102 CARNIVORA. 
Viverra angustidens, Filhol’. 
Hab. France. 
M. 1358. Anterior portion of the left ramus of the mandible, con- 
taining the canine and the four premolars, the crown of 
pm. 1 being broken off; from the Upper Kocene of Caylux 
(Tarn-et-Garonne), France. This specimen agrees with the 
type mandible figured by Filhol in the Ann. Sci. Geéol. 
vol. vii. pl. xxvi. figs. 121—2, in the extremely minute size of 
pm.1; pm. 2 is, however, rather smaller in the present 
specimen than in the type, and has apparently but one root. 
Purchased, 1884. 
Genus AMPHICTIS, Pomel’. 
This genus is mainly distinguished from existing species of Viverra 
by the form of the second lower true molar, which is elongated and 
implanted by two distinct roots. 
Amphictis antiqua, Pomel’. 
The mandible of this species was originally referred by De Blain- 
ville to his Viverra antiqua (Herpestes antiquus, Pomel), and a simi- 
lar specimen is figured by P. Gervais in the Zool. et Pal. Frangaises, 
2nd ed. pl. xxviil. figs. 8, 8a, under the same name, although a 
doubt is expressed in the text as to the correctness of the generic 
reference. It has beeen shown by Filhol* that there is a very con- 
siderable probability that the mandibles which he has figured under 
the name of Herpestes antiquus ” are really those of Viverra antiqua. 
The length of the space occupied by the six cheek-teeth in those 
specimens is 0,041. 
Hab. France. 
31057. Fragment of the right ramus of the mandible, containing 
the last three premolars, the anterior half of m.1, and the 
alveoli of the canine and pm. 1; from the Lower Miocene of 
St. Gérand-le-Puy (Allier), France. This specimen appears 
to agree exactly with the mandible figured by P. Gervais in 
the Zool. et Pal. Francaises, 2nd ed. pl. xxviii. figs. 8, 8a, 
under the name of Viverra antiqua, which is very similar 
1 Ann. Sci. Géol. vol. vii. art. 7, p. 144 (1876). 
2 Catalogue Méthodique, p. 63 (1858). 
3 [bid. 
4 Ann. Sci. Géol. vol. x. art. 3, p. 166. 
> The description in the text is headed Herpestes priscus, but in the Arch. 
Mus. Lyon, vol. iii. art. i. p. 63, this is corrected. 

