THE PROTELES LALANDII. 283 



exceed a common fox. All that is kno^vn of its 

 manners is, that it frequents clefts of rocks. 



Herewith we terminate the histoiy of digitigrade 

 carnassiers belonging to the canine family or allied 

 genera, without including in the series the new 

 genus Cynict'is, lately established by Mr. Ogilby: 

 for although it appears, from the short notices 

 published respecting two species, that they range 

 between the dogs and Civets, and are intermediate 

 between Proteles and Ryzena, not having personally 

 examined specimens, we are uncertain, but believe 

 it belongs to the elongated group, and therefore 

 should stand at the head of the Yiverrine genera. 



In conclusion, we may be permitted to remind 

 the reader, that in the groups before described, spe- 

 cific names have been bestowed upon several animals 

 which may prove to be only varieties or races of 

 species, because, while the limits of what consti- 

 tutes a species remain unsettled, it was judged, in 

 doubtful cases, preferable to point them out under 

 distinct appellations, rather than leave their indica- 

 tions mixed up with others under some common 

 name, often found to conceal more than one spe- 

 cies. In this manner, the objects are at least 

 presented to the naturalist in a form that provokes 

 his attention and invites comparison, whereas in 

 the more generalised practice Avliole groups are 

 liable to escape detection under the name of one 

 alone. AVe hope likeAvise to have rendered a ser- 

 vice to science by the application of distinctive 

 names to groups, because by their means are avoid- 



