Naturale Historical Collections. 145 
that which did not take place with an acid and its salt, occurred with two salts ; 
and M. Beudant has assured himself, that those of the same acid, and especial- 
ly those with the same atomic formula, mingle in all quantities, and that the more 
they are complicated, the more easily they unite, so that the double salts, for 
example, even of an entirely different nature, cannot be obtained pure when they 
crystallize with others in the same solution. Lastly, the facility is still greater 
when the salts are formed.in a solution, than when they are put in entirely form- 
ed,'so that, by double decompositions, we may obtain mixtures extremely varied, 
and even a great number which we could not otherwise procure. The crystals 
thus mixed, assume, however, the form of that of the composing salt whose cha- 
racter predominates ; and, according to other experiments by the same author, of 
which we gave an account in 1820, this predominating salt is not always the 
most abundant. 
These facts have appeared to M. Beudant to throw a strong light upon the 
subject with which he occupied himself. Indeed, when a salt is mixed in small 
quantities with another salt of the same acid, but of a higher order, that is to 
say, which contains an additional proportion of the acid, (if this circumstance be 
not doubtful, ) we must try, at the time of analysis, to detect a superabundance of 
acid. The same thing may take place in relation to the base, when this mixed 
salt is of an inferior order, or when it contains more of the base. 
Some experiments performed after this idea plainly confirm the fact. In dis- 
posing the solutions, so that, by double decomposition or otherwise, soluble salts 
of different orders may be formed of the same acid, M. Beudant obtained, for 
example, carbonates and sulphates of soda, which, with the crystallization and 
the other external characters proper to the bicarbonate and to the trisulphate, 
produce in the analysis excess of acid and diminution of water; which is ex- 
plained very well by comparing the composition of the constituent salts, and by 
calculating the sum of their elements. The author has thus calculated all the 
analysis of mixed salts in his experiments, so as to determine positively the relative 
qualities of the different salts united under the same crystallization, and without 
any excess of acid or of base, or, what is the same thing, any electro-negative or 
electro-positive residue. 
From this time, M. Beudant was no longer astonished at the apparent varia- 
tions observed in minerals. He saw even why they occur more frequently in si- 
licious stones, or silicates ; on the one hand, because these are the most nume- 
rous natural salts ; on the other, because they are of the greatest diversity in the 
degrees of saturation by the different bases; and lastly, because, as geology 
teaches, these are the mineral salts which are most frequently necessitated to 
crystallize together, and consequently most frequently placed in the circumstances 
best adapted to determine extremely varied mixtures. But to apply his method 
with certainty, we must have some idea of what may exist in the solution where 
the substance has been crystallized, and consequently of the sort of mixture which 
might be found there. In the absence of this knowledge, and to approach it as 
nearly as may be, it occurred to M. Beudant to make new analyses, not only of 
one mineral substance taken separately, but of all the substances which he could 
find united in the same group. He announced that he had obtained from this 
labour very positive results to assure him that all known analysis may easily be 
reduced to established laws, if we had for them data similar to those which he has 
employed for his; and the numerous examples of the latter which he has given, 
seem indeed to establish that it is with mineral substances precisely as with salts, 
and that all those which are found in the same solution, are mixed with each 
other at the moment of crystallization, and more or less according to the circum- 
stances which have accompanied it. We know however that, in the complicated 
cases, the object is to resolve equations to several terms more unknown, that is 
to say, that we have indeterminate problems, capable of many solutions, accord- 
ing to the hypothesis which we are oblig2d to make. 
M. Beudant has presented another memoir, wherein he remarks that the most 
VOL. I. T a 
