CORRECTION OF THE OXYGEN THEORY. 161 



their analogies ; thus potassium, sodium, barium, 

 have been asserted to be metals; iodine, bromine, 

 fluorine, have been arranged as analogical to chlo- 

 rine. Yet there is something vague and indefinite 

 in the boundaries of such classifications and analo- 

 gies ; and it is precisely where this vagueness falls, 

 that the science is still obscure or doubtful. We 

 are led, therefore, to see the dependence of Che- 

 mistry upon Classification ; and it is to Sciences of 

 Classification which we shall next proceed ; as soon 

 as we have noticed the most general views which 

 have been given of chemical relations, namely, the 

 views of the electro-chemists. 



But before we do this, we must look back upon 

 a law which obtains in the combination of elements, 

 and which we have hitherto not stated; although 

 it appears, more than any other, to reveal to us the 

 intimate constitution of bodies, and to offer a basis 

 for future generalizations. I speak of the Atomic 

 Theory, as it is usually termed ; or, as we might 

 rather call it, the Doctrine of Definite, Reciprocal, 

 and Multiple Proportions. 



VOL. III. M 



