South-American Mammals. 235 



the metacarpus, the sides and the fingers white; feet white, a 

 bruwiii.sh patch on the metatarsus. Tail-hairs wholly brown 

 above and below, scarcely lengthened at tip. 



Skull with the usual rounded shape found in QScomys^ but 

 the suj)raorbital ridges are unusually developed and pass 

 back as strongly marked ridges across the |)arietals; nasals 

 narrow. Antcorbital plate of zygoma more developed than 

 is usual in CEcomi/s, more as in Uryzomys, projecting about 

 1^ mm. in front of the upper bridge. Palatal foramina short, 

 broad, widely open. Mulars strong, quadrangular. 



Dinicnsions of the type (measured in spirit) : — 



Head and body 132 mm. ; tail IGtl ; hind foot 30; ear 20. 



JSkull: greatest length 3-4; basilar length 26*5 ; zygomatic 

 breadth 17; nasals 11; intcrorbital breadth 6*2; breadth 

 across parietal ridges 12"7 ; palatilar length 15; palatal 

 foramina 6x31; upper molar series 5'2. 



JIab. Joinville, Santa Catherina, S. Brazil. 



Type. Adult male. Original number 30. Collected by 

 W. Ehrhardt. 



This is a most interesting species on two accounts. Geo- 

 graphically it t'orms a great extension of the known range of 

 the group, no bush-rats, (Ecomys or Rhipidromys, being 

 known from South Brazil and no CEcomys nearer than Matto 

 Gro.-Jso, unless the peculiar ^^ Rhipidomys''^ rufescens of Rio 

 Janeiro should prove to be a member ot" this genus. 



Then in its structure (27. ctitkerince is abnormal in that its 

 zygomatic plate is as developed as in ordinary Oryzomy^, 

 thus showing that this character is not so constant as I had 

 supposed it to be. The external characters are distinctly 

 those of the bush-living (Ecomys, notably the short broad 

 feet, with proportionally long hallux and fitth toe, and the 

 well-haired unicolor tail. The tail in all true Oryzomys, 

 owing to their living on the ground, is white or whitish 

 underneath, at least proximally, while in all the bush-rats it 

 is wholly dark above and below. 



2/ie Groups referred to the Genus Oxyraycterus. 



Ever since I described, in 181)7*, the aberrant Oxy- 

 mycttrus laiiosus and Dr. Allen named the still more aberrant 

 O. aj'icalts t it has been obvious that tho genus ought to be 

 divided into three, corresponding respectively to the typical 

 Oxymycteri and to the two species above mentioned. 



Considering how different the animals are externally, it is 



• Ann. & Mrj?. Nat. Hist. (0) xx. p. 218. 

 t Rull. .Vni. MiiP. N. H. xiii. y. l'24 (IfXX)). 



