PHILOSOPHICAL ANATOMY 427 porting in the former the second, third, and fourth phalanges, that of the second very rudi- mentary; in the metatarsus the three bones are, except in the penguins, united for their whole length, including also the two tarsal bones ; the rudimentary metatarsal of the great toe is not anchylosed, and is directed back- ward, supporting the hind toe with two pha- langes ; the second toe has three phalanges, the third four, the fourth five, the fifth or little toe being wanting ; by the number of phalanges we know that, in the African ostrich, the ex- ternal or shortest toe of the two by its five joints is the fourth, and the internal the third, longer than the other, though having only four joints. In the mammal skeleton the cranial cavity is expanded, as in birds, chiefly at the expense of the neural spines, frontal, parietal, and occipital ; but in most cases the squamous portion of the temporal bone forms part of the cranial walls; the occipital is articulated to the atlas by zygapophyses or condyles developed from the neural arches, and the haemal or scapular arch is generally far removed from the skull ; the haemapophyses of the atlas, or clavicles, vary much in extent, in degree of ossification, and even in their presence; the pleurapophyses of the cervical vertebrae are very short, and are generally united to the other elements, circumscribing the foramen for the vertebral artery. The number of the cervicals is seven, except the alleged six in the manatee and the eight or nine in the three-toed sloth ; this number depends on the existence of the diaphragm, &c., determining the num- ber and distribution of the pairs of cervical nerves; in some whales and armadillos they are more or less consolidated. In the dorsals the pleurapophyses or ribs are movable, and the anterior ones are articulated between two vertebrae; the hsemapophyses are the costal cartilages, and the haemal spines are generally a distinct chain of bones, in the highest con- solidated into a sternum ; toward the loins the pleurapophyses become shorter, and are attached to their respective vertebral centres and to the diapophyses; the haemapophyses become shorter, and finally free and floating. The caudals vary much in number, size, and form ; the short pleurapophyses in the anterior ones are developed at the end of diapophyses, and the haemal arch, when it exists, is articu- lated directly to the bodies. The limbs (ex- cept in cetaceans, where the posterior are wanting) are much alike, whether adapted for flying, digging, swimming, running, or climb- ing, as will be seen under the homotypes be- low. In the fish and reptile the vertebral column is straight or nearly so; in the bird the skull forms a right angle with the neck, the latter having a sigrnoid curve, and the tail bent upward ; in the springing mammals, like the carnivora, there is a convergence of the spinous processes toward the eleventh dorsal, and in most there is a similar convergence toward the fourth cervical, these two regions being the centres of special movements of the column ; in bulky animals, like the elephant, which move with a rigid spine, these processes are all inclined a little backward, as in croco- dilians. In man the spine has several slight and graceful curves, destined to prevent shocks to the nervous system from movements inciden- tal to the erect position ; the curvature of the sacrum and of the coccyx are greatest compared with the number of vertebrae, and the anterior and posterior diverging appendages reach the maximum of development, especially the latter as compared to the length of the spine ; the thumb, which is the least constant part in the rest of the class, becomes in him the most im- portant, constituting a hand proper; in like manner the great toe, the first obliterated in other mammals, is characteristic of the genus Tiomo, as on it depend principally the erect pos- ture and biped gait of man; even the high- est monkey has a posterior thumb instead of a great toe. In the class of serial homologies, or homotypes, may be mentioned the homolo- gy of the scapula with the ilium, the hume- rus with the femur, the ulna with the fibula, radius with tibia, carpus with tarsus, metacar- pus with metatarsus, fingers with toes ; in the skull, the basi - occipital, basi - sphenoid, pre- sphenoid, and vomer are the homotypes of the vertebral bodies ; the coracoid, superior max- illary, clavicle, pubis, ischium, chevron bones, sternal or abdominal ribs and cartilages, and tendinous intersections of the rectus abdominis, are all homotypes and haemapophyses. This system of homotypes is far more natural, sat- isfactory, and intelligible than that of Oken, Spix, and Carus, who speak of the scapula, ilium, femur, humerus, &c., of the head, re- garding each part as a repetition of the whole. This is a fair representation of the principal points of philosophical anatomy, as given in the writings of Owen; in many points it is very unsatisfactory, and he labors jery hard oftentimes to make out his homologies and to refer them to his archetype. With such sources of error and room for variation, it would be useless to expect perfect agreement among authors ; from the nature of adaptive organi- zation, it must be difficult if not impossible to reduce skeletons and their parts to unex- ceptionable laws ; organic systems will not be bound down to any such narrow and clearly defined rules, and are constantly presenting to naturalists instances of inexplicable departure from what have long been considered natural laws. Philosophical anatomy will probably al- ways be an uncertain and ever changing study, assisting but not constituting the science of anatomy. Admitting the f aur cranial vertebras of Oken and Owen, there may still be recog* nized with equal propriety other vertebral cen- tres in advance of the vomer, analogous to the coccyx at the other end of the column, with no nerves belonging to them. The reason for making four cranial vertebrae seems to have been, both with Oken and with Owen, the ex-
