ANALYSIS OF MINERAL WATERS. 261 
dr nearly impracticable, to assign the precise composition, and 
the real proportions of the compound salts; and hence the 
necessity of employing the direct method of obtaining them. 
The present state of the science leads to other views. 
If the conclusion were just, that the salts obtained by eva- 
poration, or any analogous process from a mineral water, are 
its real ingredients, no doubt could remain of the superiority 
of the direct method of analysis ; and even of the absolute ne- 
cessity of employing it. But no illustrations, I believe, are re- 
quired to prove, that this conclusion is not necessarily true. 
The concentration by the evaporation, must, in many cases, 
change the state of combination, and the salts obtained are 
hence frequently products of the operation, not original ingre- 
dients. Whether they are so or not, and what the real com- 
pesition is, are to be determined on other grounds than on 
their being actually obtained; and no more information is 
gained, therefore, with regard to that composition, by their be- 
ing procured, than by their elements being discovered ; for 
when these are known, and their quantities are determined, 
we can, according to the principle from which the actual 
modes of combination are inferred, whatever this may be, as- 
sign with equal facility the quantities of the binary compounds 
they form. 
The accuracy with which the proportions of the constituent 
principles of the greater number of the compound salts are now 
determined, enables us also to do this with as much precision, 
as by obtaining the compounds themselves. And if any error 
should exist in the estimation of these proportions, the prose- 
cution of these researches could not fail soon to discover it. 
The mode of determining the composition of a mineral wa- 
ter, by discovering the acids and bases which it contains, ad- 
mits, in general, of greater facility of execution, and more ac- 
curacy, 
