In the Eocene, or Lower Tertiary, still more remarkable forms occur, along with several genera existing at the present day, such as the Cryptoprocta, Civet, Dog, and Marten, all of which are found in the upper or more recent strata of the Eocene formation. But lower down the genus _Cynodon_ also connects Dogs with Civets; and in the very lowest beds occurs a large plantigrade animal (_Arctocyon_), with a very small brain-case, wide jugal arches, a complete set of forty-four teeth, and altogether of a generalised character. In the Eocene of North America, _Limnocyon_ and _Prototomus_ occur low down, and in the Middle Eocene a form as large as a Lion has been discovered, to which the name _Limnofelis_ has been given, and also _Orocyon_, and some allies of the _Hyænodon_. [Illustration: LOWER JAW OF HYÆNODON. (_After Gaudry._)] But we have not yet learned all that Palæontology can teach us about the history of the Carnivora. In the Eocene and Lower Miocene beds are found animals referred to the genera _Hyænodon_, _Pterodon_, _Palæonictis_, and _Proviverra_ which, not content with trespassing on the boundaries between existing families, actually wander outside the Carnivorous order altogether, and approach so nearly to the Marsupials (Kangaroos, Opossums, &c.) that many competent anatomists have proposed to place them in the latter group. The premolars and molars in these extinct animals have sharp cusps, and increase gradually in size from before backwards; so that, of the whole grinding series, the first premolar is the smallest, and the last molar the largest. Now we have seen that the rule among existing Carnivora is for the last molar to be a small tooth, and for the largest of the set to be the fourth premolar in the upper jaw, and the first molar in the lower jaw. On the other hand, the regular increase in size is very characteristic of the flesh-eating Marsupials, amongst which the Thylacine, or so-called Tasmanian Wolf, shows a considerable resemblance, as to its teeth, to _Hyænodon_ and _Pterodon_, while _Palæonictis_ and _Proviverra_ are more nearly allied to the Opossums and to the Dasyure, or Tasmanian Devil. The brain-case in these forms was very small, and a cast of the interior of the skull of Proviverra, figured by M. Gaudry,[198] shows that the brain must have had an extremely low character. [Illustration: SKULL OF PROVIVERRA. (_After Gaudry._) The roof of the skull is supposed to be cut away to show the form of the brain, as deduced from a natural cast of the interior of the skull.] We thus see that a considerable number of the existing genera of Carnivora took their origin in the Eocene epoch, where they co-existed with creatures curiously intermediate between the various existing families, and with others intermediate between Carnivora and Marsupials. In the rocks of the Secondary period (chalk, oolite, lias, &c.), none of the Carnivora have as yet appeared, and only Marsupial remains are found.