1910.] Propolymastodon and iU Allies. 213 



many orders seem to abound in examples of homoplastic and convergent 

 resemblances to animals of other orders. The resemblances between 

 Propoljimadodon and Polymasfodon are perhaps not much closer than the 

 resemblances of Borhi/a;na to Oxycena, of Nesodon and Astrapotherium to 

 Metamynodoyi and Cadurcotherium, of Thoatherium to Mesohippus, of 

 Protypotherium and Arckceohyrax to Hyrax, etc.; and yet in each of the 

 cases named the resemblances are very probably due either to convergence 

 or to homoplasy. Among the Csenolestoid relatives of Propolymastodon 

 one genus, Orthodolops (Ameghino, /. c, p. 131) somewhat resembles Sciurus 

 in both upper and lower teeth, while Cephalomys, a Patagonian rodent, has 

 a deciduous p^ suggesting that of the Multituljerculate Ptilodus (Ameghino 

 /. c, p. 98); and yet Orthodolops can hardly be regarded as a relative of 

 Sciurus, nor Cephalomys of Ptilodus. 



(3) The extreme plasticity of the cheek teeth in the Multitubercidata, 

 Caenolestoidea and Diprotodontia is well illustrated in the great range in 

 form of the ultimate upper premolars and first lower molar. In Garzonia 

 and Ccenolestes mj is simple, in Ahderitrs it is highly grooved, while in Pro- 

 polymastodon it is intermediate. In Bettongia it is the posterior premolar 

 which is highly grooved, while in the more advanced Marropus the same 

 tooth is simple. In Ptilodus the anterior cheek tooth is highly grooved, in 

 the related Polymastodon it is reduced and simple. 



(4) Assuming that the early Tertiary Propolymastodon is related to the 

 Csenolestoids and also structurally ancestral to the Upper Cretaceous and 

 Basal Eocene Polymastodontid.ie, how are we to account for the ancestry 

 of the Upper Triassic Multituberculates Tritylodon, Triglyphus and Micro- 

 lestes'! Is the order Multituberculata diphyletic ? If so, where is the break 

 in the fairly close morphological sequence represented in the genera Micro- 

 lestes, Plagiaulax, Ptilodus, Meniscoessus and Polymastodon'? 



(5) The reduction of the lower lateral incisors, canines and anterior 

 premolars in Cfenolestoids does not favor the hypothesis that the tuberculo- 

 sectorial dentition of the Polyprotodont Microbiotherium might be derived 

 ultimately from the "multituberculate" type of Propolymastodon; because 

 the Csenolestoid antemolar dentition, even if it had appeared early enough 

 to give rise to that of the Triassic Multituberculates, could hardly have given 

 rise to the antemolar dentition of Amphitherium and the Polyprotodonts. 



In brief, the evidence for inferring that Propolymastodon and its allies 

 are not Multituberculates but highly modified Cfenolestoids is: (1) that they 

 seem to be connected with the Csenolestoids by a series of structurally inter- 

 mediate forms; (2) that Propolymastodon and its allies differ from the 

 true Multituberculates in certain important particulars; (3) that in view of 

 the great plasticity of the cheek teeth in the Multituberculates and Diproto- 



