OF THE HOLLOW-HORNED RUMINANTS. 47 



employed to define the genera of Ruminants, are purely trivial and accidental characters, 

 which not only exercise no assignable influence over the habits or oeconomy of the 

 animals, but which may be modified to any extent, or even destroyed altogether, without 

 in the slightest degree changing their generic relations. The small Indian Zebu, for 

 example, is not the less a Bos for having perfectly upright horns, nor the common Buf- 

 falo for being without a dewlap : the Rocky Mountain Goat does not cease to belong to 

 the genus Capra (at least as nature has formed it) because its horns are short and 

 smooth, nor the Wallachian Ram to the genus Ovis because it happens to have these 

 organs as erect and spiral as the Koodo or the Addax. All these are, in truth, accessory 

 and variable characters, mere modifications of form produced by food, climate, and other 

 casual circumstances ; they do not even indicate specific diflferences, much less generic ; 

 and their employment, in the latter sense, is but an example of that empiricism which 

 has so long burthened every department of zoology, and retarded the development of 

 its true philosophical principles. 



But besides these errors of theory, there were practical difficulties which formed an 

 insurmountable obstacle to the subdivision of the genus Antilope into definite or natural 

 groups. These arose from the extremely defective and incorrect manner in which the 

 individual characters of the different species have been hitherto observed and recorded. 

 In fact, no regular or express study of these characters had ever been undertaken in 

 detail ; detached and occasional observations were, no doubt, made upon particular 

 species in the process of general description ; but no regular system was pursued, no 

 patient and oft-repeated comparisons instituted, for the purpose of correcting the de- 

 fects or imperfections of individual specimens ; nor was there any appreciation of the 

 superior value of some characters over others : add to which the difficulty of making 

 such minute observations on dried skins or mounted specimens, where the organs are 

 not unfrequently distorted, concealed, or altogether destroyed in the process of prepa- 

 ration, and the causes will be readily perceived as well of the deficiencies and numerous 

 errors which pervade all the generic monographs and most of the specific descriptions 

 hitherto pubUshed, as of the contradictory characters occasionally assigned by different 

 observers to the same species. Such being the case, it is evident that nothing was to 

 be expected from compilation ; yet to this alternative former monographers were in a 

 great measure reduced, by the poverty of the materials at their disposal, for such facts 

 as related to the lachrymal and inguinal glands, the number of teats, and other charac- 

 ters which form the groundwork of their subdivisions ; and the circumstance is alone 

 sufficient to account for the failure which has confessedly attended their efforts. I have 

 already observed that two very distinct courses were pursued for the attainment of this 

 object. Whilst some, like Pallas, Cuvier, and Lichtenstein, assumed one or more cha- 

 racters without reference to individual species, and made their divisions depend on these 

 predetermined principles, others again, like De Blainville, Desmarest, and Hamilton 

 Smith, without any primary reference whatever to characters, but according to some 



