of the human body and that of quadrupeds, birds, and fishes, and shows how, by gradual changes from a horizontal to a vertical attitude, it is possible "to transform a cow into a bird, a quadruped into a man, &c." Finally this striking idea is seen pervading the writings of Vicq-d'Azyr, who says that "nature seems always to work after a primitive and general pattern, from which she deviates only with reluctance; that we observe every where those two characters which seem impressed on all beings, that of constancy in the type and variety in the modifications."*

Nevertheless, this opinion of a constant uniformity of design, of an astonishing analogy, of a primitive and general pattern, rested as yet only on a perception vague and more or less confused. And it is in our own times alone that this complicated question of the analogy of structure has been disentangled and divided; that it has adopted as a field of discussion determinate and precise facts; that, become a positive question, it is capable of being discussed in a rigorous and detailed manner.

This question has been termed the question of unity of organization; it might quite as well have been termed the question of variety of organization; all depends in effect on the point of view under which it is considered; for since there are different species of animals, unity here supposes necessarily a certain variety; and since, on the other hand, these different species all resemble one another, at least on that common ground which consigns them to the same kingdom, it is evident that this variety necessarily supposes also a certain unity or conformity. The true title of the question, or rather its true object, was therefore the determination of the limits where stop, by turns, the resemblances and differences in the organization of animals; an organization at once so similar and so varied. Once divided, as I have just said, the question has taken quite another aspect. The general resemblance of animals has been no longer concluded from some particular resemblances, nor limited to certain branches, to certain classes. As regards the osseous system, for example, it had been presently perceived that, only pertaining to vertebrate animals, this system can only yield results limited to that embranchment, to that type. The resemblances of the osseous system which testify so strongly to a common ground-plan, to a unity of structure, testify to them, therefore, only in reference to the sole type which possesses an osseous system, the type of vertebrates.

Considered collectively, the osseous system forms the skeleton, which is divided into several parts: the apparatus of the vertebræ, that of the cranium, those of the face, of the ear, of the hyoide, of the opercula, of the ribs, of the sternum, of the shoulder, of the pelvis, of the limbs. Now, there is not a member of this diversified apparatus which does not, in the different classes, vary in the form, the number, the complication of the pieces which constitute it. For the most part, and saving the variations just spoken of, they are reproduced in all. There are, however, some which are wanting in such and such a class; there are some which are the exclusive attribute of a single one. The question is to see what is the particular character of each apparatus in each class; that is to say, of what pieces it is there composed, what there is the form and the combination of those pieces. Now, such an examination shows that, among all these parts of which the skeleton is composed, some are essential, and hence more constant; others accessory, and hence more variable; that the vertebræ, the cranium, which lodge the spinal marrow, the encephalon, may readily vary from one class to another in the number and form of their bones, but are found in all; that, on the contrary, the ossicles of the ear, the opercula, the limbs, &c., all accessory and subordinate parts, may be wanting, and are so in effect, when the conditions of audition, of respiration, of locomotion, are no longer the same. These are analogies, then, graduated like the importance of the parts which present them; each part has its proper limits, both of variety and analogy; each should be studied apart; and it might hence be said that there is a particular * See especially his memoir " Sur le parallèle des extremités. &c."

comparative osteology of each osseous apparatus, as there is a general comparative osteology of the whole system.

The cranium, that most complicated apparatus of the skeleton, presents, in all mammifers, a composition very nearly the same; we may there follow each bone, from man to the quadrumana, from the quadrumana to the carnivora, to the rodentia, to the edentata, the pachydermata, the ruminants, the cetacea; everywhere are recognized frontals, parietals, occipitals, temporals, the sphenoid, the ethmoid; and they are everywhere recognized as well by their position as their use. It is as much if the interparietals appear to be wanting in certain species. It is the same with the face. The bones of the nose, of the cheek, of the jaws, of the palate, &c., are never wanting. The lachrymals alone fail in the phocæ, the dolphins, &c. All other differences of number are but apparent, and result only from the greater or less promptness with which, according to the species, the bones or parts of bones, constantly separated in the first stage of life, unite and are confounded in adult age. It is thus that, according to the species, the occipital, the parietal, the sphenoid, the temporal, &c., appear sometimes single, sometimes double, triple, or quadruple; but when we recur to the foetus the occipital is always divided into four parts, the parietal into two, or rather into four, counting the inter-parietals, which, in the end, constantly become united therewith, the temporal into four, the body of the sphenoid into two, &c. Thus, in the mammals, there is a normal number for the bones of the cranium; and when this number appears masked by the obliteration of the sutures in the adult state, the primitive division is always reproduced and restored in the foetal state; and what I say of the bones of the cranium may be said also of the bones of the face, and of their more numerous subdivisions in the earlier stages.

