260 A GENERAL FORMULA FOR THE 



the direct method, in which, by evaporation, aided by the subse- 

 quent application of solvents, or sometimes by precipitants, 

 certain compound salts are obtained ; and what may be called 

 the indirect method, in which, by the use of re-agents, the prin- 

 ciples of these salts, that is, the acids and bases of which they 

 are formed are discovered, and their quantities estimated, 

 whence the particular salts, and their proportions, may be in- 

 ferred. 



Chemists have always considered the former of these me- 

 thods as affording the most certain and essential information : 

 they have not neglected the latter ; but they have usually em- 

 ployed it as subordinate to the other. The salts procured by- 

 evaporation, have been uniformly considered as the real ingre- 

 dients, and nothing more was required, therefore, it was ima- 

 gined for the accuracy of the analysis, than the obtaining them 

 pure, and estimating their quantities with precision. On the 

 contrary, in obtaining the elements merely, no information, it 

 was believed, was gained with regard to the real composition, 

 for it still remained to be determined, in what mode they were 

 combined, and this, it was supposed, could be inferred only 

 from the compounds actually obtained. This method, there- 

 fore, when employed with a view to estimate quantities, has 

 been had recourse to only to obviate particular difficulties at- 

 tending the execution of the other, or to give greater accuracy 

 to the proportions, or, at farthest, when the composition is 

 very simple, consisting chiefly of one genus of salts. 



Another circumstance contributed to lead to a preference of 

 the direct mode of analysis ; — the uncertainty attending the de- 

 termination of the proportions of the elements of compound 

 salts. This uncertainty was such, that even from the most ex- 

 act determination of the absolute quantities of the acids and 

 bases existing in a mineral water, it would have been difficult, 



or 



