of the hæmal arch of the nasal vertebra with that of the frontal vertebra, and consisting of the two styliform ossicles (26 and 27) which extend from the hæmapophysis, 21, 21", to the pleurapophysis, 28: the essential relationship of the compound ray, 26 and 27, with the nasal vertebra, is indicated by their becoming confluent with its hæmapophysis (at 22"), whilst they always maintain an arthrodial articulation with the pleurapophysis (28) of the succeeding vertebra.

The bones of the splanchno-skeleton intercalated with the segments of the endoskeleton in the bird's skull are the petrosal (16), between the neural arches of the occipital and parietal vertebra, early coalescing with the elements of those vertebra with which it comes in contact: the sclerotals (17), interposed between the frontal and nasal neural arches: and the thyro-hyals (47), retained in connection with the debris of the hæmal arch of the parietal vertebra, H II. The olfactory capsule remains cartilaginous. The dermal boue (73) is well-developed and constant: a second supraorbital dermal bone is occasionally present. All the endoskeletal bones manifest, under every adaptive modification, the segmental arrangement, and it is difficult to contemplate the repetition of the arrangement of the cranial bones around the primary segments of the encephalon in the series of arches closed respectively by the bones N 1, N 11, N 111, N IV, together with that of the corresponding number of arches closed below, at H IV, H III, H II and H 1, without a conviction that the type illustrated in fig. 15 is that upon which these segments of the skull have been constructed. This conclusion might seem forced, in respect to the occipital vertebra, were its displaced hæmal arch and appendages to be viewed without reference to their relative position and connections in the lower vertebrate classes; but it will be confirmed and shown to be agreeable to nature and to the recognised kinds and grades of modification to which the elements of one and the same vertebra are subject, by observing in the young bird the distinct pleurapophysial elements of those cervical vertebræ, beyond which the corresponding elements of the occiput have retrograded, in obedience to the functions which the hæmal arch of that vertebra and its appendages are destined to perform in the feathered class.

Mammals. If the foregoing views of the general homology of the bones of the skull be agreeable to their essential nature, we should expect that the new and additional modifications, in the mammalian class, which tend to obscure those relations would be seated in the appendages and peripheral elements of the endoskeletal segments, or in the capsules and appendages of the special organs of sense.

I have selected with a view to testing such anticipation the skull of a young pachyderm, and, after successively disarticulating the segments in the order in which they have been previously described, I have given a side view of them arranged in correspondence with the figures 23, 22, and 5. (Fig. 24.)

The neural arch of the occipital vertebra, NI, agrees with that of the bird and crocodile in the coalescence of the parapophysis, 4, with the neurapophysis, 2; but the process, 4, now descends from the lower part of the arch, and, as in many other mammals, is of great length. An articular condyle is also developed from each neurapophysis which articulates with the concave anterior zygapophysis of the atlas, and is the homotype of the posterior zygapophysis in the trunk-vertebræ. The centrum (1) is reduced, like that of the atlas, to a compressed plate, and its hinder articular surface is not more

*The skull of the ruminant is perhaps still better adapted to demonstrate the vertebral relations of the cranial bones: that of the sheep is the subject of the diagram for this purpose in the concluding volume of my Hunterian Lectures.'

1846.

X

developed than is the front one of the centrum of the atlas, with which, indeed, it is loosely connected by ligament. The expanse of the occipital spine, 3, has been governed, agreeably with a foregoing remark, by the superior development of the cerebellum.

The hæmal arch of the occipital vertebra is represented, like those of the cervical vertebra, by the pleurapophysial elements only (51); but these, in most mammals, are developed into broad triangular plates with outstanding processes that called 'spine' and 'acromion' is exogenous; but that called coracoid' is always developed from an independent osseous centre (a rudimental representative of the hamapophysis, 52), which coalesces with the pleurapophysis in mammalia, and only attains its normal proportions, completing the arch with the hæmal spine (episternum) in the monotremes.

In many mammals the arch is completed by bones, which are, apparently, the hæmapophyses of the atlas, e. g. in man (fig. 25, 52′), which have followed the occipital hæmal arch in its backward displacement, but not quite to the

same extent.

The diverging appendage, though retaining the general features of its primitive radiated form, has been the seat of great development and much modification and adjustment of its different subdivisions (53-57) in relation to the locomotive office it is now called upon to perform.

