490 Mr. BicuEno on Systems and Methods 
This attempt at breaking down good orders and genera into 
many subordinate and loosely defined groups, and encumbering 
them with names, involves the subject in obscurity, and may well 
be questioned as contrary to his main design of presenting those 
comprehensive views which are afforded by a natural system. 
Mr. Brown has adopted a different mode in his ** Prodromus." 
He has attempted to combine no further than his knowledge 
would warrant, not even employing the terms class or order as 
the names of his groups. As his object is chiefly synthesis, he 
keeps his diagnostic characters apart, thus leaving the mind less 
embarrassed when it is in pursuit of analysis. It must be ad- 
mitted indeed, that his work cannot be employed with any suc- 
cess by the inexperienced, or even by those who have occupied 
themselves only in searching for species; but to have made it 
subservient to this purpose, would have been to have rendered 
it less beautiful and complete as a work of synthesis. His apho- 
risms and remarks not being reduced to exact method, * are," as 
Lord Bacon expresses it, “ still in their growth, increasing in 
bulk and substance." : | 
. Now wherever the object of the systematist is to enable his 
reader to discover species, it is necessary to define at every step; 
and where natural characters do not present themselves, we must 
adopt artificial ones. For this purpose large classes are formed, 
many of which are necessarily artificial. These again are broken 
up into orders, mostly of an artificial character; and thus the 
naturalist is led step by step from more comprehensive definitions 
to less, from class to order, from order to genus, and from genus 
to species. In this descending series it will be observed that 
the essential feature is the facility that is afforded for definition. 
. Hence the Linnæan system of botany has succeeded so well, be- 
cause its author selected chiefly as the ground of his arrange- 
ment the number and proportion of parts most obvious and 
| re least 
