THE existing families of the Carnivora, spoken of in the last lecture, do not appear to have been distinctly differentiated in the Eocene period, at all events not till towards its close, but the order was represented by other and very singular forms, the systematic position of which is not easy to determine. The earliest in point of time Arctocyonprimævus, from the lowest Eocene of La Fere, Aisne, France, an animal nearly as large as a wolf, with a long tail, and heavy, strong limb bones, and remarkable for the exceedingly small size of the brain cavity, compared with the arches and ridges of the skull developed for muscular attachments. This character has been supposed to indicate marsupial affinities, but the rest of the osteology, as far as known, does not favour this view. The lower jaw has not been found, but the cranium shows the full complement of teeth so frequent in Eocene mammals. There are three broad tubercular molars behind the trihedral sectorial, which indicate that the animal was rather omnivorous than truly carnivorous in its habits. Another genus which includes many species of various size, and having a wide geographical range, being found in late Eocene and early Miocene deposits in France, Germany, England, and North America, is called Hyænodon. It also has been by many naturalists placed among the marsupials on account of the peculiarities of its dentition, which is certainly without parallel among placenta! Carnivores. It possesses the primitive or typical dental formula of the Eocene mammals, and the incisors, canines, and premolars, are not unlike those of a dog, but the three true molars, both above and below, are shaped like the sectorial teeth of a cat or hyaena, and increase in size from the first to the last, and thus there are no teeth formed like the “tuber-culars” of ordinary Carnivores. This repetition of the sectorial character in the true molars occurs in the carnivorous marsupials, though the general structure of the skull, and limb bones as far as they are known, including the position of the lachrymal canal within the orbit, will not permit our placing Hyænodon in that group. Many of the lately discovered American Eocen e earn ivores presented the same peculiarity of several successive molars having sectorial characters. One of these from Wyoming, apparently allied to Hyænodon, has been described by Cope, under the name of Mesonyx, and another still more aberrant form, as Synoplotherium. The inferior canines project forwards, and are closely approximated, the incisors (at all events in the aged specimen on which the genus wa.s founded) being absent. The molar teeth were so much worn that their characters cannot be satisfactorily made out. The most interesting features of these animals are in the structure of the feet, the ungual phalanges being natter and broader than in any existing Carnivora, and grooved above, and the scaphoid and lunar bones of the carpus not being united as in all existing Carnivores.