lOlO.] Notoryctcs, Cmiolestes. 209 



about whose affinities some doubt has been expressed. Being very highly 

 specialized for a mode of life like that of the Cape Golden Moles (Chryso- 

 chloridffi), Notoryctcs parallels that family in so many ways that Cope (1892) 

 was inclined to believe that real genetic affinity between the two groups 

 was indicated. Even the peculiar narrow triangular form of the molars 

 in ChriisochJoris has its parallel in Notorijctes (Bensley, 1903, pp. 117-123). 

 But it is shown below (p. 255) that Chrijsochloris is a true Insectivore while 

 Notorijctes is a true Marsupial in the proximal expansion of the nasals, in 

 the possession of vestigial epii)ubic bones, of a pouch opening backward, 

 and in many other characters (Stirling, 1891). 



The pes of Notoryctcs has the appearance of retaining the last vestiges 

 of former syndactyly (Dollo, 1S99; Bensley, 1903, p. 172, fig. 6 C). It 

 seems likely that the nearer affinity of Notoryctcs is with the Australian 

 Polyprotodonts, especially the Peramelidte, (Dollo, Bensley) rather than 

 with the Didelphiidpe. 



III. The C^nolestoidea, or Paucituberculata. (Epanorthid^, 

 Abderitid^, Garzoniid^, C^nolestid^, etc.) 



Santa Cruz and recent Ccenolestokls. 



From the Santa Cruz Formation (? Miocene) of Patagonia Ameghino 

 has described, under many generic names, a group of small Diprotodont 

 marsupials that is represented in the existing fauna of Bolivia and Ecuador 

 by the very rare and important form, Coenolcstes obscurus Thomas (1895). 

 This rat-like animal exhibits the Diprotodont dentition in an interesting 

 and rather primitive stage (Fig. 15). The hypertrophy of the median pair 

 of lower incisors, the reduction of the remaining incisors and canines, and 

 the lateral and extended position of the upper and lower incisor series are 

 all conditions which are closely paralleled in different lines of marsupial and 

 placental insectivores ; as remarked by Bensley, these characters indicate 

 that the Ceenolestidee and all the higher Diprotodonts have been derived 

 from small insectivorous forms. The first and second upper molars of 

 Ccenolestes are subc[uadrate with four low rounded cusps, but the third 

 molar is triangular and has the appearance of being tritubercular. In 

 Palcpothentes (Epanorthus) intermedius of the Santa Cruz Beds (Sinclair, 

 1901, pi. Ixiii, Fig. 7) the first upper true molar is usually quadrate, while 

 the second to fourth are triangular; in P. viiyiutus even m'^ is roughlv tri- 

 angular and the remaining molars are clearly a bunodont, omnivorous 

 modification of the tritubercular type; this in turn may have been derived 



