GENERALISATION OF LAWS. 61 



clearness to descriptive natural history; but the study of 

 organised bodies, arranged in linear connection, though it 

 gives greater unity and simplicity to the distribution of 

 groups, cannot rise to the height of a classification founded 

 on a sole principle of composition and internal organisa- 

 tion, As different gradations aie presented by natural laws, 

 according as they embrace narrower or wider circles of phse- 

 nomena, so there are successive steps in empirical investi- 

 gation. It begins by single perceptions, which are after- 

 wards classed according to their analogy or dissimilarity. 

 Observation is succeeded, at a much later epoch, by experi- 

 ment, in which phenomena are made to arise under condi- 

 tions previously determined on by the experimentalist, guided 

 by preliminary hypotheses, or a more or less just intuition 

 of the true connection of natural objects and forces. The 

 results obtained by observation and experiment lead by the 

 path of induction and analogy to the discovery of empirical 

 laws ; and these successive phases in the application of the 

 human intellect have marked different epochs in the life of 

 nations. It has been by adhering closely to this inductive 

 path, that the great mass of facts has been accumulated 

 which now forms the solid foundation of the natural sciences. 

 Two forms of abstraction govern the whole of this class 

 of knowledge; viz. relations of quantity, comprehending 

 the ideas of number and magnitude; and relations of 

 quality, embracing the specific properties of heterogeneous 

 matter. The first of these forms, more accessible to the 

 exercise of thought, belongs to the domain of mathematics ; 

 the other, more difficult to seize, and apparently more mys- 

 terious, to that of chemistry. In order to submit pheno- 

 mena to calculation, recourse is had to a hypothetical con- 



