removed to show better the extent and connections of the orbitosphenoid (10) and the prefrontal (14).

petrosal (16), and the retention by a great proportion of this capsule of the acoustic labyrinth of its primitive cartilaginous state, it occupies a smaller interval between the alisphenoid (6) and exoccipital (2). It no longer protrudes as a large bony wedge (as in figs. 6 and 7, 16) into the cranial cavity, but permits the alisphenoid to come into connection with the exoccipital. The result of this further retrogradation of the alisphenoid, in regard to the relative position of the outlet of the third division of the fifth, is analogous to that which occurs in the sheep. We saw in that mammal, through the recession of the squamosal, the foramen ovale advanced from the posterior to the middle part of the alisphenoid; in the crocodile, through the further removal from the cranial cavity of the interposed petrosal, the foramen ovale is advanced to the anterior border of the alisphenoid; which border, in fact, it notches, the nerve escaping by a common foramen or trou du conjugaison' between the alisphenoid and the orbitosphenoid, the hole, however, being principally formed by the alisphenoid (fig. 9, tr). This position of the 'foramen ovale' loses all its value as an argument in favour of the petrosal character of no. 6, by analogy with the position of the foramen ovale in man or the ape, when we take into consideration the necessary consequences of the successive withdrawal of the squamosal and true petrosal from the inner surface of the cranium in descending to the reptiles. The orbitosphenoid (fig. 9, 10), notwithstanding its great relative size, retains all its essential characters: it is perforated or notched for the exit of the optic nerves (op) and first division of the fifth pair (s); it rests upon the presphenoid (9) below, and likewise, through its backward development, partly upon the basisphenoid, and it articulates with the frontal (11) above, and also through the same backward extension with the parietal (7); it constitutes the anterior border of the lateral bony parietes of the cranium, which are interrupted by the orbits, and separated by their interposition in saurians and fishes from the rhinencephalic part of the cranial cavity (at 14, fig. 9). The characters, in fact, of the orbitosphenoid are so clearly manifested in the crocodile, that Cuvier, having been led by the increased share, as compared with mammals, which the crocodile's alisphenoid (fig. 9, 6) takes in the formation of the otocrane, to regard it as the petrosal, and yet perceiving the essential characters of the orbitosphenoid in the bone (ib. 10) anterior to it, was driven to the conclusion that that bone represented both orbitosphenoid ('aile orbitaire du sphénoïde') and alisphenoid (aile temporale du sphénoïde). The cold-blooded crocodile, however, is not exactly the animal in which we should expect to find so unusual an instance of obliteration of sutures, as that between the alisphenoid and orbitosphenoid *. The actual and most characteristic modification of the orbitosphenoid in the crocodile's skull, is its retrogradation together with the alisphenoid, or rather the maintenance of its normal connection therewith by increased antero-posterior development, whereby it comes into communication above with the parietal (7) and below with the basisphenoid (5); whilst the alisphenoid, in like manner, gains a connection with the supra-occipital (3) above and the basioccipital (1) below; although it still retains its more normal relations with the parietal, and rests in great part on the basisphenoid (5), as the orbitosphenoid rests in great part upon the pre-sphenoid (9.) The superior connec* No one better appreciated the characteristic persistence of the sutures in the crocodile than Cuvier, when his attention was not diverted from it by a favourite hypothesis. "Le crocodile a cela d'avantageux à l'étude de son ostéologie, que ses sutures ne s'effacent point, du moins n'en a-t-il disparu aucune dans nos plus vieilles têtes," is the remark with which he commences his article on the determination of the bones of the head of that reptile (Ossemens Fossiles, 4to. v. pt. ii. p. 69): but at p. 76, a suture is assumed to be effaced, which is present in most mammals and all cold-blooded vertebrates, where a wider space does not intervene between the alisphenoid and orbitosphenoid.

