765 



of plants thrown together under it, and when some of them did not 

 agree with the definition, exception after exception were interlined, 

 until a great part of it looks something very like nonsense. 



It appears to me impossible that when the same characters are em- 

 ployed to designate classes, orders, genera and species, that a distinct 

 definition can ever be established. It is undeniable that certain cha- 

 racters are of ordinal, others of generic, and others again only of spe- 

 cific importance ; and characters which in some plants may be used 

 for the primary divisions, in others can only be employed for species. 

 Experience alone is the test of this ; and in consequence of this fluc- 

 tuating importance of characteristic marks, it is impossible strictly to 

 set apart such and such characters for such and such divisions ; de- 

 spite this fact, however, it is well known that there may be certain 

 rules laid down which may very generally be followed, subject to 

 some exceptions. And the tendency towards always employing the 

 same marks of distinction for all groups, most inevitably leads to con- 

 fusion and anarchy in any science. When it is borne in mind that 

 all the divisions are in every case arbitrary, and when one character 

 cannot be obtained another must be sought for — the cause of this dif- 

 ference of importance will be at once apparent. Enough however on 

 this head. As it has been proved that all the divisions of botanists, 

 under the names of classes, groups, alliances, orders, genera, sections 

 &c. have no existence in nature, it of course follows that any system 

 has for its only ground of utility the facilitating of the knowledge of 

 species ; thus in one view all systems are alike artificial, and the only 

 test whereby to know which is the best, is to be found in their practi- 

 cal utility. 



I object to the " natural " system as the best means of acquiring a 

 knowledge of species, for the following out of many reasons: — 



1. Because in this kind of system the principle of contrast or diag- 

 nosis is not sufficiently kept in view. A definition or specific charac- 

 ter ought to embrace nothing but those points which distinguish the 

 group it is defining from all others : instead of this being the case, 

 the natural school substitutes cumbersome descriptions, containing 

 little definite matter, in which differential characteristics are buried. 



2. These differential characters themselves are often injudiciously 

 chosen. This is especially apparent when some rather obscure and 

 not easily discernible characters, derived from the structure of the 

 ovules and seed, are employed to distinguish principal groupings. It 

 is a rule that the characters of primary divisions ought always, when 

 practicable, to be taken from some parts of the plant easily seen.) or 



