3 8 Bulletin American Museum of Natural History. [Vol. XX, 



across the points of the crescent to the point of greatest convexity, 

 the greatest width of the light band being 55 mm. 



This feature is of interest in connection with Mr. Thomas's 

 remarks (Ann. and Mag. Nat. Hist. (7), V, Jan., 1900, pp. 147, 

 148) in reference to three specimens from widely separated 

 localities showing this "spasmodic variation," found by him 

 in the British Museum collection. In the present subspecies 

 it amounts to almost as constant a character as the gular 

 patch. It is not present, however, in a single specimen in 

 an equal number of examples in this Museum from other 

 localities. 



Tayra barbara irara differs from typical barbara, from Ven- 

 ezuela and Brazil, in being very much smaller, apparently 

 scarcely exceeding Tayra barbara trinitatis (Thomas) from the 

 island of Trinidad, and dark chocolate brown instead of deep 

 black. Two old males from Suapure, Venezuela, and other 

 examples from Brazil, are intensely black throughout, except 

 for the head and neck, and greatly exceed the Santa Marta 

 specimens in size, the occipito-nasal length of the skull ranging 

 from 118 to 126 mm., the basal length from 115 to 118, the 

 zygomatic breadth from 70 to 79, and the upper toothrow 

 from 24 to 26, against, respectively, in, 106.9, 69, an( ^ 

 21.5 in irara. 



Various names have been applied to the South American 

 Tayras, but, as Mr. Thomas has noted (Ann. and Mag. Nat. 

 Hist. (7), VII, Feb., 1901, p. 180), most of them are strict syno- 

 nyms of Mustela barbara Linn. (1758), from " Brasilia." Ap- 

 parently Linnaeus described the animal from a specimen. 

 He cites 'Ac. Holmens.', and Brown's 'History of Jamaica,' 

 the last with a query. The former I am unable to consult. 

 Brown's "Galera . . . The Guinea Fox," is based on an 

 animal "often brought to Jamaica from the coasts of Guinea, 

 where it is a native." In the i2th edition of the ' Syst. Nat.' 

 only Brown is cited. As "Brasilia" is probably used in a 

 general sense, it seems proper to consider the type region of 

 Tayra barbara as Guiana. Mustela gulina Schinz (1821), and 

 Gulo canescens Lichtenstein (1823, ex Illiger MS.) are merely 

 new names for Mustela barbara Linn. ; Vivera [sic] poliocephalus 



