The Vertebrate Skeleton

Chapter XXIII General Account of the Skeleton in Mammalia (Continued)

The Skull and Appendicular Skeleton

The Skull.

Monotremata. In both genera the cranium is thin-walled, has a fairly large cavity, and is very smooth and rounded externally. The sutures between many of the bones early become obliterated in a manner comparable to that in birds, and the facial portion of the skull is much prolonged.

In Echidna the face is drawn out into a gradually tapering rostrum, formed mainly by the premaxillae, maxillae and nasals. The zygomatic arch is very weak, and the palate extends very far back. The tympanic forms a slender ring. The mandible is extremely slight, with no ascending portion, and but slight traces of the coronoid process and angle. The hyoid has a wide basi-hyal and stout thyro-hyals, while the anterior cornua are slender, and include ossified epi-hyals and cerato-hyals.

In Ornithorhynchus the zygomatic arch is much stouter than in Echidna. The face is produced into a wide beak, mainly supported by the premaxillae, between whose diverging anterior ends there is a dumb-bell-shaped bone. The maxillae are flattened below, and each bears a large horny tooth, which meets a corresponding structure borne on a surface near the middle of the mandible. The mandible is considerably stouter than in Echidna, but the angle and coronoid process are but little developed. The infra-orbital foramen and the inferior dental and mental foramina of the mandible are all very large.

Marsupialia. The skulls of the various types of the Marsupials frequently bear a strong superficial resemblance to those of some of the different groups of placental mammals. Thus the skull of the Dasyuridae resembles that of the Carnivora, the resemblance being most marked between the skulls of Thylacinus and the dog. The skull of Notoryctes is strongly suggestive of that of an Insectivore, and that of other Marsupials such as the wombat, recalls equally the characteristic features of a Rodent’s skull. But, however much they may differ from one another, the skulls of all Marsupials agree in the following respects. (1) The brain cavity, and especially the cerebral fossa, has a very small comparative size. (2) The nasals are always large, and the mesethmoid is extensively ossified, and terminated by a prominent vertical edge. (3) Processes from the jugal and frontal in living forms never meet and enclose the orbit, but the zygomatic arch is always complete. (4) The jugal always extends back to form part of the glenoid fossa. (5) The lachrymal canal opens either external to or upon the margin of the orbit, and the nasal processes of the premaxillae never quite reach the frontals. (6) The posterior part of the palate is commonly pierced by large oval vacuities. (7) The tympanic is small and never fused to the bones of the cranium. (8) The carotid canal perforates the basisphenoid and not the tympanic bulla. (9) The optic foramen and sphenoidal fissure are confluent. (10) In every case except Tarsipes the angle of the mandible is more or less inflected.

The skull of the extinct Thylacoleo differs from that of all other Marsupials in the fact that the postorbital bar is complete. The hyoid is constructed on much the same plan in all Marsupials. It consists of a small basi-hyal, a pair of broad cerato-hyals, and a pair of strong thyro-hyals. The epi-hyals and stylo-hyals are generally unossified.

Edentata. In Sloths (Bradypodidae) the sutures become early obliterated, the cranial portion of the skull is rather high, and the facial portion very short. The lachrymal is very small, and its canal opens outside the orbit. The zygomatic arch is incomplete, and the jugal (fig. 91, 5) is curiously forked, but in a manner differing in the two genera. The premaxillae are very small,—in Bradypus quite vestigial. The mandible is well developed, the angle being specially marked in Bradypus. In Choloepus the symphysial part is drawn out in a somewhat spout-like manner (fig. 91, 6). In both genera the thyro-hyals are ankylosed with the basi-hyal.

In Megatherium the general appearance of the skull is distinctly sloth-like, but the facial portion is more elongated, partly owing to the development of a prenasal bone, and the zygomatic arch is complete. The mandible is very deep in the middle, and is drawn out into a long spout-like process in front.

Anteaters (Myrmecophagidae) have a much modified skull, and this is especially the case in the Great Anteater, Myrmecophaga. The skull is smooth and evenly-rounded, in these respects recalling that of Echidna, but it is longer and tapers much more gradually than in Echidna. The occipital condyles are remarkably large. The premaxillae are small, and the long rostrum is chiefly composed of the maxillae and nasals with the mesethmoid and vomer. The zygomatic arch is incomplete, and there is no trace of a separation between the orbit and the temporal fossa. The palate is much elongated, the pterygoids meeting in the middle line just like the palatines. The mandible is very long and slender, there being no definite coronoid process, and a short and slight symphysis. The hyoid arch is noticeable for the length of the anterior cornu.

In the Armadillos (Dasypodidae) the skull varies a good deal in shape, but the facial portion is always tapering and depressed. The zygomatic arch is complete. In Dasypus and Chlamydophorus the tympanic bulla is well ossified.

In the Glyptodontidae the skull is very short and deep; the zygomatic arch is complete, and has a long downwardly projecting maxillary process. The mandible is massive, and has a very high ascending portion.

In the Manidae the skull is smooth and rounded, the zygomatic arch is incomplete, and the orbit is inconspicuous. The palate is long and narrow, but the pterygoids do not take part in its formation. The mandible is slightly developed and has no angle or coronoid process.

In Orycteropus the zygomatic arch is complete, and there is a small postorbital process to the frontal. The mandible is well-developed, having a coronoid process and definite ascending portion, and the hyoid is well ossified.

Sirenia. The skull, and especially the brain case of all Sirenia, is remarkable for the general density of the component bones, which, though often very thick, are without air sinuses. It is noticeable also for the roughness of the bones, and the irregular manner in which they are united together.

The cranial cavity is decidedly small, the reduction being specially noticeable in the cerebral fossa, which is not much larger than the cerebellar fossa. The foramen magnum is large, and the dorsal surface of the cranium narrow. The zygomatic arch is very strongly developed, the squamosal (fig. 92, 4) being especially prominent, and being drawn out not only into the zygomatic process, but also into a large post-tympanic process which articulates with the exoccipital. At the side of the skull between the squamosal, supra-occipital and exoccipital, there is a wide vacuity in the cranial wall, partially filled up by the very large periotic, which is ankylosed to the tympanic, but is not united to any other bones of the skull. The foramen lacerum medium is confluent with the foramen lacerum anterius, and the two together form an enormous vacuity on the floor of the skull, bounded chiefly by the exoccipital, basi-occipital, alisphenoid and squamosal. The jugal (fig. 92, 8) is large and in Manatus sends up a strong process, which nearly or quite meets the postorbital process of the frontal, completing the orbit. In the other Sirenia the orbit is completely confluent with the very large temporal fossa. The lachrymal in Manatus is very small, but is larger in Halicore. The premaxillae (fig. 92, 9) are large, but smaller in Manatus than in the other genera, in all of which they are curiously bent down in front. Their upper margin forms the anterior border of a very large aperture lying high on the roof of the skull and extending back for a considerable distance. This aperture is formed by the union of the two anterior nares. The nasals are quite vestigial or absent, and the narial aperture is bounded above by the frontals; in its floor are seen the slender vomer and large mesethmoid. The palate is long and narrow, and formed mainly by the maxillae; behind it there is a large irregular process formed by the union of the palatine, pterygoid, and pterygoid plate of the alisphenoid. The mandible is very massive and has a very high ascending portion, a rounded angle (fig. 92, 10), and a prominent coronoid process; the two rami are firmly ankylosed together. The hyoid consists principally of the broad flat basi-hyal; the anterior cornua are but slightly ossified, while the thyro-hyals are not ossified at all.

Cetacea. The skull in all Cetacea, especially in the Odontoceti, is a good deal modified from the ordinary mammalian type.

In the Archaeoceti this modification is less marked than in either of the other suborders. The nasals and premaxillae are a good deal larger than they are in living forms, and the anterior nares are placed further forward. The maxillae do not extend back over the frontals, and there is a well-marked sagittal crest.

In the Mystacoceti the skull is always quite bilaterally symmetrical, and is not so much modified from the ordinary mammalian type as in the Odontoceti. The parietals are not, as in the Odontoceti, separated by a wide interparietal, but meet; they are, however, hidden under the very large supra-occipital. The nasals are developed to a certain extent, and the nares, though placed very far back and near the top of the head, terminate forwardly-directed narial passages. Turbinal bones are also developed to some extent; this fact, and the occurrence of a definite though small olfactory fossa constituting important distinctions from the Odontoceti. The maxillae are large, but do not extend back to cover the frontals as in the Odontoceti. The zygomatic process of the squamosal is very large. The mandibular rami are not compressed, but are rounded and arched outwards, and never meet in a long symphysis.

Odontoceti. The skull departs widely from the ordinary mammalian type. The following description will apply to any of the following genera of the Delphinidae, Phocaena, Globicephalus, Lagenorhynchus, Delphinus, Tursiops, Prodelphinus, Sotalia.

The upper surface of the skull is more or less asymmetrical. The cerebral cavity is high, short and broad; and formed mainly by the cerebral fossa, the olfactory fossa being entirely absent. The supra-occipital (fig. 93, 3) is very large, and forms much of the posterior part of the roof of the skull. It has the interparietal (fig. 93, 7) fused with it, and completely separates the two parietals. The frontal (fig. 93, 10) is large and laterally expanded, forming the roof of the orbit, but is almost completely covered by an extension of the maxillae. The zygomatic arch is very slender, and is mainly formed by a rod-like process from the jugal (fig. 93, 15), the zygomatic process of the squamosal being short and stout.

The nasal passages are peculiarly modified, instead of passing horizontally forwards above the roof of the mouth, they pass upwards and even somewhat backwards towards the top of the skull (fig. 93, 23). They are bounded laterally by two processes from the premaxillae, the left of which is shorter than the right. The nasal cavities are narrow and without turbinals and the nasals (fig. 93, 19) are almost as much reduced as in Sirenia.

In front of the nasal openings the face is prolonged as a narrow beak or rostrum of varying length, formed by the maxillae and premaxillae surrounding the vomer and large mesethmoid (fig. 93, 11), which sends forwards a long partially cartilaginous process, and is fused behind with the presphenoid (fig. 93, 8). The basi-occipital (fig. 93, 1) too is fused with the basisphenoid. The foramen rotundum is confluent with the sphenoidal fissure, and the foramen ovale with the foramen lacerum medium and the foramen lacerum posterius. The palate is mainly formed by the maxillae; the premaxillae and palatines (fig. 93, 17), though both meet in symphyses, forming very little of it. The pterygoids vary in size in the different genera, sometimes as in Lagenorhynchus and Delphinus meeting in the middle line, sometimes as in Phocaena and Globicephalus (fig. 93, 18) being widely separated. The tympanic and periotic are not fused together, and the periotic has generally no bony union with the rest of the skull. The mandible is rather slightly developed, with the rami straight, compressed and tapering to the anterior end. The condyle is not raised at all above the edge of the ramus; the angle is rounded and the coronoid process is very small. Platanista has a curiously modified skull; the rostrum and mandible are exceedingly long and narrow, and arising from the maxillae are two great plates of bone which nearly meet above.

In the Physeteridae the skull is raised into a very prominent crest at the vertex behind the nares. In front of this in Hyperoödon a pair of ridges occur, formed by outgrowths from the maxillae. In the old male these ridges reach an enormous size and almost meet in the middle line. In Physeter, the Sperm whale, these ridges are not developed; the maxillae and premaxillae unite with the other bones of the crest enclosing an enormous half basin-shaped cavity, at the base of which are the very asymmetrical anterior narial apertures.

