SYSTEMS OF CLASSIFICATION 13 



artificial system have, therefore, no necessary agreement with 

 each other except in the characters selected for convenience as 

 the types of those divisions respectively. Hence such a system 

 may be compared to a dictionary, in which words are arranged, 

 for convenience of reference, in an alphabetical order, adjacent 

 words having no necessary agreement with each other, except 

 in commencing with the same letter. In the Natural System, 

 on the contrary, aU the characters of the genera are taken into 

 consideration, and those are grouped together into orders which ' 

 correspond in the greatest number of important characters ; and 

 the orders are again united, upon the same principles, into 

 groups of a higher order, namely, the classes and other divi- 

 sions. "While it must be evident, therefore, that all the know- 

 ledge we necessarily gain by an artificial system is the name of 

 an uninown plant, on the other hand, by the natural system, 

 we learn not only the name, but also its relations to the plants 

 by which it is surrounded, and hence get a clue to its structure, 

 properties, and history. Thus, supposing we find a plant, and 

 wish to ascertain its name, if we turn to the Limiasan System 

 and find that such a plant is the Menyanthes trifoliata, this 

 name is the whole amount of the knowledge we have gamed ; 

 but by turning to the Natural System instead, and finding that 

 our plant belongs to the order Gentianacew, we ascertain at 

 once from, its affinities that it . probably has the tonic and 

 other properties which are possessed by the plants generally of 

 that order, and, at the same time, we also learn that it accords in 

 its structure with the same plants. It is quite true that all the 

 orders, as at present constituted, are by no means so natural 

 as that of the Gentianacese, but this arises from the present 

 imperfection of our systems, and can only be remedied as 

 our knowledge of plants extends ; a system, devised as per- 

 fectly as possible one day, ma,y be deficient the next, in conse- 

 quence of new plants being discovered which may compel us to 

 alter our views, for at present the floras of many regions of the 

 globe are imperfectly known, and those of others almost entirely 

 unknown. Sufacient, however, is now known of plants to 

 enable us to establish certain great divisions according to a 

 natural method, which after discoveries are not likely to affect 

 to any important extent. The present imperfections of the 

 Natural System are, therefore, comparatively unimportant, and 

 wiU no doubt disappear as our knowledge of the flora of the 

 globe becomes extended. 



Having now described the general characters upon which 



