108 CARN1VORA. 



tendency. Huxley has indicated a strong resemblance between the 

 skulls of some species of Cynodictis and certain existing American 

 forms of Canis ; and the close resemblance of the skeleton of Cyno- 

 dictis to that of Amphicyon has been noticed by Filhol, who has also 

 shown that in some of the larger forms the lower dentition is very 

 like that of Amphicyon, which is again very like Cephalogale. The 

 humerus of Cynodictis (or, at all events, of most of the species) had 

 an entepicondylar foramen, thereby differing from that of Canis, 

 and agreeing with Viverra, Amphicyon, and Cephalogale. 



It thus appears that Cynodictis is an extremely generalized genus, 

 connecting the Viverroids and the Ganoids so intimately that it is 

 almost impossible to say with which family it should be classed. 

 Future discoveries will probably ere long render it impracticable to 

 generically distinguish Herpestes, Amphictis, Cynodictis, and Canis 

 from one another, or from the forms to which the names Amphi- 

 cynodon and Cynodon have been applied. 



In quite another direction Filhol 1 has indicated that in those 

 forms of Cynodictis in which the last lower molar is absent there is 

 a complete passage through Stenoplesictis to Palceoprionodon (vide 

 Ann. Soc. Sci. Phys. Nat. Toulouse, 1882, pp. 67-8) ; while from 

 the former the transition to the musteline genus Plesictis is gradual, 

 and the latter is but slightly removed from the feline Pseudcelurus. 



Cynodictis lacustris, P. Gervais 2 . 

 Syn. C. mungoides and C. antiqua, Bravard, MS. 

 Hob. France. 



27572. Fragment of the left maxilla, containing m. 1 ? m. 2 ? and a 



portion of pm. 4 ; from the Upper Eocene of Apt ( Vaucluse), 

 France. These teeth, although more perfect, appear to be 

 indistinguishable from the corresponding type specimens 

 figured by P. Gervais in the Zool. et Pal. Franchises, 2nd ed. 

 pi. xxv. fig. 2 ; together with the next specimen, they are 

 entered in Bravard's MS. Catalogue in the Museum as 

 C. mungoides, Bravard 3 . 



Bravard Collection. Purchased, 1852. 



27573. Portion of the left ramus of the mandible, containing the 



1 Ann. Sci. Geol. vol. viii. art. 1, p. 49, and Ann. Soc. Sci. Phys. Nat. 

 Toulouse, op. cit. 



2 Zool. et Pal. Fran 9 ai 8 es, 1st ed. vol. i. p. 133 (1848-52). The present writer 

 is unacquainted with the date of publication of this portion of the work. 



3 This name, together with the names C. dubia (non Filhol), C. antiqua 

 and C. angustidens, occurs in some (apparently) unpublished plates of Bra- 

 Yard's in the Museum, in which the specimens noticed aboye are figured. 