In all living Cetacea the hyoid has the same general shape, consisting firstly of a crescentic bone formed by the fusion of the thyro-hyals with the basi-hyal, and secondly of the anterior cornu formed principally by the strong stylo-hyal.

Ungulata. None of the distinctive characters separating the Ungulata from the other groups of mammals are drawn from the skull. But in the Ungulata vera as opposed to the Subungulata a distinguishing feature is found in the fact that the lachrymal and jugal form a considerable part of the side of the face, and that the jugal always forms the anterior part of the zygomatic arch, the maxillae taking no part in it.

Ungulata vera.

Artiodactyla. The skull in Artiodactyla differs from that in Perissodactyla in the fact that the posterior end of the nasal is not expanded and there is no alisphenoid canal.

The skulls in the different groups of Artiodactyla differ considerably from one another.

The skull of the Pig166 will be described as illustrative of the skull in the Suina. In the Pig as in most Artiodactyla the face is bent sharply down on the basicranial axis, the commencement of the vomer being situated below the mesethmoid instead of in front of it as in most skulls. The occipital region of the skull is small, and the line of junction of the supra-occipital and parietals is raised into a prominent occipital crest. The parietal completely fuses at an early stage with its fellow, and the exoccipital is drawn out into a long paroccipital process (fig. 94, A, 8). The frontal is large and broad and drawn out into a small postorbital process. The lachrymal too is large and takes a considerable part in forming the side of the face in front of the orbit, as does also the jugal, though to a less extent. The face is long and tapers much anteriorly. The nasals are long and narrow, as are the nasal processes of the premaxillae, which do not however reach the frontals. A prenasal ossicle is developed in front of the mesethmoid. The palate is long and narrow, the pterygoid (fig. 94, A, 10) is small, but the pterygoid process of the alisphenoid is prominent. The squamosal is small and has the tympanic fused with it; the tympanic is dilated below, forming a bulla (fig. 94, A, 9) filled with cancellous bone, and above forms the floor of a long upwardly-directed auditory meatus. The mandible has a high ascending portion and a small coronoid process (fig. 94, B, 13). The hyoid differs from that of most Ungulates, the stylo-hyal being very imperfectly ossified.

In Hippopotamus the skull though essentially like that of the pig is much modified in detail. The brain cavity is very small, while the jaws are immensely developed. The face contracts in front of the orbits and then expands again greatly, to lodge the enormous incisor and canine teeth. The postorbital bar is complete or nearly so, and the orbits project curiously outwards and slightly upwards; the lachrymal is thin and much dilated. The squamosal is drawn out into a postglenoid process, and the hamular process of the pterygoid is prominent. The tympanic bulla is filled with cancellous bone. The mandible is enormously large, the symphysis is long, the angle much expanded and drawn out into a process which projects outwards and forwards.

Among extinct forms related to the Suina, Cyclopidius is noticeable for having large vacuities in the lachrymo-nasal region, while Cotylops has the postorbital bar complete; both these forms are from the North American Miocene.

In the Tylopoda and Tragulina the skull resembles in most respects that of the Ruminants, shortly to be described; but it is allied to that of the Suina in having the tympanic bulla filled with cancellous bone. The tympanic bulla is better developed in the Tragulina than in most Ungulates.

Among Ruminants, the Bovidae, that large group including the Oxen, Sheep, and Antelopes, as a rule have the face bent on the basicranial axis much as in the Suina. The parietals are generally small and early coalesce, the frontals are large and are usually drawn out into horn cores, which are however absent in the skulls of some domestic varieties of sheep and oxen, and also in some of the earlier extinct forms of Bovidae. These horn cores are formed internally of cancellous bone, and on them the true epidermal horns are borne. In young animals there is a distinct interparietal, but this early fuses with the supra-occipital, and in the oxen also with the parietals. The occipital crest is generally well marked, but in the genus Bos becomes merged in a very prominent straight ridge running between the two horn cores; this ridge, which contains air cells communicating with those in the horn cores, is not nearly so well marked in Bison. There is often, as in Gazella, a vacuity on the side of the face between the nasal, frontal, lachrymal, and maxillae, but this is not found in oxen or sheep. The premaxillae are small, the nasals are long and pointed, and the turbinals are much developed. The Saiga antelope has a curiously specialised skull; the nasals are absent or have coalesced with the frontals and the anterior nares are enormously large. In all Ruminants the lachrymal is large and forms a considerable part of the side of the face; it often bears a considerable depression, the suborbital or lachrymal fossa, well seen in most of the smaller antelopes. The postorbital bar is complete, and the orbit is prominent and nearly circular. The palatines and pterygoids are moderately large, and the pterygoids have a backwardly-projecting hamular process. The squamosal is small, but has a postglenoid process. The tympanic is not fused to the periotic and has a small bulla not filled with cancellous bone. There is a large paroccipital process to the exoccipital and the mandible has a long slender coronoid process.

In the Cervidae and Giraffidae the face is not bent down on the basicranial axis as it is in the Bovidae. The frontals are drawn out, not into permanent horn cores as in the Bovidae, but into short outgrowths, the pedicels, upon which in the Cervidae long antlers are annually developed. These antlers are outgrowths of bone, and are covered during development by vascular integument, which dries up and peels off when growth is complete. Every year they are detached, by a process of absorption at the base, and shed. They may occur in both sexes, as in the Reindeer, but as a rule they are found only in the male. They are generally more or less branched, and are sometimes of enormous size and weight, as in the extinct Cervus megaceros. In young animals they are always simple, but become annually more and more complicated as the animal grows older.

In the Giraffe the frontals bear a small pair of bony cores, which are at first distinct, but subsequently become fused to the skull. In the allied Sivatherium, a very large form from the Indian Pliocene, the skull bears two pairs of bony outgrowths, a pair of short conical outgrowths above the orbits, and a pair of large expanded outgrowths on the occiput.

The opening of the lachrymal canal is commonly double and the lachrymal fossa is large in the Cervidae and the Giraffidae except Sivatherium. The vacuity between the frontal, lachrymal, maxillae, and nasal is specially large.

The hyoid of Ruminants is noticeable for the development of the anterior cornua, which include stout and short cerato-hyals and epi-hyals, long and strong stylo-hyals and large tympano-hyals which are more or less imbedded in the tympanics.

Perissodactyla. In the skull of Perissodactyles an alisphenoid canal is found and the nasals are expanded behind. Among the living animals belonging to this group the skull least modified from the ordinary type is that in Rhinoceros. In this form the skull is considerably elongated, the facial portion being very large. The occipital region is elevated, but the cranial cavity is small, the boundary line between the occipital and parietal regions being drawn out into a prominent crest, which is occupied by air cells. There is no postorbital process to the frontal, and the orbit is completely confluent with the temporal fossa. The nasals are fused together and are very strongly developed, extending far forwards, sometimes considerably beyond the premaxillae. In some extinct species, such as Elasmotherium and the Tichorhine Rhinoceros, R. antiquitatis, the mesethmoid is ossified as far forwards as the end of the nasals. The nasals are arched and bear one or two roughened surfaces to which the great nasal horns are attached. The premaxillae are very small and the pterygoids are slender. The palate is long, narrow, and deeply excavated behind. The postglenoid process of the squamosal is well developed, and generally longer than the paroccipital process of the exoccipital. The tympanic and periotic are both small and are fused together. The condyle of the mandible is very wide, the angle rounded, and the coronoid process moderately developed.

In the Titanotheriidae, a family of extinct Perissodactyla from the Miocene of North America, the occipital region is much elevated, as is also the fronto-nasal region, the nasals (perhaps only in the male) bearing a pair of blunt bony outgrowths. Between these two elevated regions the skull is much depressed. The cranial cavity is very small, the orbit confluent with the temporal fossa, and the zygomatic arch massive.

In Tapirus the orbit and temporal fossa are confluent. The nasals are small, wide behind and pointed in front, and are supported by the mesethmoid; the anterior nares are exceedingly large and their lateral boundaries are entirely formed by the maxillae. The postglenoid and post-tympanic processes of the squamosal are large. The periotic is not fused to the squamosal or to the small tympanic. The mandible is large and has the angle much developed and somewhat inflected.

Palaeotherium, which lived in early Tertiary times, has a skull much like that of the Tapir, especially as regards the nasal bones.

In the Horse and its allies (Equidae) the facial portion of the skull is very large as compared with the cranial portion, the nasals and nasal cavities being specially large. In the living species of the genus Equus there is no fossa between the maxillae and lachrymal, but it occurs in some extinct species. The lachrymal and jugal form a considerable part of the side of the face; and the orbit though small is complete and prominent. The postorbital bar is formed by a strong outgrowth from the frontal, which unites with a forward extension of the squamosal. The squamosal may extend forwards and form part of the wall of the orbit, a very unusual feature, as in most mammals the squamosal stops before the postorbital bar. The palate is narrow and excavated behind as in Rhinoceros; the palatines take very little part in its formation. The glenoid surface for the articulation of the mandible is very wide. The squamosal gives rise to small postglenoid and post-tympanic processes, and the exoccipital to a large paroccipital process. The tympanic and periotic are ankylosed together, but not to any other bones.

In the Subungulata, the lachrymal and jugal do not form any considerable part of the side of the face, and the maxillae commonly takes part in the formation of the zygomatic arch.

Toxodontia. The skull in the Toxodontia shows several Artiodactyloid features, while the manus and pes are of a more Perissodactyloid type. The Artiodactyloid features are (1) the absence of an alisphenoid canal, (2) the fact that the palate is not excavated behind, and that the palatines form a considerable part of it, and (3) the fusion of the tympanic to the squamosal and exoccipital, forming the floor of an upwardly directed auditory meatus. The frontal has a fairly well developed postorbital process, but the orbit is confluent with the temporal fossa. The premaxillae is well developed, as is the paroccipital process of the exoccipital, especially in Typotherium. The mandible has a rounded angle and a coronoid process of moderate size. In Typotherium the ascending portion is very massive.

Condylarthra. As far as is known the skull of these generalised Ungulates is depressed, and is frequently marked by a strong sagittal crest. The cranial cavity is small, the cerebral fossa in Phenacodus being exceptionally small. The orbit is completely confluent with the temporal fossa.

Hyracoidea. The skull of Procavia resembles that of Perissodactyles more than that of any other Ungulates, but differs strongly in the comparatively small size of its facial portion. The posterior portion of the cranium is rather high, the occipital plane being nearly vertical. There is a small interparietal. The nasals are wide behind, and the zygomatic arch is strongly developed, its most anterior part being formed by the maxillae. The jugal and parietal give rise to postorbital processes which sometimes meet, but as a rule the orbit is confluent with the temporal fossa; it is very uncommon for the parietal to give rise to a postorbital process, and even in Procavia the frontal often forms part of the process. The alisphenoid canal, and postglenoid and paroccipital processes are well developed. The tympanic bulla is large and the periotic and tympanic are fused together, but not as a rule to the squamosal. The ascending portion of the mandible is very high and broad, the angle rounded and the coronoid process moderate in size. The hyoid is singular, there is a large flat basi-hyal prolonged laterally into two broad flattened thyro-hyals. Articulating with its anterior end are two large triangular cerato-hyals, which are drawn out into two processes meeting in the middle line.

