538 PROCEEDINGS OF THE NATIONAL MUSEUM vol. no 



except for an ochi-aceous streak on each side of the white underparts 

 in one. 



In five individuals from Canelos taken March 8-12, 1924, the 

 underparts are faintly tinged with buff; in three from the same locality, 

 taken in April 1910, and in two from the Rio Copataza and another 

 from Montalvo, the underparts are pure white. The three localities 

 mentioned are within a few miles of each other in the upper Efo 

 Pastaza region. 



Specimens from Jollln, Jatun Yacu, Llunchi, Avila, Concepci6n, 

 San Jose and Rio Suno, all in the upper Rio Napo region, show the 

 same range of variation in color as the preceding. In one specimen 

 from Llunchi, however, the belly is uniformly ochraceous and hardly 

 defined from the sides. 



The series of 11 specimens from the Rio Curaray, Loreto, Peru, 

 includes individuals in dark and bright pelage. One is nearly 

 uniformly ochraceous orange on upperparts and sides. Underparts 

 are white. One specimen has a pair of small ochraceous gular patches. 



One of four specimens from Lagarto Alto, Rio Ucayali, Peru, in old 

 pelage, is nearly uniformly ocliraceous tawny on head and back. This 

 color is defined in the form of a molt line from the darker outer sides 

 of shoulders and thighs. The other tliree specimens of the series are 

 dark in unworn pelage. Underparts, upper lips and lower half of 

 cheeks in one of two mice from Satipo, Junln, are yellowish. Under- 

 parts of the other specimen are sharply defined white, the lips and 

 cheeks ochraceous. 



The remaining Peruvian material consisting of one or two individuals 

 from scattered localities reveal no peculiarities. 



The type series of Oecomys florenciae J. A. AUen, from eastern 

 Colombia, is practically indistinguishable from typical hicolor as 

 represented by any series from eastern Ecuador. Other Colombian 

 specimens at hand exhibit no individual or local variables of note. 



Judged by the original description, the only distinguishing character 

 of the Orinocoan "Rhipidomys" rosilla Thomas is the "buffy ochra- 

 ceous" belly of the type as compared with the pure white belly of the 

 type of dry as {=occidental{s) of western Ecuador. A colored belly, 

 however, is a common enough character in hicolor of eastern Ecuador. 

 In the absence of other distinctions, rosilla must be regarded as 

 identical with typical hicolor. 



Oecomys nitedulus Thomas of British Guiana is said to be "quite like 

 the Orinoco Oe. rosilla above, but the undersurface is without the 

 marked ochraceous wash characteristic of that animal." This 

 characterization implies that nitedulus is like typical hicolor. Indeed, 

 our material from British Guiana is altogether indistinguishable from 

 eastern Ecuadorian hicolor. I have not examined the type of Oecomys 



