ARTIFICIAL DEVICES 23 



arbitrary variations at the hands of authors so long as naturaUsts 

 will not agree and submit themselves to certain general principles 

 on the subject. 



Thus, even though the order of nature in a kingdom should be 

 thoroughly known, the classes which we are obUged to establish 

 in it will always be fundamentally artificial divisions. 



It is true, especially in the animal kingdom, that several of these 

 divisions appear to be really marked out by nature herself ; and it is 

 certainly difficult to believe that mammals, birds, etc., are not sharply 

 isolated classes formed by nature. This is none the less a pure illusion, 

 and a consequence of the hmitation of our knowledge of existing 

 or past animals. The further we extend our observations the more 

 proofs do we acquire that the boundaries of the classes, even appar- 

 ently most isolated, are not unhkely to be effaced by our new discoveries. 

 Already the Ornithorhyncus and the Echidna seem to indicate the exist- 

 ence of animals intermediate between birds and mammals. How 

 greatly natural science would profit if the vast region of Austraha 

 and many others were better known to us ! 



If classes are the first kind of division that can be estabhshed in a 

 kingdom, it follows that the divisions which can be established among 

 the objects of one class cannot themselves be classes ; for it is obviously 

 inappropriate to set up class within class ; that, however, is just what 

 has been done : Brisson, in his Ornithologie, has divided the class of 

 birds into various special classes. 



Just as nature is everywhere governed by laws, so too artifice should 

 be subjected to rules. If there are none, or if they are not followed, 

 its products will be vacillating and its purpose fail. 



Some modern naturaUsts have introduced the custom of dividing a 

 class into several sub-classes, while others again have carried out the 

 idea even with genera ; so that they make up not only sub-classes 

 but sub-genera as well. We shall soon reach not only sub-classes 

 but sub-orders, sub-famiUes, sub-genera and sub-species. Now this 

 is a thoughtless misuse of artifice, for it destroys the hierarchy and 

 simphcity of the divisions, which had been set up by Linnseus 

 and generally adopted. 



The diversity of the objects belonging to a class either of animals or 

 plants is sometimes so great as to necessitate the formation of many 

 divisions and sub-divisions among the objects of that class ; but it 

 is to the interest of science that artificial devices should always have 

 the greatest possible simplicity. Now that interest allows, no doubt, 

 of any divisions and sub-divisions that may be necessary; but it is 

 opposed to each division having a special denomination. A stop 

 must be put to the abuses of nomenclature ; otherwise the nomen- 



