194 EMENDATIONS 



celebrated Mr. Roscoe, in Tr. of Linn. Soc. v. 11. 65, 

 who, in showing that this method involves several 

 as unnatural assemblages as the professedly artifi- 

 cial system of Linnaeus, contends, that little is to 

 be gained by it's adoption, with respect to a confor- 

 mity to nature. Every one must also perceive, that 

 no use can be made of any such system, in the prac- 

 tical or analytical examination of plants. Natural 

 Orders indeed must, in future, be studied by all who 

 deeply contemplate theVegetable Kingdom, and some 

 links of connexion, or points of discrimination, can- 

 not but be kept in view between them. We require 

 a cabinet, as it were, with cells or drawers, where we 

 may find each Order as we want it ; and Jussieu's 

 classification, with all it's unavoidable imperfections, 

 goes much beyond any system previously invented, in 

 the natural assemblages which it produces. Never- 

 theless, Linnaeus has truly observed that Natural Or- 

 ders are related to each other by so many points, that 

 they rather resemble a geographical map, than a con- 

 tinued series ; which he has attempted to illustrate by 

 example, in his Prcrtectiones published by Giseke. 

 There remains therefore, in the study of natural clas- 

 sification, only a choice of difficulties ; and while we 

 labour to bring plants together, as naturally as pos- 

 sible, in groups or families, for their mutual illustra- 

 tion, we must perpetually relax or vary those general 

 ties, of which we can, as yet, obtain but very con- 

 fined and imperfect views. Hence therefore I am 



