320 ON THK AFFINITIES OF CiENOLESTES. 



dont has more than three incisors; Ccenolesfes has four. In 

 Diprotodonts, the canine, though often present, is always 

 of less importance than the incisors : in Ccenolestes, the upper 

 canine is larger than the incisors. The molars undoubtedly 

 are more like those of the Diprotodonts than the Polyproto- 

 donts : but when we look at the molars of some of the 

 bandicoots, we see that Ccfnolesfes is not so very far removed 

 from known Polyprotodont types : and the last two molars 

 seem still to retain a considerable amount of Polyprotodont 

 character. The lower first incisors are undoubtedly developed 

 to a degree quite unknown in any Polyprotodont, but we find 

 a tendencv to the increase of the first incisors in a number of 

 Polyprotodonts. In Phascologcde , it is so well marked, that 

 this genus might be looked upon as incipiently Diprotodont, 

 the first incisor both above and below being much longer 

 than the other incisors, and nearly as large as tlic i-aiiiiics. 



Miss Dederer, Dr. Gregory, and Sinclair, while agreeing 

 that CcFnolestes should not be placed in the Diprotodontia, 

 prefer to place it in a distinct suborder, the Pancifiihcrcu- 

 lafa. But it has long seemed to me that, as Ccenolestes differs 

 from the typical Polyprotodonts only in tooth-specialisation, 

 it should not be removed from the Polyprotodontia, but 

 merely be made the type of a distinct family, or section at 

 most. If the acquirement of a diprotodont dentition is to 

 lead to an animal s lM'iii<;- placed in a distim-t sulxn-dcr, tlien 

 Chiromi/s must be removed from the Prosimise, and Desmodus 

 from the Microchiroptera. 



