4 INTRODUCTION. 



being, not only with regard to itself but to its surrounding 

 relations. The analysis of these conditions frequently con- 

 ducts us to general laws, as certain as those that are derived 

 from calculation or experiment. 



It is only when all the laws of general physics and those 

 which result from the conditions of existence are exhausted, 

 that we are reduced to the simple laws of observation. 



The most effectual method of obtaining these, is that of 

 comparison. This consists in successively observing the same 

 bodies in the different positions in which nature places them, or 

 in a mutual comparison of different bodies ; until we have 

 ascertained invariable relations between their structures and 

 the phenomena they exhibit. These various bodies are kinds 

 of experiments ready prepared by nature, who adds to or de- 

 ducts from each of them different parts, just as we might wish 

 to do in our laboratories ; showing us, herself, at the same time 

 their various results. 



In this way we finally succeed in establishing certain laws 

 by which these relations are governed, and which are em- 

 ployed like those that are determined by the general sciences. 



The incorporation of these laws of observation with the 

 general laws, either directly or by the principle of the con- 

 ditions of existence, would complete the system of the natural 

 sciences, in rendering sensible in all its parts the mutual in- 

 fluence of every being. To this end, should those who culti- 

 vate these sciences direct all their efforts. 



All researches of this nature, however, pre-suppose means 

 of distinguishing clearly^ and causing others to distinguish, the 

 bodies they are occupied with ; otherwise we should be con- 

 tinually confounding them. Natural history then should be 

 based on what is called a system of nature; or a great cata- 

 logue in w^hich all created beings have suitable names, may be 

 recognised by distinctive characters, and be arranged in divi- 

 sions and subdivisions, themselves named and characterised, 

 in which they may be found. 



In order that each being may be recognised in this catalogue, 

 it must be accompanied by its character: liabits or properties 



