422 AFRICAN GAME ANIMALS 



of most of the members; narrow in only one species, the 

 common eland. The tail varies greatly from the short, 

 bushy tail of the bushbuck to the long, tufted tail of the 

 eland. The skull has long nasal bones, short snout and short 

 nasal cavity, and no lachrymal vacuity in front of the orbit. 

 A large sinus is present between the nasal and lachrymal 

 glands as in the waterbucks. 



The effort to divide this family into genera and species 

 is fraught with difficulties that illustrate clearly how arti- 

 ficial the terms *'genus," and "species" are. It is absolutely 

 necessary to employ the terms and in some cases they meet 

 all the needs of the situation, but in other cases, as with the 

 tragelaphs, all that can be said is that they are necessary 

 but that they are also unsatisfactory. 



Nearly all the species of the TragelaphincB are so closely 

 allied that they might all be included in a single genus. 

 Such an arrangement would, however, result in consider- 

 able geographical confusion and obscure the real relation- 

 ships of the species. Nevertheless, the attempt to make the 

 genera of equal weight so as to express the relationships 

 clearly, or avoid confusion, results in a multiplicity of 

 genera; and this means splitting into groups a number of 

 closely allied species. In fact, arguments of some weight 

 can be advanced for either uniting all the tragelaphs into 

 one genus, or for making almost as many genera as there are 

 species; and in this same way arguments can be advanced 

 for both splitting up into a large number of species, and 

 for reducing the great majority of these from specific to 

 subspecific rank. The genera adopted by most writers 

 are based almost solely upon horn characters. As a single 

 character, the shape of the horns is certainly the most re- 