It would naturally be curious to see whether this singular analogy was maintained in the other classes, in the birds, the reptiles, the fish; whether the same number of bones is there everywhere reproduced; whether, masked in the adult state, it would appear in the foetal; whether, in fine, reptiles and fish, in which the bones of the cranium are always much more numerous, could be regarded as corresponding in this respect to the early age of birds and mammals. This interesting question was successively treated by M. Cuvier in reference to reptiles (Reserches sur les ossemens fossiles, tome 5,) and to fish (Histoire naturelle des poissons ;) it will suffice to indicate here the manner in which he has resolved it relative to reptiles.

The reptile whose head presents the most striking traits of conformity with that of the mammifers is the crocodile; from the crocodile M. Cuvier passes in succession to the tortoises, the lizards, the serpents, and finishes with the batrachians, which conduct from reptiles to fish, as the crocodile from reptiles to mammals. The head of the crocodile is composed of a much greater number of bones than that of the adult mammifer, but by recurring to the foetus of this last class we recognize in the head of the crocodile and that of the mammifer very nearly the same number of bones. Thus M. Cuvier, after having found in the crocodile, and in the same place as in the mammals, the intermaxillary bones, the maxillary, the nasal, the lachrymal, the jugal, the palatine, the ethmoid,* the body of the sphenoid, the parietal, finds also, and again in the same place, the occipital, but divided into four parts, as it is in the foetus of mammals, the great wings of the sphenoid, vestiges of its lesser wings, its internal and external pterygoid wings, but all these parts separate from the body of the bone, as they all are, except the last,† in the mammal at its earliest age; finally a temporal

With its cribriform lamina, its lateral wings, its superior cornets, its vertical lamina, but all these pieces or dependences of the ethmoid, in great part, in a cartilaginous state.

For this reason M. Cuvier gives the special name of transverse bone to the external pterygoid apophysis, which, in the mammals, is not at any age effectually separated from the great temporal wing. It is, therefore, not properly a new bone, but a dismemberment of the sphenoid, as the frontal bones, anterior and posterior, are dismemberments of the frontal.

bone, but composed of four bones, as it is in the fœtal mammals, the squamosal, the mastoid, the cavity of the tympanum, and the petrous bone. There remained only the bones which correspond to the frontal to reduce to analogy, but these bones are six in number in the crocodile, and as the frontal of mammifers is never divided but into two, M. Cuvier was obliged to admit here a peculiar dismemberment of this bone, a dismemberment which, in the crocodile, or, to speak more generally, in the greater part of oviparous vertebrates, subdivides 'each of the two frontals of mammals into three others, the principal, the anterior, and the posterior frontals.

This determination of the bones of the head of the crocodile, compared with those of the head of mammals, being once established, it is easy to refer to it, as a sort of type, the bones of the head of all other reptiles, particularly tortoises, lizards, and serpents. Thus, with due regard to the differences of form and proportion, the greater part of the bones of the crocodile reappear in the head of the tortoise; but this head wants the nasal bones, which are here represented only by cartilaginous laminæ, the transverse or external pterygoid and the lachrymal bones. Moreover, the parietal, which is single in the crocodile, is double in the tortoise; but in the lizards this parietal again becomes single, the lachrymal and transverse bones reappear, a new bone is disclosed which M. Cuvier calls columella, &c.; all but slight differences, which do not hinder us from recognizing throughout the predominance of one same plan in the heads of the crocodile, the tortoise, and the lizard.

