Extinct Families of Artiodactyla

The origin of the Artiodactyla is placed by Cope in the family Pantolestidae, allied to the genus Protogonodon of the Condylarthra. As, however, this family is represented by but a few back teeth and a fragment of the hind-foot, it seems premature to regard it as the necessary starting-point of the Bunodont and Ruminant groups.

Fam. Anthracotheriidae.—This well-known and ancient family consists of creatures of for the most part a Pig-like form, with teeth approaching the selenodont shape, and a complete dentition. The carpals, tarsals, metacarpals, and metatarsals are all free. The toes are four (or five) to each foot, with the outermost beginning to be reduced. These of course are all generalised and primitive characters, pointing nowhere in particular, except, of course, to an Artiodactyle stock, on account of the teeth and the two predominating toes.

The type genus of the family, Anthracotherium, is not, as its name might seem to denote, a relic of the Carboniferous period; its remains were found in lignite, which may also show that it was at least semi-aquatic in habit. Its form, however, must have been Pig-like, so at least one would presume from the elongated skull and shortish legs. There were species as great as a Rhinoceros, and smaller forms. The genus began in the Oligocene and continued down to the Pliocene. It is known from Europe, Asia, and America.

The skull is long with a prominent sagittal crest. The facial part is also very long, and the orbits are not closed by a bony ring. The premolars are simple teeth; the molars distinctly bunodont with a tendency in one or two to the selenodont condition. The canines are powerful, as are also the incisors. The scapula has been specially compared with that of the Camel. It has no acromion, which is usually though not always absent in Ungulates. An ally of the present animal, for instance, the Hippopotamus, has the acromion developed. The radius and ulna, the tibia and fibula, are all fully developed.

Ancodus (or Hyopotamus, as it has been called) is also Oligocene in range, and its remains have been found in the same countries as have those of Anthracotherium. Both genera are indeed closely allied. Ancodus seems to be a more slightly-built creature. The skull looks weaker, but presents much the same features of organisation. In A. velaunus, a species found in French rocks, a metacarpal of digit I. was present in the manus, while A. brachyrhynchus had a completely five-fingered manus.

The Miocene genus Merycopotamus (from the lower layers of the Siwalik formation in India) is more distinctly selenodont than the forms already discussed. On this ground it has been placed in a separate sub-family. As, however, in other respects it does not depart from the Anthracotherian type of structure, this proceeding seems to be hardly necessary. There are two species known, of which one, M. nanus, is, as its name denotes, a dwarf form.

Fam. Caenotheriidae.—While the last family consisted of animals rather more akin to the Pigs, the present is more Pecorine in its characters. The molars are selenodont; but as in the Tragulidae the premolars are more of the nature of cutting teeth. The dentition, like that of so many of these early Ungulates, is complete, and the canines are not prominent. The feet are four-toed, the lateral toes not reaching the ground.

The principal genus is the Eocene and Miocene Caenotherium. Of this genus there were a considerable number of species all European in range, and of small size—not more than a foot or so in length. Their small size is suggestive of the Chevrotains. In the skull the orbital cavity is nearly or quite surrounded by bone, and the tympanic bulla is large and inflated. A common feature of Artiodactyles, a failure of the nasals and maxillae to meet at the side of the face, is to be seen in this ancient forerunner of the Pecora.

Plesiomeryx, also European, and from the same geological horizon, is a very closely allied form.

Fam. Xiphodontidae.—This family consists of slender, small Artiodactyles which are, like the Caenotheriidae, related to the Pecora. They are confined in their range to Europe.

The type genus Xiphodon has selenodont molars and elongated, slender, cutting premolars. The dentition was complete and the canines not highly developed. Like Caenotherium, Xiphodon was a hornless creature, but with only two toes, the two lateral digits being represented by the merest rudiments of metacarpals. The other metacarpals were unusually long.

Amphimeryx (also called Xiphodontotherium) is much more imperfectly known, but belongs to this family or to that of the Caenotheriidae. Dichodon is another member of the same family.

Fam. Oreodontidae.—This family, consisting of numerous genera, is limited to the North American continent. Its range in time is from the Eocene to the Lower Pliocene. The family as a whole is to be distinguished by a number of primitive characters. The dentition is complete; the feet are four- or even five-toed; the orbit is sometimes open behind. The canines of the lower jaw are not more pronounced than the incisors. The characteristics of the group will be further developed by a consideration of some of the principal genera which are included in this family.

Oreodon is a Miocene form about as large as a Peccary. The skull has a short face with a completely-closed orbital cavity. In front of the orbit is a deep pit, not a mere deficiency of ossification, such as occurs in many Artiodactyles. This is placed on the lachrymal bone, and is in fact a lachrymal fossa, such as occurs in other forms. The odontoid process of the axis vertebra is somewhat cheese-taster shaped, as in recent Artiodactyles. There are fourteen dorsal vertebrae and a very large number of caudals. The radius and the ulna are completely separated, as are the carpals. There are five digits to the fore-limbs. The fibula is complete and independent. The hind-foot is four-toed. Several species of the genus are known.