tions of the orbitosphenoid and alisphenoid are always less constant than their inferior ones. By these latter characters, and still better by their nerveoutlets and their relations to the primary divisions of the encephalon, are they rightly and truly determinable. The German authors who have followed Cuvier in his views of the special homology of the alisphenoid in reptiles, are more consistent than the great French anatomist in regard to the alisphenoid of fishes. Dr. Hallmann, accepting Cuvier's characters of the petrosal, taken from its internal position and lodgement of the whole or part of the labyrinth, naturally applies them to the alisphenoid in fishes, and adds to the grounds for regarding that bone as the 'petrosal,' that it is in some fishes perforated by the opercular branch of the great trigeminal nerve†. But, admitting the homology of the opercular nerve with the facial nerve of mammals, yet its wider homology and essential character as a motor division of the great trigeminal nerve must not be lost sight of: its origin in close contiguity with the great sensory portions of the trigeminal in fishes accords better with the character of that nerve as the great spinal nerve of the brain, than it usually presents in higher classes; and it is surely no important departure of the alisphenoid from its normal character, that it should give exit to both motory and sensory divisions of the great nerve with which it is so intimately associated from man down to the fish. Indeed, the progressive withdrawal of the bony petrosal from the interior of the skull and the concomitant backward extension, or retrogradation of the alisphenoid, ought to prepare us to expect that nerves which traverse the petrosal in mammals should perforate the alisphenoid in reptiles and fishes. And so we find in the carp that the glosso-pharyngeal even perforates the posterior border of the alisphenoid; but its origin close to the acoustic and facial nerves in fishes diminishes the force of the argument which might be drawn from this exceptional perforation, in favour of the petrosal character of the alisphenoid I concur entirely with Cuvier and M. Agassiz in their determination of the alisphenoid in fishes; but, if the great share which that bone in reptiles (figs. 9 and 10, 6) contributes to the formation of the otocrane, if the anterior position of the foramen ovale, and the superior connection of the bone with the supra-occipital, are proofs (as Cuvier believed) of its homology with the petrosal in the class Reptilia, they ought also, as Hallmann and Wagner contend, to establish the same special homology of the bone (6) in the class Pisces. But none of these are essential characters of the petrosal. The petrosal is a contentum and not a paries, or any part of the parietes of the cranial chamber or otocrane lodging the organ of hearing: it is the outermost tunic, membranous, gristly, or bony, of the labyrinth or essential part of the acoustic organ. Had the above-cited anatomists clearly appreciated the general homology of the petrosal, they could scarcely have failed to detect its special homologies in the vertebrate series. Cuvier was evidently guided to the true determination of the alisphenoid in fishes, less by its own essential characters, than by observing in certain fishes, the perch and cod for example, a partial ossification of the acoustic capsule, to which, therefore, he assigned the name 'rocher.' And, having thus satisfied himself of the existence of the homologue of the 'pars petrosa,' &c., he could not but assign to the bone which rested below upon the basisphenoid, which protected laterally the optic lobes and gave exit to the third division of the trigeminal nerve, the name of 'grande aile du sphénoïde.' But all these characters equally coexist in the bone which Cuvier calls rocher' (petrosal) in the crocodile and other reptilia. He was not aware, however, that in both gavials and crocodiles a distinct ossicle, the veritable homologue of the intra-cranial pyra

Ossemens Fossiles, 4to, t. v. pt. i. p. 81.

+ Der vergleichende Osteologie des Schläfenbeins, p. 61.

midal-shaped petrosal of mammals and birds, makes its appearance between the alisphenoid, exoccipital and basioccipital, as at 16, fig. 9. Here, however, it is necessary to offer a few observations on the sense in which I use the term 'petrosal' as applied to that ossicle.

The petrosal, properly so called, considered in its totality, as the immediately investing capsule of the labyrinth or internal organ of hearing, is wholly cartilaginous in many fishes and saurians, and in all batrachians, ophidians and chelonians, and is contained in a cavity or orbit (otocrane) which most, or all of the elements of the occipital and parietal vertebræ concur in forming. A part of the ear-capsule remains cartilaginous in the crocodile; but several portions become ossified around the semicircular canals and rudimental cochlea, which ossifications contract slender adhesions to the smooth otocranial surfaces of the supraoccipital, exoccipital and alisphenoid; and to one of these portions (on the principle on which Cuvier applies the term 'rocher' in fishes) the name petrosal might more particularly be given, as it is more distinct and moveable than the other partial ossifications of the capsule, and contributes to form the 'meatus internus' towards the cranial cavity, surrounds nearly the whole of the 'fenestra rotunda', and one-half of the 'fenestra ovalis' towards the tympanic cavity. Looking upon the inner surface of the lateral walls of the cranium (as at fig. 9), one sees at the bottom of the T-shaped suture uniting the otocranial lamina of the exoccipital, alisphenoid, and supraoccipital bones, a fourth osseous element (16), presenting a convex extremity towards the cranial cavity, and completing, with the exoccipital, the lower half of the foramen for the nervus vagus. If this little bone be pressed upon with a needle or probe, it yields and moves, being divided by smooth harmoniæ from both the exoccipital (2) and alisphenoid (6).

The protuberance in question, which thus projects into the cranial cavity, is the rounded angle of the border of the inferior plate of the petrosal, which joins the exoccipital. This lower horizontal plate of the petrosal forms the upper wall of the 'fissura lacera posterior,' and the lower wall of the 'fenestra cochleæ': the fore-part of the horizontal plate bends upwards, twisting and expanding into a vertical oval plate, articulated by its anterior surface to a corresponding sutural surface of the alisphenoid. The lower margin of this plate forms the upper boundary of the 'fenestra cochleæ,' and is continued into a thin plate of bone which divides the 'fenestra cochleæ' from the 'fenestra vestibuli' above. This thin plate of the petrosal joins and is usually anchylosed to the exoccipital: it is the only part of the true petrosal noticed by Cuvier, who describes it as a slender filament of bone which separates the two fenestræt. Seen edgewise, looking into the tympanic cavity, the plate appears like a filament: and this plate forms the sole connection, when any exists, between the petrosal and the exoccipital. I have always found the sutures persistent between the petrosal and the alisphenoid. The upper border of the 'fenestra vestibuli' is formed by a petrosal, or rather otocranial, process of the alisphenoid.

The part (fig. 9, 16) entering into the formation of the lateral walls of the brain-case, and which is here specially indicated by the name of 'petrosal,' seems to have been overlooked: it is, however, relatively to the alisphenoid or exoccipital, as large as is the petrosal (Cuvier's rocher) in the perch it has a true osseous texture, and is quite distinct from the lenticular mass of calcareous matter in the adjacent cochlear chamber which Cuvier compares to starch ('amidon durci').

* Suture à trois branches, Cuvier, l. c. p. 165.

† Du côte de la caisse la paroi est percée de deux fenêtres transversalement oblongues et séparées par un filet mince." l. c. p. 82.