Mr Glovei- on Forms of Indnct'ion. 119 



properties fixed on for its types have an actual existence in all 

 the particular instances composing the included genus. Thus, 

 the Linnaean arrangement of plants is as just a system, so fai* 

 as the mind is concerned, as that of Jussieu. It is true, the 

 one system takes cognizance of a greater number of characters 

 in composing its genera than are considered in the classes and 

 orders of the other, and also of such characters as are believed 

 most essential to the very nature of the individual plants. This 

 system is therefore physically the more perfect of the two, but 

 it is not therefore more logical than the other. In that respect 

 both are alike, — both sufficiently accurate in logical structure, 

 but framed for different ends. 



The preceding observations may in some degree illustrate a 

 great maxim of the Kantians, which makes the fundamental 

 principles of all science repose in the intellect itself; — asserting 

 the human understanding rather to dictate the laws which re- 

 gulate its acquisition of knowledge, than receive them from the 

 external world.* Indeed all general notions are the workman- 

 ship of the mind, and often cannot be ascertained to correspond 

 exactly with existences and actions of nature. And this, even, 

 on the understanding that such notions are derived in all cases 

 by an exercise of the intellective faculty from real impressions. 

 For example, the intellectual forms which are the objects of 

 geometrical reasoning, and which, being ideas of relation, have 

 somewhat of the character of general notions, are not to be found 

 pure in nature. And something similar, or at least analogous, 

 holds of physical laws also ; for these are either general terms 

 signifying the agreement of a number of particular facts or phe- 

 nomena in some common properties, or else abstractions of some 

 actions of Nature from others with which they must in many 

 cases be viewed in their real state, somewhat combined. 



The characteristic, or what may be termed the logical cha- 

 racters of all physical laws are similar. For a definition of a 

 physical law, in logic, it is enough to term it a statement im- 

 plying aconnexion observed between some properties and others, 

 in any definite class of instances. It is quite essential, that the 

 class of facts to which a law is applicable should be defined, 



• Philosophic de Kant, par Villiers, p. 301, 8vo. Metz, 1801. The 

 same truth is elaborately illustrated by Dr Brown in his 5th lecture. 



