ANALYSIS OF MINERAL WATERS. 263 
are not by another. In a water which is of complicated com- 
position, this will more peculiarly be the case. The Chelten- 
ham waters, for example, have, in different analyses, afforded 
results considerably different ; and, on the supposition of the 
salts procured being the real ingredients, this diversity must 
be ascribed to inaccuracy, and ample room for discussion, with 
regard to this is introduced. In like manner, it has often 
been a subject of controversy, whether sea-water contains sul- 
phate of soda with sulphate of magnesia. All such discus- 
sions, however, are superfluous. The salts procured are not 
necessarily the real ingredients, but in part, at least, are pro- 
ducts of the operation, liable, therefore, to be obtained or not, 
or to be obtained in different proportions, according to the me- 
thod employed. And all that can be done with precision, is 
to estimate the elements, and then to exhibit their binary 
combinations according to whatever may be the most probable 
view of the real composition. 
The process I have to state, conformable to these views, is 
essentially the same as that which I employed in the analysis of 
sea-water in a preceding memoir ; and it was the consideration 
of the advantages belonging to it, that has led me to propose 
it, with the necessary modifications, as one of general applica- 
tion. 
Mineral waters have been arranged under the four classes 
of Carbonated, Sulphureous, Chalybeate, and Saline. But all of 
them are either saline, or may be reduced under this division, 
From waters of the first class, the carbonic acid which is in ex- 
cess, is expelled by heat, and its quantity is estimated. Sul- 
phuretted hydrogen is in like manner expelled or decomposed. 
And iron may be detected by its particular tests, and removed 
by appropriate methods. In all these cases the water remains, 
with any saline impregnation which it has, and of course is es- 
sentially 
