VIVERRIDJE. 45 



in. The hinder jmH of the tarsus bald, callous. 

 a. Tail thick, strong, jji-ehensile. Arctictidina. 



9. Arctictis. 



b. Tail very long, stibconvolute : freniim naked, glandular ; head elongate. 

 Paradoxm-ina. 



10. Nandinia. riesh-tooth elongate, triangular ; tubercular 

 teeth triangular, transverse. Orbit rather incomplete. Pa- 

 late narrow, short. 



11. Pakadoxurus. Flesh-tooth elongate, triangular ; tubercular 

 teeth oblong. Orbit very incomplete. 



12. Paguma. Flesh-tooth short, triangular, large. Orbit vciy 



imperfect. 



13. Arctogale. Flesh-tooth triangular, small. Orbit nearly 



complete. Palate very narrow, elongate. 



The shortness of the characters that I give to some genera has 

 been objected to by several writers, especially by amateurs who have 

 not studied the Linnean brevity and method of description. They 

 overlook the fact that the characters of the sections and subsections 

 of the family that precede the genus form an essential part of the 

 generic character, in the same manner that the section of the genus 

 is part of the specific character of the species that the section con- 

 tains. The definition of the subsections of the famihes and genera 

 requires more study, analysis, and consideration than the writing 

 out of a long generic character, that contains particulars that ai-e 

 common to a number of allied genera, such as the writers who make 

 the complaint usually give. At the same time, the use of such de- 

 tailed characters requires a greater exertion on the reader's part to 

 eliminate the essential particulars, which ai'e the real characters of 

 the group. In the above table, the most easily seen and often 

 empirical characters are purposely chosen, for facility of use and 

 bre\nty. I have even used the colour of the animal for this pur- 

 pose ; for it has a great influence on the formation of a natural 

 genus — more than many zoologists are willing to admit. Even 

 those who know this fact avoid making use of it, ajjparently fearing 

 that it might not be considered scientific ! In the body of the essay, 

 longer generic characters are given. Those who object to analytic 

 characters forget the immense number of animals now known, and 

 the great advantage of a rapid way of discovering the name of the 

 animal they seek, and whose history they desire to know. As Mr. 

 \V. S. Macleay justly observes, " the modern art of describing is too 

 long, often insufl^crably long, while human life remains as short as 

 ever." — Illust. Zool. !South Africa, p. 54. 



