The placental group is first known from primitive insectivorous representatives from Mongolia, as already noted ; apparently it originated somewhere in the north and by basal Eocene times its diversified descendants are found in western North America and western Europe but as yet nowhere else. These were small brained forms, many belonging to families that became extinct before the close of the Eocene. Here belong most of the archaic carnivores or creodonts, amblypods, tillodonts and others. Be sides the "archaic" placental families that became wholly extinct were others apparently related to later groups. Here belong (I) the Plesiadapidae (apparently related to the tree-shrews and thus representing an early phase in evolution of the primates) ; (2) early forerunners of the tarsioid primates; (3) the insectivore Palaeoryctes, representing a primitive phase of the modern cente toid insectivores; (4) the progressive creodont Didymictis (in or near to the ancestry of the modern carnivores) ; (5) the palaeano donts, including primitive relatives of the armadillo group of edentates. The bats (Chiroptera, q.v.), too, appear to be an old group, possibly dating back to the Basal Eocene or even earlier. Conspicuously absent from the Basal Eocene are the modernized or caeneutherian placental orders, including especially the perissodactyls, the artiodactyls, the rodents, the Probos cidea.
The condylarths or primitive ungulates of the Basal and Lower Eocene are not regarded by modern authorities as ancestral either to the perissodactyls or artiodactyls; it is however probable that some of the condylarths gave rise to the peculiar South Ameri can orders of litopterns, toxodonts, etc. By the Lower and Middle Eocene we find in Europe and western North America, side by side with the later families of meseutherian placentals, the earliest known representatives of the more progressive modern families : Eohippus, representing the horses (Equidae) ; Hyrachyus and other genera, forerunners of the lophiodonts and rhinoceroses; Trigonolestes, a primitive representative of the artiodactyls; Metachiromys, a specialized armadillo-like edentate; Viveravus, close to the ancestors of the later carnivores ; Paramys, a fore runner of the sciuroid rodents; Pelycodus and Notharctus, early lemuroids, and the numerous anaptomorphids, relatives of the primate Tarsius. Even the Cetacea were represented by the pe
culiar zeuglodont group. During the Lower Eocene, western Europe and North America had several fossil genera in common (including the earliest stages of the horse family) which may perhaps have spread from central Asia.
In many cases the molar teeth of the Eocene placental mam mals, while variously specialized, still retained distinct traces of derivation from an earlier tritubercular type characteristic of the basal Eocene; the hands and feet were either five-toed or bore clear traces of derivation by reduction from the five-toed type; the humerus had an entepicondylar foramen and the femur a third trochanter. So that it seems reasonable to infer that all the varied placental orders of Eocene times were descendants of some group that lived perhaps in the Lower Cretaceous; but as to the more precise interrelationships of the mammalian orders, extended research on the anatomy and osteology of the recent forms and on the osteology and dentition of the fossil forms has so far rather revealed the complexity of the problem than solved it. From present evidence the tree-shrews, lemuroids and primates, with Galeopithecus and the bats, appear to form one great superordinal assemblage, probably derived from arboreal Cretaceous insec tivores. On the other hand, the old order of ungulates, formerly supposed to be a natural group, seems to consist of a hetero geneous assemblage of orders (condylarths, taligrades, ambly pods, perissodactyls, hyracoids, proboscideans, sirenians, artio dactyls, notoungulates, arsinoitheres, etc.), related chiefly by descent from various as yet unknown protoungulates.
The aardvarks, formerly classed with the edentates, may prove to be highly specialized descendants of the condylarths, while the edentates themselves are probably the descendants of the palae anodonts, which in turn may be remotely related to the insec tivore-creodont stock. The earliest known cetaceans, the zeuglo donts, were not directly ancestral to the true cetaceans; never theless they tie in that order with the insectivore-creodont divi sion of the placentals. On the other hand, the Sirenia, although resembling the cetaceans in general body-form, are herbivores and their real affinities are with the ungulates.