668 Bulletin American Museum of Natural History. [Vol. XXIV, 



size and form, but the skull in the two species is markedly different in con- 

 tour and in details, as is well shown in the accompanying figures of the type 

 skulls of each (Figs. 5-12). The anterior base of the zygoma is much 

 lower in B. gabbi than in B. richardsoni and less arched upward; the post- 

 palatal region is shorter and broader, as is also the postpalatal fossa and 

 the pterygoid processes; the bullse are shorter and more inflated; the ros- 

 trum descends much more abruptly and the anterior jiarial opening is 

 lower and more oblique; the coronoid process is narrower, shorter, and 

 directed less backward. 



As now known, Bassaricyon is represented by three well marked species, 

 inhabiting respectively the eastern lowlands of Nicaragua, the lowlands of 

 southeastern Costa Rica and adjoining parts of Panama, and the vicinity 

 of Sarayacu, Ecuador, east of the Andes and well within the upper drainage 

 of the Amazon. If the British Guiana record for B. alleni be correct, the 

 group may have a wide range in South America; in which case it seems 

 strange that it has been so long overlooked, and that so few South American 

 examples of it have thus far been obtained. As long since made known 

 by Thomas, Huet, and Beddard, Bassaricyon bears a close resemblance 

 externally to the kinkajou, although much smaller and with a non-prehensile 

 tail. It may thus readily be mistaken by collectors and travellers for a 

 young kinkajou. 



48. Nasua narica bullata Allen. Three specimens, males, Lavala, 

 Oct. 12, Tuma, Nov. 29, 1907, and Ocotal, May 7, 1908. 



49. Urocyon cinereoargenteus guatemalae Miller. One specimen, 

 adult male, Matagalpa, April 20, 1906. 



This is one of the smallest specimens of Urocyon that has thus far come 

 under my notice, or of which measurements have been published. It is a 

 male, fully adult, as shown by the skull. No external measurements were 

 taken by the collector, and the occipital part of the skull, including most 

 of the braincase, is lacking, rendering the skull measurements incomplete. 

 The palatal length, zygomatic and interorbital breadth, and the breadth 

 across the postorbital processes are practically the same as in the type of 

 Urocyon parvidens Miller, from Merida, Yucatan, while the measurements 

 of the teeth correspond exactly with those of parvidens. It is much smaller 

 than a female of U. guatemalce Miller (type locality, Nenton, Guatemala), 

 of corresponding age, from Pozo Azul, Costa Rica. 



While the present specimen may represent a dwarfed form peculiar to 

 the interior of Nicaragua, it seems better to refer it to its nearest geographical 

 ally rather than to the more distant Yucatan form (parvidens), or than to 

 make it the basis of a new name. 



