Réévaluation des Peramura (Mammalia, Cladotheria) sur la base de nouveaux spécimens du Crétacé inférieur d'Angleterre et du Maroc

L\u27analyse de dents mammaliennes nouvellement extraites du gisement britannique de Durlston Bay et du gisement marocain du synclinal d\u27Anoual (Crétacé inférieur) nous a conduit à identifier trois nouveaux taxons Minimus richardfoxi n.g., n.sp., Magnimus ensomi n.g., n.sp., Afriquiamus nessovi n.g., n.sp.) et à réexaminer la question (validité, définition, contenu) des Peramura McKenna, 1975. Ce groupement de mammifères thériens, dont les dents établissent une liaison morphologique entre celles des symmétrodontes primitifs et celles des mammifères tribosphéniques, ne présente en réalité, dans l\u27état actuel des connaissances au moins, aucune synapomorphie exclusive ; il s\u27agit plutôt de quelques représentants d\u27un grade évolutif dont la vaste répartition géographique et stratigraphique laisse supposer une importante diversification, ainsi qu\u27une origine précoce pour la lignée prétribosphénique.The analysis of several mammalian teeth exhibiting peramurid characters and recently obtained from the Early Cretaceous localities of Durlston Bay (England) and Anoual (Morocco), led to the identification of three new taxa (Minimus richardfoxi n.g., n.sp., Magnimus ensomi n.g., n.sp. and Afriquiamus nessovi n.g., n.sp.) and to the reexamination of Peramura McKenna, 1975 (definition, contents, validity). It appears that this group of therian mammals, whose molars establish a morphological link between those of primitive Symmetrodonta and those of tribosphenic mammals, does not in fact show any exclusive synapomorphy, at least in our present state of knowledge; we are confronted, rather, with a few representatives of an evolutionary grade whose vast distribution in space and time suggests an even wider diversification, as well as an early origin for the pretribosphenid line. Moreover it is shown that Amphitherium is closer to tribosphenids than to dryolestoids, while Vincelestes does not qualify as a pretribosphenic mammal. We also discuss the question of the metacone, the stylocone and of the lingual cingulum on the upper molars of Peramus, to conclude that the ancestors of this genus did not go through a dryolestoid stage: the metacone is considered homologous to that of tinodontids, so that its lingual situation is in fact a primitive character. Finally it is concluded that it was the individualization of a hypoconid and a distal metacristid on the lower molars, hence the introduction of a different masticatory mode, that created, early in the pretribosphenid line, the conditions favorable to the later elaboration of a protocone on the upper molars; but the primitive stage of this cusp remains unknown.