INTRODUCTION. 57 



tion, so happily recognised by modern chemists, and pro- 

 claimed under the ancient form of atomic symbols, still 

 remains isolated and independent of mathematical laws of 

 motion and gravitation. 



Those productions of nature which are objects of direct 

 observation may be logically distributed, in classes, orders, and 

 families. This form of distribution undoubtedly sheds somei 

 light on descriptive natural history, but the study of organ- 

 ised bodies, considered in their linear connection, although it 

 may impart a greater degree of unity and simplicity to the 

 distribution of groups, cannot rise to the height of a classifi- 

 cation based on one sole principle of composition and internal 

 organisation. As different gradations are presented by the 

 laws of nature according to the extent of the horizon, or the 

 limits of the phenomena to be considered, so there are like- 

 wise differently graduated phases in the investigation of the 

 external world. Empiricism originates in isolated views, 

 which are subsequently grouped according to their analogy or 

 dissimilarity. To direct observation succeeds, although long 

 afterwards, the wish to prosecute experiments, that is to 

 say, to evoke phenomena under different determined condi- 

 tions. The rational experimentalist does not proceed at 

 hazard, but acts under the guidance of hypotheses, founded 

 on a half indistinct and more or less just intuition of the con- 

 nection existing among natural objects or forces. That which 

 has been conquered by observation or by means of experi- 

 ments, leads, by analysis and induction, to the discovery of 

 empirical laws. These are the phases in human intellect 

 that have marked the different epochs in the life of nations ; 

 and by means of which that great mass of facts has been 

 accumulated which constitutes at the present day the solid 

 basis of the natural sciences. 



Two forms of abstraction conjointly regulate our knowledge, 

 namely, relations of quantity, comprising ideas of number and 

 size, and relations of quality, embracing the consideration of 

 the specific properties and the heterogeneous nature of matter. 

 The former, as being more accessible to the exercise of thought, 

 appertains to mathematics, the latter, from its apparent mys- 

 teries and greater difficulties, falls under the domain of the 

 chemical sciences. In order to submit phenomena to calcu- 

 lation, recourse is had to a hypothetical construction of matter, 



