enable us to throw them mto groups approximating more or less to 

 the scheme of Nature, and in that sense natural. 



Mr. Edmonston's remarks on the process by which botanical know- 

 ledge is or may be most easily acquired, are lor the most part good ; 

 still I cannot help considering him mistaken on some points. For 

 example, he says that a finished botanist has no need of classifica- 

 tion ; and here I differ from him. I grant that a botanist who knows 

 all the British plants, does not need classification for the purpose of 

 distinguishing species, since he knows them and their characters; but 

 still he needs it to show their analogies and affinities : and I contend 

 that were a man capable of comprehending in one view the whole ve- 

 getable kingdom, with all the distinctions and analogies of species, 

 the natural tendency of his mind would compel him to throw them 

 into groups ; and supposing his knowledge perfect, he would con- 

 struct a perfectly natural system. Hence the natural system will 

 become more and more perfect by its progressive changes, and ap- 

 proximate more and more to a conformity with Nature. It appears 

 to me that Mr. Edmonston, throughout his letter, lays more stress on 

 the distinguishing of species from each other, than combining them 

 into a coherent whole. He would consider Botany as an analytic ra- 

 ther than a synthetic science, or both combined. 



Descending from the high ground we have hitherto occupied, I 

 must be permitted to notice a few of the detailed objections put for- 

 ward by Mr. Edmonston against the natural system. He says that 

 there has never been a system which was not liable to exceptions, yet 

 he expects that the natural system should be liable to 7ione : and 

 commencing with the three primary groups — Mono- Di- and Acoty- 

 ledons, he discovers a few exceptions, and these he may be permitted 

 to make the best of. But what would Mr. Edmonston say, if the 

 same species (Nature only creates species) should be found usually 

 with two, but occasionally with three or four cotyledons ? Would he 

 separate the individuals and place them in the two primary divisions, 

 although in other respects they perfectly resembled each other ? Yet 

 this would be necessary to produce his idea of uniformity. Such a 

 circumstance has been observed in Sinapis ramosa, and is figured in 

 Decandolle's ' Organographie,' plate 53. It is true that Acotyledo- 

 nous plants may be found among Dicotyledons ; indeed we should 

 not expect two seed-leaves in a plant entirely destitute of leaves, like 

 the Cuscuta, although in all other respects it may be and is accordant 

 with our idea of Dicotyledonous structure. But if Mr. Edmonston 

 will cast his eye over the following table, he will perceive that the in- 



