xii PREFACE. 



oddly termed Systeme de la Nature, which has been 

 defined by Cuvier to be a great catalogue in which 

 all organized beings have suitable names, can be 

 recognised by suitable characters, and may be dis- 

 tributed into divisions and subdivisions which are 

 themselves also named and characterized. If then 

 such a system as this be termed a dictionary, the 

 true natural system may be reputed the language to 

 which it refers; and as a dictionary is nothing but 

 an useless assemblage of words, without some gram" 

 mar or rules of syntax by means of -which a know- 

 ledge of the structure of the language may be ac^ 

 quired, so an artificial system is a dry unmeaning 

 collection of names, unless it be made subservient 

 to the discovery of the natural one. An artificial 

 -system, according to what has been said, must al- 

 ways be more or less a violation of natural order; 

 since it has no higher pretensions, no other merit, 

 than the readiness and facility with which it may 

 enable an object to be named; and again, because 

 this facility seems principally to depend on the as^- 

 sumption of arbitrary distinctions with which nature 

 is altogether unacquainted. But it is different with 

 respect to the natural system ; since an injury to 

 the order of creation in this ought to be as offensive 

 and as readily perceptible to the naturalist, as an 

 error of syntax is to the grammarian. An artificial 

 system depends solely on observation, and maj: 



