LAVOISIER. 331 



compound of hydrogen, nitrogen, and carbon ; hydro- 

 bromic; fluoric-acid; ferro-cyanic acid; sulpho-cyanic ; 

 hydro-selenic; hydriodic: xanthic. Even if fluoric be 

 omitted, here are nine undeniable acids, and all without 

 a particle of oxygen in their composition. Again, the 

 mere fact of calcination should have prevented him 

 from so generalizing, for all calces contain oxygen, 

 and many of them have no acid qualities. Indeed, 

 his own conjecture, since fully confirmed by experi- 

 ment, that the fixed alkalis are oxides, is a still more 

 striking disproof of his theory ; for it appears that he 

 might just as well have called oxygen the alkalizing 

 principle as the acidifying, or rather much better, 

 since all the alkalis save one contain it and the alka- 

 line earths to boot. But he also should have recol- 

 lected that no acid of them all contains so much oxygen 

 as water, and yet nothing less like an acid can well be 

 imagined. We now have still further instances of the 

 same kind against this theory, and which might justify 

 us in calling hydrogen the acidifying principle as well 

 as oxygen. Upwards of two hundred acids contain 

 hydrogen either with or without oxygen present. 

 Hence he might really have reckoned hydrogen the 

 acidifying principle upon fully better grounds than 

 support his choice of oxygen ; and the truth appears 

 to be, that there is no one substance which deserves 

 the name. 



It is, then, quite clear that M. Lavoisier committed 

 a great error in his induction, and that he framed a 

 theory which was in the extent to which he pressed it 

 wholly without foundation not merely without suffi- 

 cient proof from the facts, but contrary to the facts. 

 Newton gives it as a fundamental rule of philosophising, 

 that we are to state the inferences from phenomena 

 with the exceptions which occur, and if a first induc- 

 tion should be made from imperfect views of the phe- 

 nomena, then to correct it by the exceptions afterwards 

 found to exist. But from this rule Lavoisier has 