A new and more difficult study commences with the batrachians. First, the general composition of the head is singularly simplified. We find here only the two lateral occipitals, with neither upper occipital nor basilar; a single sphenoid with neither temporal nor orbital wings; a single bone replaces at once the principal frontal and the ethmoid; there are no posterior frontals, but there are two anterior frontals, two parietals, and two petrous bones. Nor is the face less simplified, for the transverse forms only one with the pterygoid, the temporal but one with the tympanic, and there is no mastoid. The cranium of the frog, therefore, has but 10 bones, one ethmoid, two frontals, two parietals, two occipitals, one sphenoid, two petrous bones; its face has but 16, two intermaxillars, two maxillars, two nasals, two palatines, two vomers, two pterygoids, two tympanics, and two jugals or zygomatics. In all, its head has but 26 bones, while that of the crocodile has nearly 40. And this difference of number presents itself in each particular apparatus of the face; thus the lower jaw of the crocodile has six bones on each side, and each side of the jaw of the frog has but three, &c.

I have said that the apparatus of the vertebræ is, with that of the cranium, the most constant; each vertebra may itself be considered as a small distinct арраratus, composed of a certain number of bones, which is not the same for all the vertebræ in each species, nor for each vertebra in all the species; the atlas of the crocodile has six bones, its axis has five; the atlas of the tortoise has only four, that of the monitor three, &c. But it is chiefly by their total number that the vertebræ vary from one class to the other, and even in the different orders, the different genera of each class. Not to depart here from the reptiles, the crocodile, for example, has 26 vertebræ-7 cervical, 12 dorsal, 5 lumbar, and 2 sacral; 200 are counted in the adder, the boa, &c.; the frog has but nine. As regards other apparatus of the animal system, being only accessory, the greater part may be wanting, and is wanting, in effect, in such or such a class, such or such an order, such or such a genus, &c. The hinder extremities fail in the cetacea, both the anterior and posterior extremities in the serpent, the ribs are absent in the frog, the auricular apparatus in fishes, &c.

Nothing is better calculated to give a just idea of the manner in which a certain general conformity is combined in certain cases with all the variations of detail than what is seen in the shoulder and the sternum. The shoulder, which is composed, in the mammifer, of but one bone, the scapula, or of two, the scapula and

the clavicle, has always three in the bird: the scapula or omoplate, the clavicle, and the coracoid bone; it has only two in the crocodile, the scapula, and the coracoid, the true clavicle being wanting; the three bones reappear in the lizards; there are two in the tortoise, the scapula and coracoid, or perhaps three, for there are traces of a clavicle; there are certainly four in the frog, the clavicle, coracoid, and an omoplate, divided into two pieces; and, what is remarkable, it is precisely of these two pieces of the omoplate that the shoulder of fishes is composed. The sternum in the crocodile is a single bone; in the tortoise it is always composed of nine pieces; it recovers in the lizard the simplicity which it exhibits in the crocodile; it has but two ossified pieces in the frog; a sort of sternum is hardly discoverable in certain fishes; in the mammals, on the contrary, it is highly developed; here we count as many as seven, eight, nine pieces, placed ordinarily on a single line; and as to birds, there are five pieces in the gallinacea; there are not more than two in ducks; its composition is again changed in pigeons, in sparrows, in birds of prey, &c. Thus the sternum not only varies from one class to another, it varies in the same class, and that in the very class of birds where, in general, unity or conformity of organization is so constant and conspicuous.

But, in reference to this question of osteological unity in vertebrate animals, there are two apparatus which have a peculiar importance: these are the auricular and the hyoid apparatus. By the auricular apparatus we here designate a chain of small bones, placed within the tympanum of the ear, and which extends from its membrane to the fenestra ovalis. In the mammals we count always four of these small bones, the malleus, the incus, the lenticular, and the stapes; in the birds we find but one, formed of two branches, one of which adheres to the tympanum, the other presses upon the fenestra ovalis; in like manner a single ossicle replaces, in the crocodile, the four small bones of the ear of mammals; it is a stapes still more simple than that of birds; * there is but a single ossicle in the tortoise, the lizard, the serpent; in the frog, the auricular chain might be pronounced somewhat complex did it not remain in great part cartilaginous; lastly, in the salamanders, the sirens, the protei, the last auditory ossicle itself, the stapes, is reduced to a simple cartilaginous plate. From this to the complex apparatus of the mammals is certainly a wide interval, and when we thus follow step by step this successive simplification, when we arrive thus at that final reduction of the whole apparatus to a simple cartilaginous plate, we recognize the full force of M. Cuvier's opinion that this apparatus, after having disappeared in the aerial vertebrates, is not all at once restored in the class of fishes, there to form the opercula, and that these opercula' are consequently a special apparatus appropriate to this latter class.

