306 ORIGIN AND DISTRIBUTION OF MAMMALS 



Allotheria and partly from the Didelphys-like types of the Upper Cretaceous 

 Laramie beds. The rodents, which appear later, hoAvever, have another origin. 

 Their ancestors lived in Europe and JSTortli Africa. They probably emigrated 

 passivelj^ having been carried from island to island by birds of prey ; they 

 therefore did not need a continuous bridge of land — a row of islands between 

 West Africa and Brazil was sufficient. 



The oldest deposits in South America from which mammals have been 

 obtained in noteworthy numbers are the Notostylops beds of Casa Mayor, in 

 the Gulf of St. George, Patagonia, which appear to be of about Upper Eocene 

 age. Except the rodents, almost all the types of the South American fossil 

 mammal world are already represented here, though at first by small or 

 medium-sized forms always with brachyodont molar teeth, the Typotheria by 

 the Notojnthecidae, and perhaps also by the somewhat more distantly related 

 Archaeopithecidae and Acoelodidae, the Entelonychia by the Notostylopidae, Isotem- 

 nidae, and Homalodontotheriidae — Thomaslmxleya, the A strapotherioidea by the 

 Trigonostylopidae and Albertogaudryidae, and the Litoptcrna by the bunodont 

 genus Didolodus and a number of other very imperfectly known forms. There 

 are also already representatives of the very problematical Pyrotheria, of which 

 Carolozittelia seems almost to suggest some relationship with Pantolamhda. The 

 Polydolopidae, which belong to the marsupials, remind us of the Allotheria ; and 

 the predaceous marsupials are represented by forms, such as Procladosidis, 

 Pharsophorus and the Didelphys-\\ke genus Ideodidelphys, Avith which may best 

 be associated the peculiar bunodont Carol oameghiniidae. Edentata are by no 

 means lacking, but most of the very problematical genera are already closely 

 related to the existing DasyjMdidae, and the single gravigrade genus, Proto- 

 hradys, is only very incompletely knoAvn. 



The next younger mammal faima of Patagonia, that of the Oligocene 

 Astraponotm beds, ma}" be best described as the continuation of the Notostylops 

 fauna, but it is certainly much poorer in genera and species, and is hitherto 

 only very imperfectl}' knoAvn. As neAv types, there appear only the Noto- 

 hippidae, of Avhich the tooth-croAvns are for the first time considerably hypsodont, 

 the Archaeohyraddae Avith prismatic teeth, the Astrapiotheriidae — Astraponotus — 

 and among edentates the first Glyptodontia. 



The succeeding fauna, that of the Pyrofhermm beds south of Cape Blanco, 

 shows much advance on the tAvo preceding faunas, and has so many genera in 

 common Avith the best known and richest fauna of Patagonia, the Santa 

 Cruzian, that there cannot be an}' doubt about its Miocene age. The prob- 

 lematical Pyrotheria, Avhich occur here for the last time, now reach the highest 

 point of their development in the ponderous species of the genus Pyrotherium. 

 The Typotheria and To.wdontia uoav all have prismatic teeth. The Notohippidae 

 are also especiall}' numerous here, and the Nesodontidae appear for the first 

 time, even a large JVesodon, besides Interatheriidac, Pla giarthrus. Among 

 Entelonychia the Isotemnidae are indeed extinct, but the Leontiniidae already 

 attain great size of body — Leontinia, Ancylocoelus — likeAA'ise the Astrapotheriidae, 

 and even the genus Asmodeiis among the Homalodontotheriidae. The Lifopterna 

 are now specialised as Proterotheriidae and Macrnucheniidac, but the latter are 

 represented only by rather small brachyodont forms. Among the Edentata 

 the Doiiypoda are the direct successors of the forms in the Notostylops beds, 

 Avhile the Glyptodontia and Gravigrada are scarcely distinguishable from the 

 later forms, and the same is the case AA^ith the Dasyuridae. On the other hand, 