Amblypoda. In the Uintatheriidae (Dinocerata) the skull has a very remarkable character, being long and narrow and drawn out into three pairs of rounded protuberances, a small pair on the nasals, a larger pair on the maxillae in front of the orbits, and the largest pair on the parietals. The cranial cavity, and especially the cerebral fossa, is extraordinarily small. The orbit is not divided behind from the temporal fossa. The mandible has a prominent angle, and a long curved coronoid process; its symphysial portion bears a curious flattened outgrowth to protect the great upper canines.

In Coryphodon the skull is of a more normal character, being without the conspicuous protuberances. The cranial cavity though very small is not so small as in Uintatherium.

Proboscidea. The character of the skull in the young elephant differs much from that in the old animal. In very young individuals the skull is of a normal character, and the cranial cavity is distinctly large in proportion to the bulk of the skull. But as the animal gets older, while its brain does not grow much, the size of its trunk and especially of its tusks increases greatly; and consequently the skull wall is required to be of very great superficial extent in order to afford space for the attachment of the muscles necessary for the support of these heavy weights. This increase in superficial extent is brought about without much increase in weight of bone by the development of an enormous number of air cells in nearly all the bones of the skull; sometimes, as in the case of the frontal, separating the inner wall of the bone from the outer, by as much as a foot. This development of air cells is accompanied by the obliteration of the sutures between the various bones. The most noticeable point with regard to the cranial cavity is the comparatively large size of the olfactory fossa. The supra-occipital (figs. 96 and 97, 9) is large—exceedingly large in the adult skull; the parietals (figs. 96 and 97, 2) are also very large. The frontals send out small postorbital processes, but these do not meet processes from the small jugal, which forms only the middle part of the slender zygomatic arch, the anterior part being formed by the maxillae. The lachrymal (fig. 96, 15) is small and lies almost entirely inside the orbit. The anterior narial aperture (fig. 97, 8) is wide and directed upwards, opening high on the anterior surface of the skull. It is bounded above by the short thick nasals and below by the premaxillae. The narial passage is freely open, maxillo-turbinals not being developed. The palatine is well developed, the pterygoid is small and early fuses with the pterygoid process of the alisphenoid. The tympanic is united with the periotic but not with the squamosal, and forms a large auditory bulla. There are no paroccipital or postglenoid processes. The exoccipital is not perforated by the condylar foramen,—a very exceptional condition.

The mandible has a high ascending portion, is rounded off below and has no angle. The symphysial portion is long, narrow, and spout-like, and the coronoid process is small. The thyro-hyals are ankylosed with the basi-hyal, which is connected with the large forked stylo-hyals by ligament only.

Rodentia. The cranial cavity is depressed, elongated, and rather small, and the cerebral fossa lies entirely in front of the cerebellar fossa. The occipital plane is vertical or directed somewhat backwards, and the supra-occipital does not form much of the roof of the cranium. The paroccipital processes of the exoccipitals are generally of moderate size; in the Capybara (Hydrochaerus), however, they are very long, and are laterally compressed and directed forwards. The parietals are small, and often become completely fused together; there is sometimes a small interparietal. The frontals in most genera have no trace of a postorbital process; in Squirrels, Marmots and Hares, however, one occurs, but in no case does it meet a corresponding process from the zygomatic arch, so the orbit and temporal fossa are completely confluent. In Hares the postorbital process of the frontal is much flattened, and has an irregular margin. The temporal fossa is always small, and in Lophiomys is arched over by plates arising respectively from the parietal and jugal; a secondary roof is thus partially developed in a manner unique among mammals, but carried to a great extent in many Chelonia. The nasal bones and cavities are large, attaining their maximum development in the Porcupines (fig. 98, 1). The premaxillae is always very large, and sends back a long process which meets the frontal. The vomer is occasionally found persisting in two separate halves, a feature recalling the arrangement in Sauropsids. In many Rodents there is an enormous vacuity at the base of the maxillary portion of the zygomatic arch. It is sometimes as large as the orbit, and attains its maximum development in the Capybara and other Hystricomorpha; in the Marmots, Beavers, and Squirrels (Sciuromorpha), and in the Hares it is undeveloped. In Lagostomus the maxillae bears an upwardly directed plate of bone, shutting off from this vacuity a space which is the true infra-orbital foramen.

The zygomatic arch is always complete, and in many cases the jugal extends back to form part at least of the glenoid surface for articulation with the mandible. In Coelogenys the jugal and maxillary portion of the zygomatic arch is greatly expanded and roughened, and the maxillary portion encloses a large cavity. The palate in Rodents is narrow, and the space between the incisor and molar teeth passes imperceptibly into the sides of the face. The anterior palatine foramina form long, rather narrow slits in this region. The bony palate between the grinding teeth is sometimes as in the Hares very short, sometimes as in the Capybara very long. The maxillae extends back beneath the orbit to unite with the squamosal. The pterygoid is always small, but sometimes has a well-marked hamular process which in Hystrix, Lagostomus, and some other genera unites with the tympanic bulla. The periotic is large, and fused with the tympanic, which forms a prominent bulla, and is generally drawn out into a tubular meatus. The bulla attains its maximum development in Chinchilla and Dipus.

The mandible is narrow and rounded in front, the two halves meeting in a long symphysis. The angle is generally drawn out into a long backwardly-projecting process, which is often pointed and directed upwards. In the Hares the angle is rounded. The coronoid process is never large.

There are a number of points in which the skull of the Duplicidentata (Hares and Rabbits) differs from that of other Rodents. (a) The sutures between the basi-occipital and basisphenoid, and between the basisphenoid and presphenoid remain open throughout life. (b) Much of the maxillae forming the side of the face in front of the orbit is fenestrated. (c) The optic foramina are united to form a single hole, much as in birds. (d) The coronoid process is slightly differentiated from the ascending portion of the mandible. The first two of these points have been thought to indicate degradation of the hares and rabbits as compared with higher mammals.

Carnivora167. It is characteristic of the skull in Carnivora that the glenoid fossa is deep, and the postglenoid process (fig. 75, 23) well developed. The condyle of the mandible is much elongated transversely. The orbit and temporal fossa in the great majority of forms communicate freely, the postorbital bar being incomplete.

Carnivora vera. The axis of the facial portion of the skull is a direct continuation of that of the cranial portion. The cranial cavity though rather depressed is large, and generally long, though in Cats it is comparatively short and wide. The occipital plane is nearly vertical, and the exoccipitals are developed into fairly prominent paroccipital processes. The interparietal is commonly distinct, and the parietals unite in a long sagittal suture, which is often developed into a crest. The nasals (fig. 73, 4) are well developed, especially in Cats, and the nasal processes of the premaxillae do not nearly reach the frontals. A considerable part of the palate is formed by the palatine, and the maxillary portion is pierced by rather long anterior palatine foramina. The pterygoid has a hamular process. The zygomatic arch is strong, especially in Cats. Postorbital processes are developed on the frontal (fig. 73, 10) and jugal, but never form a complete postorbital bar. A carotid canal is well seen in the Ursidae, and to a less extent in the Felidae; in the Canidae there is an alisphenoid canal (fig. 75, 21).

The auditory bulla differs a good deal in the different groups. In the Bears (Ursidae) it is not much inflated, and is most prominent along its inner border; it is not closely connected with the paroccipital process. In the Cats it is very prominent, and its cavity is almost divided by a septum into two parts, the inner of which contains the auditory ossicles. The paroccipital process is closely applied to the bulla. In the Dogs the bulla is intermediate in character between that of the Cats and that of the Bears; it is partially divided by a septum, and is moderately expanded.

The mandible is well developed with a prominent angle (fig. 72, 26), and a large coronoid process. The hyoid consists of a broad basi-hyal, a long many-jointed anterior cornu and short thyro-hyals (fig. 72, 33).

The skull in the Creodonta is in most respects allied to that of the Canidae, but presents some ursine affinities. The tympanic bulla is fairly prominent, but has no well-developed septum. The cranial cavity is very small and narrow, the zygomatic arch standing away from it. The temporal fossa is of great size.

In the Pinnipedia the cranial cavity is large and rounded. The skull is much compressed in the interorbital region, and in correlation with this compression the ethmo-turbinals are little developed, while the maxillo-turbinals are large. The orbit is large, and the temporal fossa smaller than in the Carnivora vera. In the Walrus (Trichechus) the anterior part of the face is distorted by the development of the huge canines. The Otariidae have an alisphenoid canal. The tympanic bulla is small in Otaria, large in the Phocidae, and flattened in the Walrus. The hyoid is similar to that in Carnivora vera.

Insectivora. The skull varies much in the different members of the order Insectivora, but the following points of agreement are found. The cranial cavity is of small size, and is never much elevated. The facial part of the skull is generally considerably elongated, and the nasals and premaxillae are well developed. The zygomatic arch is usually slender or incomplete, and the coronoid process and angle of the mandible are commonly prominent.

In some Insectivora, such as Galeopithecus, Tupaia, and Macroscelides, the skull shows a higher type of structure than is met with in most members of the order. In these genera the cranial cavity is comparatively large, and the occipital plane is nearly vertical. The zygomatic arch is fairly strong, and the frontal and jugal give rise to postorbital processes which nearly or quite (Tupaia) meet. The tympanic bulla is well developed, and produced into a tubular auditory meatus, this being specially well marked in Macroscelides.

In the other Insectivora the cranial cavity is of smaller comparative size, and the orbit and temporal fossa are completely confluent, often without any trace of a postorbital bar. The occipital plane commonly slopes forwards. In the Hedgehogs (Erinaceidae) and Centetidae the tympanic is very slightly developed, forming a small ring. The zygomatic arch of Hedgehogs and Gymnura is very slender, the jugal being but little developed and the squamosal and maxillae meeting one another; in the Centetidae the jugal is absent and the arch is incomplete.

The Moles (Talpidae) have an elongated, depressed and rounded skull with a very slender zygomatic arch formed by the squamosal and maxillae. The nasals are fused together, and the mesethmoid is ossified very far forwards. In the Shrews (Soricidae) there is no zygomatic arch; the tympanic is ring-like, and the angle of the mandible is very prominent. The hyoid has a transversely extended basi-hyal, a long anterior cornu with three ossifications, and thyro-hyals which are sometimes fused to the basi-hyal.

Chiroptera. In the frugivorous Flying Foxes (Pteropidae) the skull is elongated, and the cranial cavity is large and arched, though considerably contracted in front. There are commonly strong sagittal and supra-orbital crests. The parietals take a great part in the formation of the walls of the cranial cavity, the supra-occipital and frontals being small. The frontal is drawn out into a long postorbital process, but the zygomatic arch, which is slender, and formed mainly by the squamosal and maxillae, gives rise to only a small postorbital process, so that the orbit and temporal fossa are confluent. There is no alisphenoid canal, and the tympanics are very slightly connected with the rest of the skull. The mandible has a large coronoid process, a rounded angle, and a transversely expanded condyle.

In Insectivorous Bats the skull is generally shorter and broader than in the Pteropidae. The cranial cavity is large and rounded, and has thin smooth walls. The zygomatic arch is slender, and postorbital processes are not generally well developed. The premaxillae is generally small, sometimes absent. The tympanics are ring-like and are not connected with the surrounding bones. The angle of the mandible is distinct. The hyoid in most respects resembles that of the Insectivora.

Primates. The characters of the skull differ greatly in the two suborders of Primates, the Anthropoidea and the Lemuroidea.

