318 
ON POTASH AND SODA IN SOILS. 
method might not be devised for accomplishing this object, 
and I am now prepared to place before your Society my new 
plan, and to detail some of the experimental results which 
confirm its exactness and general applicability. Its simplici- 
ty will be best demonstrated by contrasting it with the direc- 
tions given by that most excellent chemist Berzelius, than 
whom no analyst deserves more of our praise or confidence, 
or has done more towards bringing analytical chemistry to 
that high standard of perfection to which it has now at- 
tained. 
After setting before chemists the two processes, and assur- 
ing them that the second does not yield in accuracy to the 
first, I shall leave it for them to choose which they will in 
future employ. 
Berzelius's words are these, which I translate from the 
French of M. Essligen. 
u If it should happen that both alkalies be present, it is 
very (assez) difficult to separate them, so as to determine 
their exact proportions. Of several modes I have tried, the 
following has succeeded the best ; we mix the salt obtained 
(chloride or chlorides,) with three and three quarter times its 
weight of crystallized chloride of platinum and sodium, this 
being the precise quantity necessary for the potassium to re- 
place the sodium contained in the double salt, on the suppo- 
sition that the salt we are examining is nothing but chloride 
of potassium. 
" We dissolve the mixture in a very little water, and evap- 
orate to dryness, by a gentle heat, then we treat it with alco- 
hol, which dissolves the chloride of sodium, and the chloride of 
platinum and sodium,but it does notdissolve the chloride of pla- 
tinum and potassium, which is to be washed with alcohol, and 
then dried and weighed ; 100 parts of this salt contain 30.86 
of chloride of potassium. 
" The deficiency of the original weight is chloride of sodi- 
um. The quantity of pure alkali corresponding to these chlo- 
rides iscalculated from the tables. ; 
" Remark, — It is not equally advantageous to use the sim- 