With the exception of this excess of development of the appendage, the defective development and displacement of the hæmal arch, and the coalescence of the parapophyses in the neural arch, there are few points of resemblance which are not sufficiently salient between the segment NI, HI in the mammal, and that so marked in the fish (fig. 5). And, if the interpretation of the more normal condition of this segment in the lower vertebrate, according to the archetypal vertebra, fig. 15, be accepted, then the explanation of the nature of the modifications of the special homologues of the constituents of the occipital segment by which that archetype is masked in the mammal, may be confidently left to the judgement of the unbiassed student of homological anatomy.

In commencing his comparisons of the second segment of the skull with the typical vertebra, he will be unexpectedly gratified by finding, in the immature mammal, the centrum, 5, naturally distinct, and the hæmal arch, HII, retaining its natural connections with the rest of the segment, and by means of a more complete development of the pleurapophyses (38) than in any of the inferior air-breathing vertebrates. He may now separate, without artificial division of any compound bone, the entire parietal segment, but he brings away with it the petrified capsule of the acoustic organ, and the anchylosed distal piece (27) of the maxillary appendage, which more or less encumbers and conceals the typical character of the neural arch of the parietal vertebra in every mammal : least so, however, in the monotremes and ruminants. The neurapophyses (0) of the parietal vertebra, like the mesencephalic segment of the brain, are but little more developed in mammals than in the cold-blooded classes: they are notched in the hog and perforated in the sheep by the larger divisions of the trigeminal, and they send down an exogenous process, which articulates and sometimes coalesces with the appendage (24) of the palato-maxillary arch. The neural spine (7), always developed from two centres, often vastly expanded, and sometimes complicated with a third intercalary or interparietal osseous piece, is occasionally uplifted and removed from its neurapophyses by the interposed squamous expansion of the bone 27; but this, which reminds one of the occasional separation of the neural arch from the centrum of the atlas in fishes, is a rare modification in the mammalian class. A still rarer one is the separation of the halves of the parieto-neural spine

from each other by the extension and mutual junction at the median line of the occipital and frontal spines. A specimen of this, in a species of Cebus, which repeats the common modification of the parts in fishes, is preserved in the museum of the Royal College of Surgeons. The parapophysis (s) always commences as an autogenous element by a distinct centre of ossification, as shown in the human fœtus, fig. 11, 8; it speedily coalesces with the petrosal, but otherwise retains its individuality in some of the lower mammals, as e. g. in the echidna (fig. 12, s): or it coalesces with the curtailed frontal pleurapophysis 28, or with the maxillary appendage 27, or with both these and the pleurapophysis of its own vertebra (38), when the complex "temporal bone' of anthropotomy is the result. In most mammals the pleurapophysis (38) retains its primitive independency and rib-like form, with usually the 'head' and 'tubercle'; but by reason of its arrested growth it has been called 'styloid' bone or process. Sometimes it is separated from the short hamapophysis, 40, by a long ligamentous tract, sometimes it is immediately articulated with it, or by an intervening piece. The hamal spine, 41, is usually small, but thick and always single. The rudiments of hypobranchial elements (46) are retained as diverging appendages of the parieto-hæmal arch in all mammals, and have received the special names of 'posterior cornua,' or thyrohyals,' from their subservient relationship to the larynx.

In the frontal segment the centrum, 9, and neurapophyses, 10, very early coalesce. Two separate osseous centres mark out the body (fig. 26, C, 9), and each neurapophysis has two distinct centres (ib. 10, 10), the optic foramina (op) being first surrounded by the course of the ossification from these points. The superior development of the neurapophysial plates (10), as compared with those of the parietal vertebra (6), in most mammals, harmonizes with the greater development of the prosencephalon; but the chief bulk of this segment of the brain is protected by the expanded spines of the frontal (11) and parietal (7) vertebræ, and by the intercalated squamosals (27). And the appendicular piece (27) not only usurps some of the functions of the proper cranial neurapophyses, but, likewise, the normal office of the frontal pleurapophysis (29), in the support, viz. of the distal elements of the hæmal arch (29, 32), which now articulate directly with 27, in place of 28 as in all oviparous vertebrates. The true pleurapophysis of the frontal vertebra (28) is almost restricted in the mammalian class to functions in subserviency to the organ of hearing, is sometimes swollen into a large bulla ossea, like the parapophyses and pleurapophyses of the cervical vertebræ of Cobitis; it is sometimes produced into a long auditory tube, and sometimes reduced to the ring supporting the tympanic membrane. Yet, under all these changes, since its special homology is demonstrable with 28 in the bird (fig. 23) and crocodile (fig. 22) as well as with the teleologically compound bone, 28 a, b, c, d, in the fish (fig. 5), so likewise must its general homology, which is so plainly manifested in the fish, be equally recognised. The frontal hamapophysis (fig. 24, 29, 30), and the corresponding half of the hamal spine (ib. 32) are connate on each side in all mammals, and become confluent at H 111, in most. The hæmal arch of the frontal segment of the skull, as in other air-breathing vertebrates, has no diverging appendage, unless the tympanic otosteals be so regarded, an idea which is not borne out by their development.