In the Lemuroidea the general relative proportions of the cranium and face are much as in most lower mammals, and the occipital plane forms nearly a right angle with the basicranial axis. The postorbital processes of the frontals are commonly continued as a pair of ridges crossing the roof of the cranium and meeting the occipital crest. Though the postorbital bar is complete, the orbit and temporal fossa communicate freely below it. The lachrymal canal opens outside the orbit, and the lachrymal forms a considerable part of the side of the face. The tympanic is developed into a large bulla. The hyoid apparatus much resembles that of the Dog.

In the Anthropoidea the skull differs greatly from that in the Lemuroidea. The cranial portion of the skull is very large as compared with the facial portion, though the comparative development varies, some monkeys, such as the baboons (Cynocephali) having the facial portion relatively large. The comparative size of the jaws does not vary inversely with the general development of the animal, some of the Cercopithecidae having comparatively larger jaws than some of the Cebidae. The great size of the cranial part of the skull is mainly due to the immense development of the cerebral fossa, which commonly completely overlaps the olfactory fossa in front, and the cerebellar fossa behind. This development also has the effect of making the ethmoidal and occipital planes lie, not at right angles to the basicranial axis, but almost in the same straight line with it. This is, however, not always the case, as the Howling Monkey (Mycetes) and also some of the very highest monkeys, the Gibbons (Hylobates), have the occipital plane nearly vertical to the basicranial axis. In adult Man the basi-occipital, exoccipitals and supra-occipital coalesce, forming the so-called occipital bone; while the basisphenoid, presphenoid, alisphenoids, orbitosphenoids and pterygoids form the sphenoid bone. The roof of the skull is partly formed by the large supra-occipital and frontals, but mainly by the parietals (fig. 99, 1), which in Man are of enormous extent.

In Man and in most monkeys, at any rate when young (fig. 99, B), the roof of the skull is smooth and rounded, but in many forms, such as the Baboons, in the adult the supra-orbital and occipital ridges are much developed. In the Gorilla this is also the case with the sagittal crest (fig. 99, A, 2). The bones of the upper surface of the cranium interlock with wavy outlines. The nasals vary much in length, being much shorter in man than in most monkeys; they commonly become early fused together, as do also the frontals. The vomer is well developed, and the ethmo-turbinal always forms part of the boundary of the orbit. There are frequently, as in many Lemuroidea, a pair of more or less well-marked ridges, crossing the roof of the skull from the postorbital processes of the frontals to the occipital crest. The orbit is completely encircled by bone, and the alisphenoid assists the jugal and frontal in shutting it off from the temporal fossa, leaving however a communication between the two as the sphenomaxillary fissure. In most cases the frontals meet one another in the middle line between the mesethmoid and orbitosphenoid, but in Man, Simia, and some Cebidae this does not take place. In nearly all Cebidae the parietal and jugal meet one another, separating the frontal and alisphenoid on the skull wall; in Man and all Old World monkeys, on the other hand, the alisphenoid and frontal meet and separate the jugal and parietal. The premaxillae nearly always send back processes which meet the nasals. The palate is rather short and both the palatine and the premaxillae take a considerable part in its formation. The pterygoid plate of the alisphenoid is decidedly large, and there is no alisphenoid canal. There is never any great development either of the paroccipital process of the exoccipital, or of the postglenoid process of the squamosal. The periotic and tympanic are always fused together; in Cebidae they form a small bulla, but a bulla is not developed in any Old World forms. The periotic is large, especially the mastoid portion, which forms a distinct portion of the skull wall between the squamosal and exoccipital. In Man and still more in Old World monkeys, the external auditory meatus is drawn out into a definite tube, whose lower wall is formed by the tympanic; in the Cebidae the tympanic is ring-like. The perforation of the periotic by the carotid canal is always conspicuous.

The mandible is rather short and broad, and the angle formed by the meeting of the two rami is more obtuse than in most mammals. The coronoid process is fairly well developed, and the angle is more or less rounded. In most Primates the condyle is considerably widened, but this is not the case in Man. In Mycetes the mandible is very large, its ascending portions being specially developed. The hyoid of Primates is remarkable for the large expanded basi-hyal, which is generally concave above and convex below. The anterior cornu is never well ossified, but the thyro-hyal is always strong. In Mycetes the basi-hyal is enormously large, forming a somewhat globular thin-walled capsule.

Auditory ossicles.

There are in mammals four auditory ossicles forming a chain extending from the fenestra ovalis to the tympanic membrane. Three of these, the malleus, incus and stapes, are always distinct, while the fourth, the lenticular, is smaller than the others and is sometimes not distinct. The names are derived from human anatomy and indicate in the case of the first three a more or less fanciful resemblance respectively to a hammer, an anvil and a stirrup. The ossicles are homologous as a whole to the hyomandibular of fishes and to the columellar chain of Sauropsids and Amphibians. The malleus is homologous to the extra-columella of Crocodiles and the stapes to the columella. The malleus when typically developed consists of a rounded head (fig. 100, 1) which bears a surface articulating with the incus, and a short neck continued into a process, the manubrium (fig. 100, 5), which comes into relation with the tympanic membrane. From the junction of the neck and manubrium two processes are given off, a processus longus or gracilis (fig. 100, 4), which in the embryo is continuous with Meckel’s cartilage, and a processus brevis (fig. 100, 6). The incus generally consists of a more or less anvil-shaped portion which articulates with the malleus, and of a process which is connected with the stapes by the small lenticular. The stapes is generally stirrup shaped, consisting of a basal portion from which arise two crura separated by a space the canal through which a branch of the pharyngeal artery runs The lenticular is frequently cartilaginous and sometimes is not developed at all.

The above is the arrangement of the auditory ossicles met with in the higher Mammalia, but in the lower Mammalia the characters approach more nearly to those met with in Sauropsids.

In Monotremes the ossicles, though distinctly mammalian in character, show a very low type of development. The incus is articulated, or often fused, with an outgrowth from the head of the malleus. The stapes is very much like a reptilian columella, having a single crus with no perforation.

In Marsupials the ossicles are of a low type, but not so low as the rest of the skeleton might have led one to expect, and all or almost all the points showing a low grade of development may be paralleled among the Monodelphia. The lowest Marsupials as regards the ossicles are the Peramelidae, whose ossicles are of a frail papery consistence. The Didelphyidae on the other hand have the most highly developed ossicles, the malleus much resembling that of many Insectivores, and the stapes having two definite crura separated by a canal.

In Edentates the character of the ossicles varies much. In Sloths the stapes approaches that of Sauropsids in its narrowness and the slight trace of a canal; this character is however still more marked in Manis, whose stapes is as Sauropsidan as that of Monotremes, and consists of a nearly circular basal plate bearing a column which does not show any sign of division into crura. The stapes of other Edentates, such as ant-eaters, aard varks, and most armadillos, is of a high type and has well-developed crura. Priodon has a lower type of stapes than Dasypus and Tatusia.

The ossicles of the Sirenia differ widely from those of all other mammals in their great density and clumsy form.

In Cetacea the ossicles are solid, though not so solid as in Sirenia, and their details vary much. The malleus is always firmly fused to the tympanic by means of the processus longus, and the manubrium is very little if at all developed. The incus has the stapedial end greatly developed, and the stapes has very thick crura with hardly any canal. The ossicles of the Mystacoceti are apparently less specialised than are those of the Odontoceti.

The auditory ossicles of the Ungulata do not present any characters common to all the members of the group.

Among Ruminants they are chiefly remarkable for the development of a broad lamellar expansion between the head and the processus longus of the malleus. In some cases the malleus of the foetus differs strikingly from that of the adult. Among Perissodactyla the Rhinoceros and Tapir have the malleus of a low type, recalling those of Marsupials; while in the Horse the head is well developed, and the malleus is of a higher type.

The ossicles of Procavia, which recall those of the Equidae, are chiefly remarkable for the small size of the body of the incus. In Elephants the ossicles are large and massive.

In the Rodentia (fig. 100, C) the malleus is generally characterised by a very broad manubrium. In many genera such as Bathyergus, and most of the Hystricomorpha such as Hystrix, Chinchilla and Dasyprocta, the malleus and incus are ankylosed together.

Carnivora. In Carnivora vera the most striking feature of the malleus is the occurrence of a broad lamellar expansion between the head and neck and the processus longus. This however does not occur in some Viverridae. In the Carnivora vera the incus and stapes are small as compared with the malleus, but in the Pinnipedia they are large. In the Pinnipedia the auditory ossicles have a very dense consistence, and except in the Otariidae are very large. The stapes frequently has no canal, or only a very small one.

In Insectivora the characters of the auditory ossicles are very diverse. Many forms such as shrews, moles, hedgehogs, and the Centetidae have a low type of malleus resembling that of Edentates. Chrysochloris has very extraordinary auditory ossicles. The head of the malleus is drawn out into a great club-shaped process, the incus is long and narrow, and differs much from the ordinary type.

In Chiroptera the ossicles and especially the malleus much resemble those of shrews. The stapes is always normal in character, never becoming at all columelliform.

Primates. In Man and the Anthropoid Apes the malleus has a rounded head, a short neck, and the manubrium, a processus longus and a processus brevis. The incus consists of an anvil-shaped portion from which arises a long tapering process. The stapes has diverging crura and consequently a wide canal. The crura in other monkeys do not diverge so much as in man and anthropoid apes. The New World monkeys have no neck to the malleus.

The Sternum168.

In Monotremes and most Marsupials the sternum does not present any characters of special importance. The presternum is strongly keeled in Notoryctes.

The sternum in Edentates is very variable: in the Sloths it is very long, the mesosternum of Choloepus having twelve segments. In the ant-eaters and armadillos the presternum is broad and sometimes as in Priodon strongly keeled. In Manis macrura the xiphisternum is drawn out into a pair of cartilaginous processes about nine inches long.

In the Sirenia the sternum is simple and elongated, and of fairly equal width throughout, in the adult it shows no sign of segmentation. Its origin from the union of two lateral portions can be well seen in Manatus.

Two distinct types of sternum are met with in the Cetacea. In the Odontoceti the sternum consists of a broad presternum followed by three or four mesosternal segments, but with no xiphisternum. Indications of the original median fissure can be traced, and are very evident in Hyperoödon. In the Mystacoceti, on the other hand, the sternum consists simply of a broad flattened presternum which is sometimes more or less heart-shaped, sometimes cross-shaped. Only a single pair of ribs are united to it.

The sternum in Ungulata is generally long and narrow and formed of six or generally seven segments. The presternum is as a rule small and compressed, often much keeled, especially in the horse and tapir. The segments of the mesosternum gradually widen as followed back and the xiphisternum is often terminated by a cartilaginous plate.

In the Rodentia the sternum is long and narrow and generally has a large presternum, and a xiphisternum terminated by a broad cartilaginous plate.

In the Carnivora, too, the sternum (fig. 76) is long and narrow and formed of eight or nine pieces, all of nearly the same size. The xiphisternum generally ends in an expanded plate of cartilage.

In Insectivora the sternum is well developed but variable. The presternum is commonly large and is sometimes as in the Hedgehog (Erinaceus) bilobed in front, sometimes as in the Shrew (Sorex) trilobed. It is especially large in the Mole (Talpa) and is expanded laterally and keeled below.

In the Chiroptera the presternum is strongly keeled and so is sometimes the mesosternum.

Among Primates, in Man and the Anthropoid Apes the sternum is rather broad and flattened; the mesosternum consists of four segments which are commonly fused together and the xiphisternum is imperfectly ossified.

The Ribs.

