EVOLUTION OF THE EARLY TERTIARY RODENTS

Abstract

Representatives of the order Rodentia first appear in the late Paleocene (Tiffanian) deposits of the Fort Union Series. The known material is, however, exceedingly fragmentary, for only a single lower molar tooth and some incisors have been found. The specimens have been referred to the genus Paramys (sensu lato). They demonstrate that the characteristic rodent incisor and molar were already developed at the time. These teeth suggest that most of the other characteristic rodent features were also present, although it is, of course, possible that some primitive features absent in other rodents were still retained. Paramys (sensu lato) is relatively well known from the early Eocene (early Wasatchian) deposits of the Willwood formation, and similar rodents are known from the early Eocene of Europe. In their comparable parts, these later rodents do not differ in any important respect from the late Paleocene species. All rodents known from the late Paleocene and early Eocene pertain to the superfamily Aplodontoidea and to the subfamily Paramyinae (or family Paramyidae of other authors), the most primitive group of rodents. These early paramyines are nearly structurally antecedent to all later rodents, if indeed they are not their actual ancestors. Primitive as they are, they already have acquired the formally diagnostic characters of the order insofar as skeletal structures are concerned; and these characters, fully developed, show no transition to those of the Insectivora (sensu stricto), or to any known primitive placental. Several early insectivore or primate groups (mixodectids, apatemyids, Phenacolemur) exhibit some degree of convergence in the incisors and lower jaws toward a rodent type of structure. Osborn years ago placed the mixodectids in a distinct suborder of ancestral rodents, the Proglires (Osborn, 1902, p. 203). Further detailed examination has failed to reveal any evidence for such a relationship. Structures common to the mixodectids (presumptive insectivores) and ro'dents are shown also in several other groups which have no special relationship to the Rodentia. The apatemyids and Phenacolemur are regarded currently as probable primates. It seems that in the early Cenozoic several placental groups had similar adaptations of the front teeth with attendant, correlated modifications of the jaws. Perhaps these groups indicate a series of competing forms which finally passed to extinction as the rodents slowly rose to ascendancy. Many workers have accepted the conclusions of W. D. Matthew (1910), and have followed his thesis that the paramyines are actually ancestral to all later rodents, and that rodent diversification in the higher categories has been a postPaleocene event. Others, particularly Miller and Gidley (1918), have at least implied that the known stratigraphic record is faulty, and that all major groups of rodents (approximately the superfamilies of the Miller and Gidley classification of 1918) were in existence by early Eocene time. Advocates of Miller and Gidley's view perhaps tend to think of the origin of the order as occurring at an earlier time than do those advocating an Eocene diversification. According to various suggestions, the time of origin may range from far back in the Cretaceous to early Paleocene. If we accept one of these dates, particularly a Cretaceous one, and agree