The facts which concern the inverse progression of the hyoid apparatus, that is to say, its gradual development from the mammals to fishes, are still more important, and, in relation to theories of the skeleton, more decisive. In man this apparatus is composed of five parts: of a body, of two branches or anterior processes which suspend the hyoid to the cranium, and of two posterior ones which suspend the larynx to the hyoid. Even in the mammals the apparatus undergoes great modifications, depending on the form of its body, the soldering of this to the posterior branches more or less promptly, the number, shape, and proportion of the anterior branches. In birds these anterior branches are no longer attached to the cranium, but simply pass around and behind it; to the back part of the body of the bone is soldered a single slender bone, on which rests the larynx, and which, in itself alone, represents the two posterior branches; in front is another bone which penetrates into the tongue, being the lingual bone.

The hyoid of the crocodile is one of the most simple. Its body consists of a

*We might, in truth, give the name of malleus to that branch which in birds and crocodiles is inserted in the membrane of the tympanum, but still there would be neither incus nor lenticular ossicle.

*

*

thick and broad cartilaginous plate, without any very distinct vestige of the posterior branches, and a single ossified piece, representing the anterior branches. The hyoid of lizards is much more complex, and that of the tortoise even more so than that of the lizard. But it is in the batrachians, chiefly, that the hyoid acquires importance, and thus leads by degrees to the hyoid of fishes, so rich and complicated. To explain this richness of the hyoid apparatus in fish, recourse had been had to a pretended intercalation which has taken place of pieces borrowed at once from the sternum, the larynx, and the ribs. It will be perceived that the metamorphosis of the frog, which, in its first period, respires by bronchiæ or gills, like the fish, which at a later period respires by lungs, like the land animals, and the bronchial apparatus of which changes by degrees, and visibly, into a true hyoid, should have settled all difficulty in this respect. M. Cuvier, therefore, made this singular metamorphosis a detailed study; he followed it in its whole progress; he saw the branchiæ and branchial arches successively fall; he saw the proper hyoid of the adult frog take shape proportionably; and at no time, even at that of the greatest complication, when the branchial arches and the branchiæ existed, neither did the sternum nor the larynx take, nor could they take, any part in this whole composition; for the branchial apparatus still very distinctly exists, with all its parts, which are clearly to be seen, as well as the larynx, with its dependent lungs, and the sternum, with the bones which rest against it. The hyoid of the salamander is metamorphosed like that of the frog, and the branchial apparatus, in the same way, still very distinctly subsists, although the larynx, the lungs, and the sternum, are also present; and all this acquires new force from what is observed so plainly in the axolotl, the proteus, the siren, animals in which the branchial apparatus exists simultaneously, and in a constant manner, with the larynx, the tracheal artery, &c. The branchial apparatus is, therefore, only a more complex hyoidian apparatus, and not one resulting from the combination of parts foreign to it, and derived from neighboring organs.

Each apparatus has, therefore, its proper constitution; it has its marked increments and decrements; its parts change from one class to another in form, in number, in complexity; and it is these very changes which determine the organic characters of classes, of orders, of genera, of species. What, then, must be understood by unity, or, to speak more exactly, by conformity of organization, by conformity of plan, in the vertebrate animals, at least in what regards their osseous system, if not an assemblage of graduated analogies, more constant in the essential apparatus, more variable in the accessory, and of which the limit cannot be given, for each apparatus, except by the direct and consecutive study of all the modifications of that apparatus, in all the classes? Now this consecutive study of an apparatus through all the classes, and of all the graduated modifications which it undergoes from one class to another, is precisely what constitutes the most distinctive feature of Cuvier's method, and the point which ought perhaps most to fix the attention of right-minded inquirers; for it is on the rigorous and special adaptation of the method to its object, that depends the exactness of the results. Now, what is here the question? Of following, of recognizing an apparatus through all its metamorphoses of number, of form, of complication of parts. And is it not plain that to lose sight of a single intermediate metamorphosis would suffice to render impossible all recognition of those which follow, would be to lose the thread which connects one with the other, nay, to lose the apparatus itself? The principle of successive and graduated modifications, employed by M. Cuvier, is one, then, of the most fruitful as well as most ingenious means of investigation with which he has enriched science, and the only one which can give, in a sure and precise manner, both the determination of each apparatus, and the limit of its analogies or its dissimilarities in each class.