386 On new Rodents from Western South America. 



In another specimen from tlie same locality the whole 

 body is dull blackish rufous, nearly as rufous on the sides of 

 the rump as on the shoulders. 



Skull not appreciably different from that of S. vartabtli's 

 from Santa Marta. Premolars ]. 



Dimensions of the type (measured in the flesh) : — 



Head and body 230 millim. ; tail 225 ; hind foot (s. u.) 58, 

 (c. u.) 63 ; ear 27. 



Skull : greatest length 57 ; basilar length 43*3 ; greatest 

 breadth 33*5; nasals (diagonally) 18 x 8*8; palate length 25; 

 length of upper tooth-row 9*5. 



JJab. (of type). Cachabi, Prov. Esmeraldas, N. Ecuador. 

 Alt. 160 m. 



Type. Male. B.M. no. 97. 11. 7. 32. Original number 25. 

 Collected 3rd Jan., 1897, by Mr. W. R. Rosenberg. 



The British Museum contains a large number of squirrels 

 from the north-western corner of South America which have 

 been referred to S. variabilis, Geoff. ; but now that Mr. Bangs 

 has shown that the true S. variabilis is constantly white- 

 bellied (as is also the type of Gray's S. Gerrardi), it is 

 evident that the red-bellied forms require a special name. 



As is the case with S. variabilis, this species is a very 

 variable one, and each of the several series of it in the Museum 

 presents some striking variation among its members. The 

 general tone varies in depth and degree of grizzling : the 

 centre of the back varies from deep shining black to hardly 

 darker than the sides, the ear-patches are occasionally 

 suppressed, the middle part of the tail may be either yellow, 

 orange, or red, the end of the tail may be either red or black 

 (more frequently the latter), and even the distinctive red 

 belly may be picked out with white markings in the axillee, 

 middle line, and groins. It is possible therefore that other 

 variations will lead into S. variabilis, in which case it will 

 have to be considered as a subspecies of that form. 



Other localities from which the Museum has specimens of 

 S. versicolor are — in Ecuador, Paramba [Rosenberg, Miketta) ; 

 in Colombia — Valdivia, Lower Cauca {Pi-att), Medellin 

 {Salmon), San Pablo [Ilopke), and Bogota [Child). 



A rather similar coloration to that of S. versicolor is found 

 in the true S. griseogena, Gray, of Venezuela ; but that is a 

 smaller form, as shown by its skull, and has a more uniformly 

 grizzled back, less blackened behind and less reddened in 

 front. 



Mr. Nelson's Sciurus caucensis is not a member of this 

 group, but is allied to — indeed, I do not see how it is diffe- 

 rent from — Gray's S. medellinensis, also from the Valley of 