Free ribs are borne as a rule only by the thoracic vertebrae; ribs may be found in other regions, especially the cervical and sacral, but these are almost always ankylosed to the vertebrae. As a general rule the first thoracic rib joins the presternum, while the succeeding ones are attached between the several segments of the mesosternum. Some of the posterior ribs frequently do not reach the sternum; they may then be attached by fibrous tissue to the ribs in front, or may end freely (floating ribs). There are generally thirteen pairs of ribs, and in no case do they have uncinate processes.

In Monotremes (fig. 102, B) each rib is divided not into two but into three parts, an intermediate portion being interposed between the vertebral and sternal parts. The sternal ribs are well ossified, and some are very broad and flat. The intermediate portions are unossified, those of the anterior ribs are short and narrow, but they become longer and wider further back.

In Marsupials there are almost always thirteen pairs of ribs, whose sternal portions are very imperfectly ossified. Notoryctes has fourteen pairs of ribs, eight of which are floating: the first rib is very stout, and is abruptly bent on itself to join the sternum. It has no distinct sternal portion. All the other ribs are slender.

Of the Edentates the Sloths have very numerous ribs; twenty-four pairs occur in Choloepus, and half of these reach the sternum. In the Armadillos there are only ten or twelve pairs of ribs, but the sternal portions are very strongly ossified. The first rib is remarkably broad and flat, and is not divisible into vertebral and sternal portions.

In the Sirenia there are a very large number of ribs noticeable for their great thickness and solidity, but not more than three are attached to the sternum.

Cetacea. In the Whalebone whales the ribs are remarkable for their very loose connection both with the vertebral column and with the sternum. The capitula are scarcely developed, and the attachment of the tubercula to the transverse processes is loose. The first rib is the only one connected with the sternum. In the Toothed whales the anterior ribs have capitula articulating with the centra, as well as tubercula articulating with the transverse processes; in the posterior ones, however, only the tubercula remain. Seven pairs of well-ossified sternal ribs generally meet the sternum. In the Physeteridae most of the ribs are connected to the vertebrae by both capitula and tubercula.

In the Ungulata the ribs are generally broad and flattened, and this is especially the case in the genera Bos and Bubalus (fig. 101, 6). The anterior ribs are short and nearly straight, and sternal ribs are well developed. The Artiodactyla have twelve to fifteen pairs of ribs, the Perissodactyla eighteen or nineteen, and Procavia twenty to twenty-two. The Elephant has nineteen to twenty-one pairs, seven of which may be floating ribs.

In the Rodentia there are generally thirteen pairs of ribs, which do not present any marked peculiarities.

The Carnivora have thirteen to fifteen pairs of ribs, whose vertebral portions are slender, nearly straight and subcylindrical, while their sternal portions are long and imperfectly ossified (fig. 76, 5). There is nothing that calls for special remark about the ribs, in either Insectivora or Chiroptera.

Primates. In Man and the Orang (Simia) there are generally twelve pairs of ribs; in the Gorilla and Chimpanzee (Anthropopithecus), and Gibbons (Hylobates), there are thirteen, in the Cebidae twelve to fifteen, and in the Lemuroidea twelve to seventeen pairs. The first vertebral rib is shorter than the others, and the sternal ribs generally remain cartilaginous throughout life, though in man the first may ossify.

Appendicular Skeleton.

The Pectoral Girdle.

By far the most primitive type of the pectoral or shoulder girdle is found in the Monotremata. The scapula (fig. 102, A, 1) is long and recurved, and has only two surfaces, one corresponding to the prescapular[1] fossa, the other to the postscapular[1] and subscapular169 fossae. The coracoid is a short bone attached above to the scapula and below to the presternum; it forms a large part of the glenoid cavity. In front of the coracoid there is a fairly large flattened epicoracoid (fig. 102, 6); there is also a large T-shaped interclavicle (fig. 102, 4), which is expanded behind and rests on the presternum. The clavicles rest on and are firmly united to the anterior border of the interclavicle. This shoulder girdle differs greatly from that of any other mammals, and recalls that of some Lacertilia.

In Marsupials, as in all mammals except the Monotremes, the shoulder girdle is much reduced; there are no epicoracoids and interclavicle, and the coracoid forms simply a small process on the scapula, ossifying from a centre separate from that giving rise to the rest of the bone. The scapula has a long acromion, and a clavicle is always present except in Perameles. Unossified remains of the precoracoids are found at either end of the clavicle. The scapula of Notoryctes has a very high overhanging spine, and there is a second strong ridge running along the proximal part of the glenoid border.

The shoulder girdle of the Edentata shows some very curious variations. In Orycteropus the scapula is of very normal form and the clavicle is well developed. In the Pangolins and Anteaters the scapula is very broad and rounded; there is no clavicle in the Pangolins, and generally only a vestigial one in Anteaters. In Armadillos, Sloths, and Megatheriidae, the acromion is very long and the clavicle is well developed. In the Sloths, Megatherium, and Myrmecophaga, a connection is formed between the coracoid, which is unusually large, and the coracoid border of the scapula, converting the coraco-scapula notch into a foramen. In Bradypus the clavicle is very small, and is attached to the coracoid, which sometimes forms a distinct bone170.

In the Sirenia the scapula is somewhat narrow and curved backwards: the spine, acromion, and coracoid process are moderately developed, and there is no clavicle.

Cetacea. In nearly all the Odontoceti the scapula is broad and somewhat fan-shaped; the prescapular fossa is much reduced, and the acromion and coracoid process form flattened processes, extending forwards nearly parallel to one another. Some of the Mystacoceti, such as Balaenoptera, have a broad, fan-shaped scapula, with a long acromion and coracoid process, extending parallel to one another. Others, such as Balaena, have a higher and narrower scapula, with a smaller coracoid process.

In Ungulata the scapula is always high and rather narrow, and neither acromion nor coracoid process is ever much developed. In no adult Ungulate except Typotherium is there any trace of a clavicle, but a vestigial clavicle has been described in early embryos of sheep171.

Ungulata vera. In the Ruminantia the suprascapular region (fig. 103, 5) is very imperfectly ossified, and when this is removed the upper border of the scapula is very straight (fig. 101, 5). The spine is prominent, and generally has a fairly well-marked acromion. In Hippopotamus the acromion is fairly prominent, but in the other Suina, though the spine is prominent, the acromion is not developed. The Perissodactyla have no acromion, but while the Equidae and Hyracotherium have the scapula long and slender, with the spine very slightly developed, the other living Perissodactyla have the spine prominent and strongly bent back at about the middle of its length.

Subungulata. Typotherium (Toxodontia) differs from all other known Ungulates in having well-developed clavicles; its scapula has a strong backwardly-projecting process, much like that in Rhinoceros.

Phenacodus (Condylarthra), has a curiously rounded scapula, with the coracoid and suprascapular borders passing imperceptibly into one another. The scapula resembles that of a carnivore more than does that of any existing Ungulate.

Procavia has a triangular scapula with a prominent spine and no acromion; there is a large unossified suprascapular region.

The scapula in the Proboscidea has a large rounded suprascapular border and a narrow, slightly concave glenoid border. The spine is large, and has a prominent process projecting backwards from about its middle. The spine lies towards the front end of the scapula, so that the postscapular fossa is much larger than the prescapular fossa.

In Rodentia the shoulder girdle is of a rather primitive type. The scapula is generally high and narrow, somewhat as in Ruminantia; it differs, however, from the Ruminant scapula in having a high acromion, which is often, as in the Hares and Rabbits, terminated by a long metacromion. The development of the clavicle varies, and sometimes it is altogether absent. It is frequently connected by cartilaginous bands or ligaments (fig. 104, 7 and 9), on the one hand with the scapula, and on the other with the sternum. These unossified bands are remains of the precoracoid. Epicoracoidal vestiges of the sternal ends of the coracoids (fig. 104, 11) are also often present.

In the Carnivora vera the scapula is large, and generally has rather rounded borders. The spine and acromion are well developed, and the prescapular and postscapular fossae are nearly equal in size. The coracoid is very small, and the clavicle is never completely developed, being often absent, as in the Bears and most of their allies. In the Seals (Phocidae) the scapula is elongated and curved backwards, and has a very concave glenoid border. In the Eared Seals (Otariidae) the scapula is proportionally much larger and wider, the prescapular fossa being specially large, and being traversed by a ridge, which converges to meet the spine.

In the Insectivora the shoulder girdle is well developed and, as in Rodents, remains are met with of various parts not generally seen in mammals. In the Shrews the scapula is long and narrow, and has a well-marked spine, whose end bifurcates, forming the acromion and metacromion. The clavicle is long and slender, and is connected with the sternum and acromion by vestiges of the precoracoid. Considerable remains of the sternal end of the coracoid are also found. In Potamogale, however, there are no clavicles. In the Mole the shoulder girdle is greatly developed, and of very remarkable form. The scapula is high and very narrow, with the spine and acromion very little developed. The other shoulder girdle element is an irregular bone, which articulates with the humerus and presternum, and is connected by ligaments with the scapula. This bone appears to represent both the coracoid and the clavicle, being formed partly of cartilage bone, partly of membrane bone.

In the Chiroptera the scapula is large and oval, and has a moderately high spine and a large acromion. The coracoid process is well developed and is often forked. The clavicles are also well developed, and vestiges of the precoracoid and of the sternal end of the coracoid are often found.

In Primates the clavicle and coracoid process are always well developed. In Man and the Gorilla the scapula has a long straight suprascapular border, a well-developed coracoid process and spine, and a large curved acromion. Vestiges of the precoracoid occur at each end of the clavicle. The shape of the scapula varies much in the lower Primates.

The Upper arm and Fore-arm.

In the Monotremata the humerus is short, very broad at each end and contracted in the middle. The radius and ulna are stout and of nearly equal size, while the ulna has a greatly expanded olecranon.

In the Marsupialia the humerus is generally a strong bone, broad at the distal end and having well marked deltoid and supinator ridges, which are specially large in Notoryctes. An ent-epicondylar or supracondylar foramen (fig. 105, 5) is almost always present except in Notoryctes. The radius and ulna are always distinct and well developed, and a certain amount of rotation can take place between them. The ulna of Notoryctes has an enormous hooked olecranon which causes the bone to be nearly twice as long as the radius.

Edentata. The Sloths have long slender arm bones; the humerus is nearly smooth and has a very large ent-epicondylar foramen in Choloepus, but not in Bradypus. The radius and ulna can be rotated on one another to a considerable extent. The humerus in all other Edentates is very strong and has the points for the attachment of muscles much developed, especially in the Armadillos and Megatheriidae. An ent-epicondylar foramen is found in all living forms. The radius and ulna are well developed, but are not capable of much rotation.

In the Sirenia the humerus is well developed and of a normal character. It is expanded at each end and has a prominent internal condyle, a small olecranon fossa, and no ent-epicondylar foramen. In the Dugong and Rhytina there is a bicipital groove and the tuberosities are distinct, but in the Manatee there is no bicipital groove, and the tuberosities coalesce. The radius and ulna are about equally developed and ankylosed together at both ends.

In the Cetacea the arm bones are very short and thick. The humerus has a globular head, and a distal end terminated by two equal flattened surfaces to which the radius and ulna are united. There is no bicipital groove, and the tuberosities coalesce. The radius and ulna are flat expanded bones fixed parallel to one another, but the ulna has a definite olecranon. Scarcely any movement can take place between them and the humerus, and in old animals the three bones are often ankylosed together.

