E. LÖNNBERG, MAMMALOGY OF ECUADOR. 39 



Cararay with this description there is especially one charac- 

 teristic which is found to be entirely absent. Waterhouse 

 says: »on the upper parts of tlie animal the hairs are black 

 towards the point, yellowish white at the point, and with 

 the hidden parts of a deep rust colour.» — The last remark 

 cannot be due to »the progressive deterioration and bleaching 

 of the type specimen» (Thomas 1. c). — In the specimen 

 from Rio Curaray the hairs of the upper parts are black 

 below the whitish tips to the basal parts which are dark 

 grey. There is not the slightest trace of any »rust colour» 

 of the hidden parts, neither on the back, nor on the flanks, 

 where the basal fur only becomes somewhat paler grey. This 

 appears the more important as even in Thomas' new sub- 

 species D. d. canescens the »rusty colour of underfur» is 

 »strongly marked along the middle dorsal area». The spe- 

 cimen from Rio Curaray thus differs in a very conspicuous 

 manner from the latter as well. The lower flanks of this 

 specimen are duU greyish buff, but the back of the hind legs 

 »rust red» (Oberthur: 318, 1). The head and lower parts 

 agree with Waterhouse's description. 



The specimen from Rio Curaray appears also to be a 

 smaller animal (head and body about 28 cm, tail about 34 

 cm. Its cranial dimensions are considerably smaller than 

 those recorded by Thomas for D. d. canescens. 



Condylobasal length 69 {D. d. c. 73 ) mm 



Condyloincisive » G6 ( 



Zygomatic breadth 35,3 ( 



Upper tooth-series 18,6 ( 



» » 70 ) 

 > » 38,5) 

 » >' 21,ö) 



Jentink^ has unfortunately not communicated any mea- 

 surements of the skull of his Dactylonys, but he has published 

 some very fine figures in nat. size. From one of these can 

 be seen that the occipitonasal length of his specimen is about 

 79,5 mm, while the corresponding dimension of the present 

 specimen is only 71,5 mm. On the same figure may be seen 

 that the nasals with their posterior ends reach further back 

 than the frontal processes of the premaxillary which is quite 

 the opposite to the condition of the present specimen. Jen- 

 TiNK says (1. c): »The molars of Daclylomys daciylinus are 



' Notes Leyden Mus. Vol. XIII, 1801. 



