214 Bulletin American Museum of Natural History. [Vol. XXVII, 



donts the general resemblances between Propolymastodon and Polyviastodon 

 are possibly convergent; (4) that in fact the resemblances between South 

 American and other groups are in many cases the result of convergence, and 

 not indicative of close affinity; (5) that if Polymastodon is assumed to be a 

 INIultituberculate we are led into seemingly contradictory hypotheses of the 

 derivation and relationships of Multituberculates, Cjenolestoids and true 

 Diprotodonts. 



The most reasonable provisional conclusions from these very puzzling 

 facts seem to be the following : 



(1) Propolymastodon is a "pseudo-Multituberculate" offshoot of the 

 Csenolestoids. The latter (p. 211), seem to be derived from the Polyproto- 

 donts, or at least from Mesozoic forerunners of the same. 



(2) Propolymastodon and the series of Csenolestoids serves to illustrate 

 in what manner the true ]\lultitu1)erculate lower molar miglit have been 

 derived from a tuberculo-sectorial lower molar and vice versa: but they do not 

 by any means prove that either type was derived from the other (see p. 168). 



(3) The known forerunners of the Csenolestoidea in the Xotostylops 

 Beds represented an early specialized offshoot and are not structurally 

 ancestral to the later Csenolestoids. 



At the same time the contrary hypothesis should not be forgotten, namely : 



(1) That the Notostylops Beds genera Polydolops, Propolymastodon, 

 etc., are survivors of a group of Triassic Diprotodonts which gave rise to 

 the Multituberculates on the one hand and to the Csenolestoids and Austra- 

 lian Diprotodonts on the other. 



(2) That the resemblances of both the Diprotodonts and Csenolestoids 

 to the Polyprotodonts are due partly to convergence and partly to inheritance 

 of primitive characters from a remote Triassic Marsupial stock. 



A clearer knowledge of the genetic relations of these Patagonian forms 

 is of the utmost importance in the study of Marsupial phylogeny. The wide 

 difference between the theoretical relations of the Csenolestoids to other 

 Marsupials and the imperfect indications of the geological record are revealed 

 in the tables on pages 228, 229. 



Wynyardia bassiana Spencer. 



This species (Spencer, 1900), which is from the Eocene (or Oligocene) 

 of Table Cape Tasmania, is another form that helps to bridge over the 

 structural gap between the Polyprotodontia and the Diprotodontia. The 

 skull (which unfortunately lacks the dentition) approaches that of the 

 Dasyuridse in many features of the cranium proper and zygoma, but the 

 premaxillaries are stated to approximate in form to those of Trichosurus 