In the Ungulata vera the humerus is stout and rather short. The great tuberosity is always large and often overhangs the bicipital groove, it is especially large in Titanotherium (Brontops). There is never an ent-epicondylar foramen. The radius is always large at both ends, but the condition of the ulna is very variable. Sometimes, as in Tapirus, Rhinoceros, Macrauchenia, Suina and Tragulina, the ulna is well developed, and quite distinct from the radius; but in most forms, although complete, it is much reduced distally, and is fused to the radius. Sometimes, as in the Horse and Giraffe, it is reduced to the olecranon and to a very slender descending process which does not nearly reach the carpus. In the Tylopoda, though the ulna is complete and its distal end is often distinct, it has coalesced with the radius throughout its whole length; the olecranon is generally very large.

Subungulata. In the large Condylarthra the humerus has an ent-epicondylar foramen, and the radius and ulna are stout bones nearly equal in size.

In Procavia the humerus is rather long, and has a very prominent greater tuberosity, and a large supra-trochlear fossa, but no ent-epicondylar foramen.

In the Proboscidea the humerus is marked by a greatly developed supinator ridge, and is very long, longer than the radius and ulna. The ulna has a remarkable development, having its distal end larger than that of the radius, it has also a larger articular surface for the humerus than has the radius.

In Rodentia the humerus varies much in its development according to the animal’s mode of life. In the Hares it is long and straight, with a small distal end, and a slight deltoid ridge. In the Beaver on the other hand the deltoid and supinator ridges are considerably developed. There is generally a large supra-trochlear fossa, but no ent-epicondylar foramen.

Carnivora. In the Carnivora vera the humerus has large tuberosities, a prominent deltoid ridge and a deep olecranon fossa. The shaft is generally curved, and an ent-epicondylar foramen is often found, though not in the Canidae, Hyaenidae, and Ursidae. The radius and ulna are never united. The radius (fig. 77, B) has a very similar development throughout its whole length, while the ulna has a large olecranon (fig. 77, C, 11) and a shaft tapering somewhat towards the distal end.

In the Pinnipedia the arm bones are very strongly developed. The humerus has a very prominent deltoid ridge, and the proximal end of the ulna and distal end of the radius are much expanded.

In the Insectivora the arm bones are well developed, and the radius and ulna, though sometimes united, are generally distinct; as a rule there is an ent-epicondylar foramen, but this is absent in the Hedgehog. The Mole has an extraordinary humerus, very short and curved, and much flattened and expanded at both ends. It articulates both with the scapula and coraco-clavicle. The ulna has a greatly developed olecranon.

In the Chiroptera both humerus and radius are exceedingly long and slender; the ulna is reduced to little more than the proximal end and is fused to the radius. There is no ent-epicondylar foramen.

All Primates have the power of pronation and supination of the fore-arm, by the rotation of the distal end of the radius round that of the ulna.

In Man and the Anthropoid Apes the humerus is long and straight, and has a globular head; neither of the tuberosities, nor the deltoid nor supinator ridges are much developed. The olecranon fossa is deep and there is no ent-epicondylar foramen. The radius is curved and has a narrow proximal, and expanded distal end, the ulna is straighter than the radius and has the distal end much smaller than the proximal; the olecranon is not much developed.

In the lower Primates, although the radius and ulna are always quite separate, the power of pronation and supination is not nearly so great as in the higher forms. In most of the Cebidae and Lemurs an ent-epicondylar foramen occurs.

The Manus.

The Manus is divisible into two parts, viz. the carpus or wrist, and the hand which is composed of the metacarpals and phalanges. The carpal bones are always modified from their primitive arrangement, sometimes more, sometimes less. One modification however is always found in mammals, viz. the union of carpalia, 4 and 5 to form the unciform bone. Two sesamoid bones are commonly developed, one on each side of the carpus, the pisiform or one on the ulnar side being much the larger and more constant: it has been suggested that these represent respectively vestiges of a prepollex and a post-minimus digit172.

One or more of the five digits commonly present may be lost, and sometimes all are lost except the third. The terminal or ungual phalanges of the digits are commonly specially modified to support nails, claws, or hoofs. There are as a rule two small sesamoid bones developed on the ventral or flexor side of the metacarpo-phalangeal articulations, and sometimes similar bones occur on the dorsal or extensor side.

Monotremata. In Echidna the carpus is broad, the scaphoid and lunar are united and there is no centrale. The pisiform is large and several other sesamoid bones occur. Each of the five digits is terminated by a large ungual phalanx. In Ornithorhynchus the manus is more slender, but the general arrangement is the same as in Echidna.

Marsupialia. The carpus has no centrale and the lunar is generally small or absent. Five digits are almost always present. In Choeropus however the only two functional digits are the second and third, which have very long closely apposed metacarpals; the fourth digit is vestigial, but has the normal number of phalanges, while the first and fifth are absent. The manus in Notoryctes is extraordinarily modified, the scaphoid and all the distal carpalia are apparently fused, the first, second, and fifth digits are very small, the third and fourth, though having only one phalanx apiece, bear each an enormous claw. Lying on and obscuring the ventral surface of the manus is a large bone, probably a sesamoid.

Among the Edentata there is a great diversity in the structure of the manus, the centrale is however always wanting, and except in Manis the scaphoid and lunar are distinct. In the Sloths the manus is very long, narrow, and curved, and terminated by two or three long hooked claws, borne by the second and third, or the second, third and fourth digits. The fifth digit is absent, and the fourth is represented only by a small metacarpal. In the Anteaters the third digit is greatly developed and bears a long hooked claw. In Myrmecophaga all five digits are fairly well though irregularly developed, in Cycloturus the first, fourth, and fifth, are vestigial. In the Armadillos the manus is broad, and has strongly developed ungual phalanges. The digits, though almost always five in number, vary much in their relative arrangement. In Dasypus they are regular, but are remarkably irregular in Priodon. The pollex is absent in Glyptodonts and in Megatherium. In Megatherium the fifth digit is clawless while the second, third, and fourth bear enormous claws. In the Manidae the scaphoid and lunar are united; five digits are present, the third and fourth being very large, and all being terminated by deeply cleft ungual phalanges. In Orycteropus the pollex is absent, while the other digits are terminated by pointed ungual phalanges.

In Sirenia the general structure of the manus is quite of the ordinary mammalian type. In Manatus most of the bones of the carpus are distinct, but in Halicore many, especially those of the distal row, have coalesced. The digits are always five in number and have the normal number of flattened phalanges.

In the Cetacea, on the other hand, the manus is much modified by the fact that the number of phalanges may be greatly increased above the normal number of three, thirteen or fourteen sometimes occurring in each digit. These are believed to be duplicated epiphyses. In the Mystacoceti the manus remains largely cartilaginous, in the Odontoceti it is better ossified, and the phalanges commonly have epiphyses at both ends. In Physeter the carpal bones also have epiphyses. The carpus generally consists of six bones arranged in two rows of three each. Five digits are generally present, but sometimes as in Balaenoptera musculus, there are four, the third being suppressed. Their relative development varies much. The Sperm Whale which till recently was placed in the entrance hall of the Natural History Museum at South Kensington has one phalanx to the first digit, four to the second, five to the third, four to the fourth, and three to the fifth. Generally the manus is short and broad, but sometimes, as in Globicephalus, it is much elongated owing to the great development of the second and third digits.

Ungulata173. The manus of the members of this great order is of very great classificatory and morphological importance. All the members agree in having the scaphoid and lunar distinct, and in almost every case the ends of the digits are either encased in hoofs or provided with broad flat nails. It is by means of characters derived from the manus and pes that the group is subdivided into the Ungulata vera and the Subungulata.

In the Ungulata vera the manus is never plantigrade, and there are not more than four digits, the pollex being almost always completely suppressed: in Cotylops among extinct Artiodactyla however a vestigial pollex is found. The centrale is absent, and the magnum articulates freely with the scaphoid, and is separated from the cuneiform by the unciform and lunar. All the bones of the carpus interlock strongly, and the axis of the third digit passes through the magnum and between the scaphoid and lunar.

There is a very strong distinction between the manus of the suborders Artiodactyla and Perissodactyla. In the Artiodactyla the axis of the manus passes between the third and fourth digits, which are almost equally developed and, except in the Hippopotami and some extinct forms such as Anoplotherium, have their ungual phalanges flattened on their contiguous surfaces.

In all Artiodactyla the third and fourth digits are large, but a gradual reduction in the second and fifth can be well traced. Thus in the Suina the second and fifth digits, though smaller than the third and fourth, are well developed and all four metacarpals are distinct. In the Tragulina too all four metacarpals are developed, and in Dorcatherium the third and fourth commonly remain distinct as in the Suina. In the other Artiodactyla however the third and fourth metacarpals are almost always united, though indications of their separate origin remain. In some Ruminantia, such as many Deer, the second and fifth digits are reduced to minute splint bones attached to the proximal end of the fused third and fourth metacarpals, and to small hoof-bearing phalanges, sometimes attached to splint-like distal vestiges of the metacarpals, sometimes altogether unconnected with any other skeletal structures. In some other Ruminants, such as the Sheep and Oxen, the only remnants of the second and fifth digits are nodules of bone supporting the hoofs, and in others, such as the Giraffe, Anoplotherium commune, some Antelopes and the Tylopoda, all traces of these digits have disappeared. The Camels differ from all living Ungulata vera in not having the distal phalanges completely encased in hoofs, and from all except the Hippopotami in placing a considerable amount of the manus on the ground in walking.

While the manus of the Artiodactyla is symmetrical about a line drawn between the third and fourth digits, that of the Perissodactyla is symmetrical about a line drawn through the middle of the third digit, which is larger than the others and has its ungual phalanx evenly rounded and symmetrical in itself. The most reduced manus in the whole of the mammalia is found in the Horse and its allies, in which the third digit, terminated by a very wide ungual phalanx, is the only one functional. Small splint bones representing the second and fourth metacarpals are attached to the upper part of the third metacarpal. In Hipparion174 and other early horse-like animals the second and fourth digits, though very small and functionless, are complete and are terminated by small hoofs. In Rhinoceros the second and fourth digits are equally developed and nearly as large as the third, and reach the ground in walking, a vestige of the fifth is also present. In the Tapir (fig. 106, A) and Hyracotherium the fifth digit is fully developed but is scarcely functional. In Titanotherium (Brontops) (fig. 106, B) it is nearly as well developed as any of the others, and there is little or no difference between the relative development of the third and fourth digits.

The Chalicotheriidae175, though distinctly Perissodactyles in various respects such as their cervical vertebrae and teeth, differ not only from all other Perissodactyles, but from almost all other Ungulates, in the very abnormal character of their manus. For while the carpus and metacarpus are like those of ordinary Perissodactyles, the phalanges resemble those of Edentates, each second phalanx having a strongly developed trochlea, and each distal one being curved, pointed and deeply cleft at its termination (fig. 106, C).

The Macraucheniidae, while agreeing with Perissodactyles in having only three digits, with the limb symmetrical about a line drawn through the middle of the third, have a carpus which approaches closely to the subungulate condition, the magnum articulating regularly with the lunar, and only to a slight extent with the scaphoid.

In the Subungulata the manus sometimes has five functional digits, and a considerable part of it rests on the ground in walking. The bones of the carpus retain their primitive relation to one another, the magnum articulating with the lunar, but not with the scaphoid. This character does not however hold in the Toxodontia, for in most of the animals belonging to this group the magnum does articulate with the scaphoid. The corner of the scaphoid just reaches the magnum also in Amblypoda.