The nasal segment (N IV, HIV) is chiefly complicated by the confluence of parts of the enormously developed olfactory capsules (18) in the mammalian class, and its typical character is masked by the compression and mutual coalescence of the neurapophyses, 14. The centrum is usually much elongated, as at 13, and soon coalesces with both neurapophyses (14) and nasal capsules in the hog. The neural spine (15) is usually divided, but is sometimes single,

e. g. in Simia. In the rhinoceros it supports a dermal spine or horn. The pleurapophysis (20) or proximal element of the hæmal arch of the nasal vertebra has its real character and import almost concealed by the excessive development of the second element of the arch (21), which resumes in mammals all those extensive collateral connections which it presented in the crocodile; and to which are sometimes added attachments to the expanded spine of the frontal vertebra, as well as to that of its own segment. The pleurapophysis however, besides its normal attachment to its centrum, 13, sends up a process to the orbit, in order to effect a junction with its neurapophysis which sometimes appears there, as the os planum' of anthropotomy. The hamal spine (22) is developed in two moieties, which never coalesce together, although, in the higher apes, and at a very early period in man, each half coalesces with the hæmapophysis, and repeats the simple character of the corresponding elements (rami) of the succeeding (mandibular) arch.

The appendicular element (24) which diverges from the pleurapophysis (20), contributes to fix and strengthen the palato-maxillary arch by attaching it to the descending process of the parietal centrum (5); with which, in most mammals, it ultimately coalesces. The other elements of the diverging member of the arch correspond in number and in the point of their divergence with those in birds, chelonians and crocodiles. They are two in number, succeeding each other, and both become the seat of that expansive development which is followed by the multiplication of their points of connection; thus the proximal piece ('malar' 26) articulates in the hog not only with the hæmapophysis (21) from which it diverges, but likewise with the muco-dermal bone, 73. The distal piece of the appendage (squamosal, 27) expands as it diverges, and fixes the naso-hæmal arch not only to the frontal pleurapophysis (28), but also to the frontal, parietal and occipital neurapophyses and spines: it also affords, in the hog, as in other mammals, an articular surface to the frontal hæmapophysis (29).

The development of an osseous centre in the cartilage of the snout of the hog, and the homologous prenasal' ossicle in certain fishes, the carp, e. g., might be regarded as rudiments of terminal abortive segments more anterior than the nasal vertebra. The multiplied points of ossification in the vomer have been, also, deemed indications of that bone being, like the vomerine coccygeal bone in birds, a coalescence of several vertebral bodies. Of course, à priori, the segments in the cranial region of the endoskeleton might as reasonably be expected to vary in number in different species, as the segments in the thoracic or sacral regions. I have not, however, been able to determine clear and satisfactory representatives of more than four vertebræ in the skull of any animal; and the special ossifications in the nasal cartilages appear to me to belong to the same category of osseous parts, as the palpebral bones in certain crocodiles and the otosteals.

Man.-Arriving, finally, in the ascensive survey and comparison of the archetypal relations of the bones of the vertebrate skull, at Man, the highest and most modified of all organic forms, in which the dominion of the controlling and specially adapting force over the lower tendency to type and vegetative repetition is manifested in the strongest characters, we, nevertheless, find the vertebrate pattern so obviously retained, and the mammalian modification of it, as illustrated in the preceding paragraph and diagram, so closely adhered to, as to call for a brief notice only of those developments of the common elements which impress upon the human skull its characteristic form and proportions.

The neural arch of the occipital vertebra differs from that of the hog by a much greater development of the neural spine (fig. 25, 3) and a much less