W. Davies — Neiv British Eocene Carnivores. 435 



The left ramus of the lower jaw is nearly entu-e and in situ, 

 the teeth being in contact with those of the opposing maxilla, 

 the condyle is just within the glenoid cavity of the squamosal. 

 The length is 34. 



The detached portion of the right ramus contains the third and 

 fourth premolars, and the first and second true molars. Premolar 3 

 is smaller relatively to premolar 4 than is the corresponding tooth 

 in Herpestes, the base is broad in the direction of the long axis of 

 the jaw, the crown is low, and there is a well-defined posterior 

 tubercle ; premolar 4 resembles the same tooth in Herpesles. We 

 have, however, in the first molar (the carnassial) another type of 

 tooth, distinct from Herpestes, but like the corresponding tooth in 

 Cynodictis. The posterior lobe of the blade is acutely conical and 

 high, the anterior lobe being relatively low, barely half the height 

 of the posterior lobe, the inner cusp is also high and stout. In 

 Herpestes the two lobes of the blade do not differ greatly in height 

 and the inner cusp is relatively smaller, and less elevated than 

 the anterior lobe of the blade. The talon has a central conical 

 tubercle, and there is a small cingulum on the outer anterior base 

 of the blade. The tooth is short antero-posteriorly, the length not 

 exceeding the vertical height of the posterior blade. The second 

 molar is proportionally small, it has three tubercular points and a 

 short posterior talon, which is not shown in m. 2 in Herpestes, but is 

 present in Cynodictis. The dimensions of the mandibular teeth are — 



Antero -posterior diameter of premolar 3 0'27 



Vertical diameter O'lo 



Antero -posterior „ premolar 4 0-3o 



Vertical ,, ,, 0-23 



Antero-posterior ,, carnassial 0'37 



Vertical diameter of anterior blade of ditto 0-21 



ditto posterior ,, 0-36 



Antero-posterior diameter of molar 2 0-21 



Compai'ing the upper molars of our fossil, with the published 

 figures of fossil Yiverridee, I find that in size and general character 

 they most nearly resemble the Viverra antiqua of De Blainville,^ 

 a species founded upon a portion of a maxilla containing four teeth 

 from the Miocene Freshwater deposits, St. Gerand le Puy, Allier. 

 Gervais reproduces De Blainville's figure,^ and accepts his deter- 

 mination of the genus. Pomel subsequently referred the fossil to 

 the genus Herpestes,^ as he considered that the teeth indicated a 

 transition from the Civets in the direction of the Genets. M. Filhol 

 has more recently described and figured * a very perfect skull, also 

 from St. Gerand le Puy, which he refers to Viverra antiqua, and 

 although he adopts De Blainville's generic appellation, he admits 

 that in many points of structure it is distinct from any true Viverra, 

 and that by its dentary system it is allied to Genetta ; yet on the 

 other hand he enumerates certain cranial characters that he considers 

 indicate a still nearer relationship with the Civets than with the 



1 " Osteographie, " Genus Viverra, p. 71, pi. 13. 



- " Zool. et Falajont. Franc." 1848-52, Explication, pi. 28, p. 11. 



2 "Catalogue Methodique," 1853, p. 64. 



* " Ann. des Sc. Geol." Paris, 1879, tome x. art. 3, p. 152, pi. 19. 