As far as is known the Toxodontia generally have three, sometimes five digits to the manus, and the third is symmetrical in itself—a Perissodactyloid feature.

In Phenacodus (fig. 107, B) (Condylarthra) all five digits are well developed, the pollex being the smallest. The carpal bones retain their primitive arrangement, the magnum articulating with the lunar and not with the scaphoid. There is no separate centrale.

In the Hyracoidea (fig. 107, C) the manus is very similar to that in Phenacodus, but a centrale is present and the pollex is much reduced.

The manus of the Amblypoda, such as Coryphodon (fig. 107, A) and Uintatherium, is short and broad, with five well developed digits and large carpal bones. The carpals however interlock to a slight extent, and the corner of the magnum reaches the scaphoid.

In the Proboscidea the manus is very short and broad, with large somewhat cubical carpals which articulate by very flat surfaces and do not interlock at all. All five digits are present, and none of them are much reduced in size. The manus in Proboscidea and in Coryphodon is subplantigrade.

In the Tillodontia the manus is plantigrade and has pointed ungual phalanges, in this respect approaching the Carnivora. It differs however from that of all living Carnivora in having the scaphoid and lunar distinct.

In Rodentia the manus nearly always has five digits with the normal number of phalanges: the pollex may however be very small as in the Rabbit, or absent as sometimes in the Capybara. The scaphoid and lunar are generally united, and a centrale may be present or absent. In Pedetes caffer the radial sesamoid is double and the distal bone bears a nail-like horny covering. In Bathyergus the pisiform is double. It is upon these facts that the contention for the former existence of prehallux and post-minimus digits has partly been based.

In living Carnivora the scaphoid, lunar and centrale are always united, forming a single bone. All five digits are present, but as a rule in Carnivora vera the pollex is small, and in Hyaena is represented only by a small metacarpal. Sometimes, as in Cats and Dogs, the manus is digitigrade, sometimes, as in Bears, plantigrade. The ungual phalanges are large and pointed, and in forms like the Cats, whose claws are retractile, they can be folded back into a deep hollow on the ulnar side of the middle phalanx; a small radial sesamoid is often present.

In Pinnipedia the manus is large and flat and the digits are terminated by ungual phalanges which are blunt (sea lions and walrus), or slightly curved and pointed (seals). The pollex is nearly or quite as long as the second digit, and as a rule the digits then successively diminish in size.

The Creodonta differ from living Carnivora in the fact that the scaphoid and lunar are usually separate.

In Insectivora the scaphoid and lunar are sometimes united, sometimes separate, and a separate centrale is usually present. There are generally five digits, but sometimes the pollex is absent. In the Mole the manus is greatly developed and considerably modified. It is very wide, its breadth being increased by the great development of the radial sesamoid which is very large and sickle-shaped. The ungual phalanges are also large and are cleft at their extremities.

In the Chiroptera the manus is greatly modified for the purpose of flight. The pollex is short and is armed with a rather large curved claw, the other digits are enormously elongated, the elongation in the case of the Insectivorous bats being mainly due to the metacarpals, and in the Frugivorous bats to the phalanges. In the Frugivorous bats the second digit is clawed as well as the pollex, in other bats this claw is always absent, and so is often the ungual phalanx, the middle phalanx then tapering gradually to its termination.

In Primates as a rule the manus is moderately short and wide. The carpus has the scaphoid and lunar distinct, and generally also the centrale; sometimes however, as in Man, the Gorilla, Chimpanzee, and some Lemurs, the centrale has apparently fused with the scaphoid. There are almost always five well-developed digits, but in the genera Colobus and Ateles the pollex is vestigial.

The magnum in man is the largest bone of the carpus. The pisiform also is well developed, but there is no radial sesamoid. In Man, the Gorilla, Chimpanzee, and Orang, the carpus articulates only with the radius, in most Primates it articulates also with the ulna. The third digit of the Aye-Aye (Chiromys) is remarkable for its extreme slenderness.

The Pelvic Girdle.

The pelvic girdle in all mammals except the Sirenia and Cetacea consists of two halves, usually united with one another at the symphysis in the mid-ventral line, and connected near their upper ends, with the sacral vertebrae. Each half forms one of the innominate bones, and includes at least three separate elements, a dorsal bone, the ilium, and two ventral bones, the ischium and pubis. Very often a fourth pelvic element, the acetabular or cotyloid bone, occurs.

In the Monotremata the pelvis is short and broad, and the pubes and ischia meet in a long symphysis. The acetabulum is perforated in Echidna as in birds, but not in Ornithorhynchus. A pair of elongated slender bones project forwards from the edge of the pubes near the symphysis; these are sesamoid bones formed by ossifications in the tendons of the external oblique abdominal muscles, and are generally called marsupial bones.

In the Marsupialia the ilia are generally very simple, straight, and narrow, while the pubes and ischia are well developed and meet in a long symphysis. Marsupial bones are nearly always prominent, but are not developed in Thylacinus or Notoryctes. The ischium often has a well-marked tuberosity and in Kangaroos the pubis bears a prominent pectineal process on its anterior border close to the acetabulum. The pelvis in Notoryctes differs much from that in all other Marsupials, the ilium and ischium being ankylosed with six vertebrae in a manner comparable to that of many Edentates.

In the Edentata the pelvis is generally well developed, but the symphysis is very short. In the Sloths the pelvis is rather weak and slender, the obturator foramina are very large and the ischia do not meet in a symphysis. In the Megatheriidae the pelvis is exceedingly wide and massive, and is firmly ankylosed with a number of vertebrae. In the Armadillos, Glyptodonts, Anteaters, and Pangolins it is much developed and firmly united to the vertebral column by both the ilia and the ischia. In Orycteropus however the ischium does not become united to the vertebral column, and the pubis generally has a strongly developed pectineal process.

In the Sirenia the pelvis is quite vestigial. In the Dugong it consists on each side of two slender bones, one of which represents the ilium and the other the ischium and pubis; the two bones are placed end to end and are commonly fused together. The ilium is attached by ligament to the transverse process of one of the vertebrae. In the Manatee each half of the pelvis is represented by a triangular bone connected by ligaments with its fellow and with the vertebral column. In neither Manatee nor Dugong is there any trace of an acetabulum but one can be made out in Halitherium.

In the Cetacea the pelvis is even more vestigial than in the Sirenia, consisting simply of a pair of small straight bones which probably represent the ischia, and lie parallel to and below the vertebral column at the point where the development of chevron bones commences.

In Ungulata vera the pelvis is generally rather long and narrow. The ilium is flattened and expanded in front (fig. 103, 8), but becomes much narrower and more cylindrical before reaching the acetabulum. Both pubis and ischium contribute to the symphysis which is often very long. The ischia are large and have prominent tuberosities, especially in Artiodactyles. In most Ruminantia there is a deep depression, the supra-acetabular fossa above the acetabulum, but this is not found in the Suina or Tylopoda.

Subungulata. In Procavia the pelvis is long and narrow, and bears resemblance to that in Artiodactyles.

The Proboscidea have a very large pelvis set nearly at right angles to the vertebral column; the ilium is very wide, having expanded iliac and gluteal surfaces, and a narrow sacral176 surface. The pubes and ischia are rather small, but both meet their fellows in the symphysis. Uintatherium (suborder Amblypoda) also has a large and vertically placed pelvis (fig. 108) with a much expanded ilium. The pelvis however differs from that of the Proboscidea in the fact that the ischia do not meet in a ventral symphysis.

In many Rodentia the ilia have their gluteal, iliac, and sacral surfaces of nearly equal extent; in the Hares, however, the gluteal and iliac surfaces are confluent. The pubes and ischia are always well developed and sometimes, as in the Hares, the acetabular bone also. In these animals the pubis does not take part in the formation of the acetabulum, and the ischium bears on its outer side a well-marked ischial tuberosity.

In the Carnivora the pelvis is long and narrow. The iliac surfaces (fig. 78, A, 5) are very small and the sacral large; the crest or supra-iliac border is formed by the union of the sacral and gluteal surfaces. The symphysis is long and includes part of both pubis and ischium. The ischial tuberosity (fig. 78, A, 10) is often well marked, and sometimes as in Viverra the acetabular bone is distinct. In the Pinnipedia the pelvic symphysis is little developed, or sometimes not developed at all, and the obturator foramina are remarkably large.

In some Insectivora such as Galeopithecus, there is a long pelvic symphysis, in others such as Erinaceus and Centetes, it is very short, in others again such as Talpa and Sorex, there is no pelvic symphysis. The acetabular bone is exceptionally large in Talpa and Sorex.

In Chiroptera the pelvis is small and narrow, and in the great majority of cases the two halves do not meet in a ventral symphysis. The pubis has a strongly developed pectineal process, which occasionally unites with a process from the ilium enclosing a large pre-acetabular foramen.

Primates. In Man and the Anthropoid Apes the pelvis is very large and wide, and the ilium has much expanded iliac and gluteal surfaces. The symphysis is rather short and formed by the pubis alone. The acetabulum is deep and the obturator foramen large, and there is frequently a well-marked ischial tuberosity. In the lower Anthropoidea the ilium is long and narrow and has a small iliac surface. The ischial tuberosities are large in the old world monkeys.

The Thigh and Shin.

In the Monotremata the femur is short, rather narrow in the middle, and expanded at each end. The great and lesser trochanters are large and about equally developed, but there is no third trochanter. The fibula is very large and is expanded at its proximal end, forming a flattened plate much resembling an olecranon. The patella is well developed.

In the Marsupialia there is no third trochanter to the femur, the fibula is well developed but not the patella as a general rule. Notoryctes has a femur with a prominent ridge extending some little way down the shaft from the great trochanter; the tibia has a remarkably developed crest, and the fibula has its proximal end much expanded and perforated; there is an irregularly shaped patella closely connected with the proximal end of the tibia.

Edentata. In the Sloths the leg bones are all long and slender. The femur has no third trochanter, and the fibula is complete and nearly equal in size to the tibia. In the Megatheriidae the leg bones are extraordinarily massive, the circumference of the shaft of the femur in Megatherium equalling or exceeding the length of the bone. There is no third trochanter in Megatherium. In most of the remaining Edentata the leg bones are strongly developed. The femur in the Armadillos and Aard Varks has a strong third trochanter, and the tibia and fibula are both large and are commonly ankylosed together at either end. The limb bones are very massive also in the Glyptodonts.

Sirenia. In no living Sirenian is there any trace of a hind limb, but in Halitherium a vestigial femur is found, which articulates with the pelvis by a definite acetabulum.

In the Mystacoceti among the Cetacea small nodules of bone or cartilage occur connected with the vestigial pelvis, and may represent the femur and tibia. No trace of the skeleton of the hind limb is known in the Odontoceti.

In the Ungulata vera the femur is noticeable for the size of the great trochanter (fig. 109, 2); there is no definitely constricted neck separating the head from the rest of the bone, and the lesser trochanter (fig. 109, 3) is not very prominent. All Perissodactyles except the Chalicotheriidae show a strongly marked third trochanter, but this is absent in all known Artiodactyles. The development of the fibula in general corresponds to that of the ulna. In Rhinoceros, Macrauchenia, Tapirus and the Suina it is distinct and fairly well developed; in the Tragulina on the other hand it is vestigial, being reduced to the proximal end only. In the Ruminantia and Tylopoda also, it is much reduced forming merely a small bone attached to the distal end of the tibia, sometimes, as in the Red deer a slender vestige of the proximal end also is preserved quite detached from the distal portion; in the Horse this proximal portion is all that there is found of the fibula. The progressive diminution of the fibula can be well seen in the series of forms that are regarded as the ancestors of the Horse. The patella of the Ungulata vera is well ossified, but fabellae177 are not usually found.

