BY R. BROOM. 317 



teeth might necessitate the separation of Ccenolestes from the 

 Diprotodonts, and the placing of it and its fossil allies in a 

 (listiiK-t suhoi-flor. 



Gregory, in his most important work on the orders of 

 mammals, accepts Miss Dederer's conclusion, and places 

 ( '(rnolesfes in a separate suborder — the Paucituberculata — 

 equivalent to the suborders Polyprotodoutia and Diproto- 

 doiilia. Ill the diagram he gives of the phylogeny of the 

 marsupials, he derives the Ca^nolestoids quite independently 

 of the Phalangeroids from a generalised Polyprotodont which 

 lived in Upper Jurassic times. 



It is admitted by all that ('iciioJesten resembles the Diproto- 

 dont in only two points, which, as stated by Miss Dederer, 

 are — "(1) Condition of teeth: (a) one large lower incisor, 

 cutting, projecting forward : (b) other incisors and canine 

 in lower j;i\v vestigial as in Epanorthiche; (c) anterior pre- 

 molars small, showing tendency towards condition seen in 

 Phalangers, where they are vestigial. (2) Pattern of teeth 

 - -molars like Phalangcr molars rather than the Polyproto- 

 dont type." 



On the other hand, Miss Dederer gives a list of ten Poly- 

 protodont characters shown by Ccrnolestes, though one of 

 those given, the equality of the fore and hind limbs, is of 

 little importance, and had better be omitted. The remaining 

 Polyprotodont characters are : — 



"(1) Dental formula like that of the Dasyurid genera 

 TJij/hictnits, J'JKixroloijdJc, namely: i.J c.\ pm.i! m. | ; in this 

 family, the incisors are numerous, small, subequal ; canines 

 larger than incisors. This agrees with the condition in the 

 upper jaw of Ccfnolestes." 



"(2) Close resemblance in external form to PJidnrolof/nle — 

 rat-like or opossum-like in form (Thomas)." 



"(3) Resemblance to Dasyxirus skull, {a) in general shape, 

 {li) ptervgoid processes of palatine slender ami delicate, (e) 

 alisphenoid bullae similar in general form." 