Merycochoerus is an allied Miocene genus. It is more massive in form than the last, but otherwise does not present differences of importance.

Mesoreodon is another genus of this family which presents some curious features of organisation. In the skull and teeth there is nothing very noteworthy, but the hyoid is remarkable. This appendage of the skull is by no means always preserved, and when it is, it might be denied that it belonged to any particular skull. In the present case there appears to be no doubt as to the identity of the bones, which resemble the corresponding bones of the Perissodactyla much more than they do those of other Artiodactyles. Associated with the bones an ossified thyroid cartilage of the larynx was found. As the skull was that of a male, this character may be a sexual one. It is quite comparable to the ossification of the same cartilage in the American monkey Callithrix. "The function of the bone," observes Professor Scott, "was probably similar to that performed by the enormously-inflated basihyal of the howling monkeys, and must have given to these animals most unusual powers of voice." Another important anatomical fact about Mesoreodon is the apparent existence of a clavicle. It is of course conceivable that the remains of some other animal have got mixed up with that of the individuals upon which the present genus is founded; but failing that, here is a clavicle in an Ungulate. The spine of the scapula possesses a metacromion. This greater development of the spine of the scapula in Artiodactyles than in Perissodactyles is, it is suggested, to be correlated with the earlier loss of the clavicle in the latter group of Ungulates.

Cyclopidius (synonymous with Brachymeryx) is a kind of pug form of Oreodon. The skull is short and broad, and the end of the snout a little turned up. The upper incisors are small and drop out early. On each side of the nasals is a large oval vacuity which is perhaps to be compared to the lateral deficiency to be found in other Artiodactyles. One species of this singular-looking form is appropriately called C. simus.

Other allied genera are Merychyus and Leptauchenia. The former extends as far down as the Lower Pliocene, and is thus one of the newest forms of Oreodontidae.

Agriochoerus (Fig. 178) is placed in a separate sub-family from the types which have just been considered. It is Miocene in range. It differs from Oreodon and its closer relatives by the fact that the orbit is open behind and not closed. The most remarkable fact about this creature is that the terminal phalanges of the digits (five in the fore- and four in the hind-feet) being pointed, seem to suggest their encasement with claws rather than hoofs. The pollex, though small, seems to have been opposable. As with other Oreodonts, the molars are selenodont. The premaxillae are toothless—at least in adults, for two teeth are present in the young. There are several species. Agriochoerus, like Oreodon and primitive Ungulates in general, had a long tail. The genus thus shows a mixture of ancient and specialised characters.

Fig. 178.—Skeleton of Agriochoerus latifrons. × ⅛. (After Wortman.)

The most ancient form of Oreodont is Protoreodon. This is Eocene, and became extinct during that period. It had a complete dentition, open orbit, and no lachrymal fossa. The fore-feet were five-toed, the hind four-toed.

Fam. Anoplotheriidae.—This family is entirely Eocene in point of time, and is unknown outside Europe. The dentition of the group is complete; the molars are seleno-bunodont, like those of the Anthracotheriidae. The bones of the carpus, tarsus, metacarpus and metatarsus are all free; the toes are four to two in number on each foot. The orbit is widely open behind. The tail is long, as in Xiphodon, etc.

These general characters only just serve to differentiate the family; but they illustrate its archaic character, in which it resembles the Xiphodontidae, and even more the Anthracotheriidae. A survey of some of the genera which have been assigned to the family will bring out other features in the organisation of these very ancient Artiodactyles.

Anoplotherium is so called on account of the fact that it is, like all ancient Artiodactyles, without horns or claws. Tusks it might have, but as a matter of fact has not. There are, as in Artiodactyles generally, nineteen dorso-lumbar vertebrae; the long tail has numerous chevrons. The shoulder blade has a well-marked acromion and a distinct coracoid process; it is wide proximally. The bones of the fore-arm and fore-leg are, as is usual in primitive Artiodactyles, separate.

In the skull the chief features, in addition to that mentioned in the definition of the family, are the large size of the paroccipital processes; there is no fossa lachrymalis or deficiency in the side of the face. The animal is three-toed, both in the fore- and hind-limbs. The second toe is nearly as large as the Artiodactyle third and fourth. There are tiny rudiments of the two remaining fingers. The hind-foot is also three-toed, and there is a trace of the hallux. The fingers are so widely separated and divergent from each other that it has been suggested that the animal had webbed feet and inhabited marshes, in which it swam by the aid of its long tail. The creature was the size of a Tapir.

Closely resembling Anoplotherium are a number of other genera.

Diplobune ( = Hyracodontotherium) was much like the last, but was a more delicately-formed animal. The fingers and toes (three of each) end in such sharply-pointed phalanges that claws seem to be almost suggested. There are several species of this genus. Dacrytherium differs by the presence of a lachrymal fossa.

Dichobune has four-toed extremities, of which the lateral ones are more slender and shorter than the two middle ones. As in other Anoplotheriidae, the anterior premolars are furnished with a sharp cutting edge.