Subungulata. Of the Toxodontia, Toxodon has no third trochanter while Typotherium and Astrapotherium have one. In the Condylarthra the femur has well-marked lesser and third trochanters, and the fibula and patella are well developed. In the Hyracoidea there is a slight ridge on the femur in the place of the third trochanter, the fibula is complete, but is generally fused to the tibia at its proximal end.

Of the Amblypoda, Coryphodon has a third trochanter, but Uintatherium has none; in this respect, in the vertical position and general appearance (fig. 108) of the limb, and in the articulation of the fibula with the calcaneum, the leg of Uintatherium closely approaches that of the Proboscidea.

In the Proboscidea the femur is very long and straight, the development of trochanters is slight, and the fibula though slender is complete and articulates with the calcaneum.

A third trochanter is found in the Tillodontia.

In Rodentia the femur is variable, the great trochanter is generally large and so sometimes is the third as in the Hares. In most Rodents as in the Beaver the fibula is distinct, sometimes as in the Hares it is united distally with the tibia. The patella is well developed, and so too are the fabellae as a general rule.

Carnivora. In the Carnivora vera the femur (fig. 79, A) is generally rather straight and slender, and has a very distinct head. The fibula (fig. 79, C) is always distinct and there is generally a considerable interval between it and the tibia. Fabellae (fig. 79, 7) are commonly present.

In the Pinnipedia the femur is short, broad and flattened, having a prominent great trochanter. The fibula is nearly as large as the tibia, and the two bones are generally ankylosed together at their proximal ends.

The Creodonta differ from all living Carnivores in having a femur with a third trochanter.

In the Insectivora a third trochanter is sometimes developed. The fibula is sometimes distinct, sometimes fused distally with the tibia, thus differing from that of a Carnivore.

In Chiroptera the femur is straight, slender and rather short, with a small but well-developed head. The fibula may be well developed or quite vestigial or absent. Owing to the connection of the hind limb with the wing membrane the knee joint is directed backwards.

In Primates the femur is rather long and slender, having a nearly spherical head and large great trochanter. The tibia and fibula are always distinct and well developed. Fabellae are not found in the highest forms but are generally present in the others.

The Pes.

The skeleton of the pes is in most respects a counterpart of that of the manus. Just as in the manus if one digit is absent it is the pollex, so in the pes it is the hallux. But while in the manus the third digit is always well developed, however much the limb may be modified, in the pes any of the digits may be lost. In all mammals the tibiale and intermedium fuse to form the astragalus, and the fourth and fifth tarsalia to form the cuboid. Sesamoid bones are considerably developed. In almost every case the phalanges and first metatarsal have epiphyses only on their proximal ends, while the remaining four metatarsals have epiphyses only on their distal ends.

In the Monotremata all the usual tarsal bones are distinct, and the five digits have the normal number of phalanges. Several sesamoid bones are developed, the most important one, found only in the male, being articulated to the tibia and bearing the curious horny spur. The ungual phalanges of the pes like those of the manus, are deeply cleft at their extremities. In the Echidnidae the pes is turned outwards and backwards in walking.

In the Marsupialia the pes is subject to great modifications, but in every case the seven usual tarsal bones are distinct. In the Didelphyidae the foot is broad, all five digits are well developed, and the hallux is opposable to the others. In the Dasyuridae the foot is narrow, and the hallux may be very small, or as in Thylacinus completely absent. In Notoryctes the pes is much less abnormal than the manus, and all five digits have the usual number of phalanges. The fifth metatarsal has a curious projecting process, and there is a large sesamoid above the hallux. In the Wombats (Phascolomyidae) the foot is short and broad, the digits are all distinct, and the hallux is divaricated from the others.

In the remaining marsupials the second and third metacarpals and digits are very slender, and are enclosed within a common integument. This condition is known as syndactylism, and its effect is to produce the appearance of one toe with two claws. In the Kangaroos (Macropodidae) the pes is very long and narrow, owing to the elongation of the metacarpals. The fourth digit is greatly developed, the fifth moderately so, while the hallux is absent, and the second and third digits are very small. The Peramelidae have the foot constructed on the same plan as in the Kangaroos, and in one genus Choeropus the same type of foot is carried to a greater extreme than even in the Kangaroos. Thus the fourth digit is enormously developed, the second and third are small, and the fifth smaller still, while the hallux is absent. In the Phalangers and Koalas though the second and third toes are very slender, the hallux is well developed and opposable.

Edentata. In the Sloths the pes much resembles the manus, being long and narrow, but in both genera the second, third and fourth digits are well developed. Most of the other Edentates have a but little modified pes with the normal number of tarsal bones and the complete series of digits. In Cycloturus however the hallux is vestigial and it is absent in Glyptodonts. Megatherium has a greatly modified pes, the hallux is absent, and the second digit vestigial, while the third is very large, having an enormous ungual phalanx. The calcaneum too is abnormally large.

No trace of the pes occurs in either Sirenia or Cetacea.

In the Ungulata the pes like the manus is subject to much variation and is of great morphological importance.

In the Ungulata vera the pes is never plantigrade and never has more than four digits, the hallux being absent. The cuboid always articulates with the astragalus, and the tarsal bones strongly interlock. As was the case also with the manus, the pes is formed on two well-marked types characteristic respectively of the Artiodactyla and Perissodactyla.

Artiodactyla. Just as in the manus, the third and fourth digits are well and subequally developed; their ungual phalanges have the contiguous sides flat, and the axis of the limb passes between them, and between the cuboid and navicular. The astragalus has both the proximal and distal surfaces pulley-like, and articulates with the navicular and cuboid by two facets of nearly equal size. The calcaneum articulates with the lower end of the fibula if that bone is fully developed.

In the Suina four toes are developed, and though in the Peccaries the third and fourth metatarsals are united, they are all distinct in most members of the group, as are all the tarsal bones. In the Hippopotami the four digits are of approximately equal size, and the middle ones do not have the contiguous faces of their ungual phalanges flattened.

In the Tragulina the cuboid, navicular, and two outer cuneiforms are united forming a single bone; all four metatarsals are complete and the two middle ones are united. In the Tylopoda and Anoplotherium commune only the third and fourth digits are developed, their metatarsals are free distally, but are elsewhere united. In the Ruminantia the cuboid and navicular are always united and so are the second and third cuneiforms, while in Cervulus all four bones are united together. The third and fourth metatarsals in Ruminants are always united in the same way as are the third and fourth metacarpals, while the second and fifth are always wanting. In Deer the second and fifth digits are usually each represented by three small phalanges, but in the Giraffe and most Bovidae the bones of these digits are wanting.

In the Perissodactyla the pes like the manus is symmetrical about a line drawn through the third digit; this line when continued passes through the external cuneiform, navicular and astragalus. The astragalus has its distal portion abruptly truncated, and the facet by which it articulates with the cuboid is much smaller than that by which it articulates with the navicular. The calcaneum does not articulate with the fibula. The tarsus in Macrauchenia like the carpus differs from that of other Perissodactyles and resembles that of Subungulates in having the bones arranged in lines with little or no interlocking. The calcaneum resembles that of Artiodactyles in having a small facet for articulation with the fibula. Tapirus (fig. 110, A), Rhinoceros (fig. 110, B) and Titanotherium have a short and broad foot with the usual tarsal bones and three well-developed digits,—a number never exceeded by any Perissodactyle. From this tridactylate limb a series of stages is exhibited by various extinct forms leading gradually to the condition met with in the Horse (fig. 110, D) in which the third toe is greatly developed, while the second and fourth are reduced to slender metatarsals attached to the proximal half of the third metatarsal.

In Chalicotherium and Agriochoerus the pes has the same abnormal characters as the manus, the digits being clawed and the ungual phalanges in Chalicotherium deeply cleft.

In the Subungulata the pes is sometimes plantigrade and pentedactylate, the cuboid sometimes does not articulate with the astragalus, and the tarsal bones sometimes do not interlock.

In Typotherium (Toxodontia) the hallux is absent and the other four digits are well developed; in Toxodon and Nesodon the pes is tridactylate. The tarsal bones have the regular Subungulate arrangement, the cuboid not articulating with the astragalus. The calcaneum articulates with the fibula as in Artiodactyles. The astragalus in most forms, but not in Astrapotherium, resembles that of the Ungulata vera in having a grooved proximal surface.

In Phenacodus (Condylarthra) the tarsus is very little modified, five digits are present, the first and fifth being small and not reaching the ground.

In Procavia only the three middle digits are present with a vestige of the fifth metacarpal.

In the Amblypoda the pes (fig. 108) is very short and broad, all five digits are functional, and at any rate in Coryphodon plantigrade, the hallux being the smallest. The astragalus is very flat, and the tarsals interlock to a slight extent, the cuboid articulating with both calcaneum and astragalus.

The pes in the Proboscidea much resembles that in the Amblypoda, but differs in that the astragalus does not articulate with the cuboid, the tarsals not interlocking at all.

In the Rodentia the structure of the foot is very variable. In Beavers the foot is very large, all five digits being well developed; the fifth metatarsal articulates with the outer side of the fourth metatarsal, and not with the cuboid, and there is a large sesamoid bone on the tibial side of the tarsus. In the Rats, Porcupines and Squirrels, there are five digits, in the Hares only four, and in the Capybara and some of its allies only three. In the Jerboa (Dipus) a curious condition of the pes is met with, as it consists of three very long metatarsals fused together and bearing three short toes, each formed of three phalanges. Lophiomys differs from all other Rodents in having the hallux opposable.

Carnivora. In the Carnivora vera the pes is regular and shows little deviation from the normal condition. All the usual tarsal bones are present, but sometimes as in the Dogs, Cats, and Hyaenas, the hallux is vestigial. Sometimes as in the Bears the pes is plantigrade, sometimes as in the Cats and Dogs it is digitigrade. In this respect and in the character of the ungual phalanges, the pes closely corresponds with the manus. In the Sea Otter (Latax) the foot is large and flattened and approaches in character that of the Pinnipedia.

In the Pinnipedia the pes differs much from that in the Carnivora vera. In the Seals in which the foot cannot be used for walking, and is habitually directed backwards, the first and fifth digits are much longer and stouter than any of the others. In the Sea Lions which can use the pes for walking, the digits are all of nearly the same length, and in the Walrus the fifth is somewhat the longest.

In the Insectivora the pes is almost always normal, and provided with five digits.

In the Chiroptera the pes is pentedactylate, and the digits are terminated by long curved ungual phalanges. In some genera the toes have only two phalanges. The calcaneum is sometimes produced into a long slender process which helps to support the membrane between the leg and the tail.

Among the Primates Man has the simplest form of pes. In Man all five digits are well developed, the hallux being considerably the largest. Sesamoid bones occur only under the metatarso-phalangeal joint of the hallux.

In the other Primates the internal cuneiform has a saddle-shaped articulating surface for the hallux, which is obliquely directed to the side of the foot and opposable to the other digits. Two sesamoid bones are usually developed below each metatarso-phalangeal joint, and one below the cuboid. The second digit in Lemurs, and all except the hallux in Chiromys have pointed ungual phalanges; in all other cases the ungual phalanges are flat. In some of the Lemuroidea, especially Tarsius, the tarsus is curiously modified by the elongation of the calcaneum and navicular.