TRANSITION TO CLASSIFICATORY SCIENCE. 305 



great steps of its advance, from the most unconnected facts and the 

 most imperfect speculations, to the highest generalization at which 

 chemical philosophers have yet arrived. 



Yet it will appear to our purpose to say a few words on the connex- 

 ion of this science with those of which we are next to treat ; and that 

 T now proceed to do. 



CHAPTER X. 



TRANSITION FROM THE CHEMICAL TO THE CLASSIFICATORY SCIENCES. 



IT is the object and the boast of chemistry to acquire a knowledge 

 of bodies which is more exact and constant than any knowledge 

 borrowed from their sensible qualities can be ; since it penetrates into 

 their intimate constitution, and discloses to us the invariable laws of 

 their composition. But yet it will be seen, on a little reflection, that 

 such knowledge could not have any existence, if we were not also atten- 

 tive to their sensible qualities. 



The whole fabric of chemistry rests, even at the present day, upon 

 the opposition of acids and bases : an acid was certainly at first known 

 by its sensible qualities, and how otherwise, even now, do we perceive 

 its quality ? It was a great discovery of modern times that earths and 

 alkalies have for their bases metals : but what are metals ? or how, 

 except from lustre, hardness, weight, and the like, do we recognize a 

 body as a metal ? And how, except by such characters, even before 

 its analysis, was it known to be an earth or an alkali ? We must sup- 

 pose some classification established, before we can make any advance 

 by experiment or observation. 



It is easy to see that all attempts to avoid this difficulty by referring 

 to processes and analogies, as well as to substances, bring us back to 

 the same point in a circle of fallacies. If we say that an acid and alkali 

 are known by combining with each other, we still must ask, What is 

 the criterion that they have combined ? If we say that the distinctive 

 qualities of metals and earths are, that metals become earths by oxida- 

 tion-, we must still inquire how we recognize the process of oxidation ? 

 \Ve have seen how important a part combustion plays in the history 

 of chemical speculation ; and we may usefully form such classes of 

 VOL. II. 20. 



