REPORT ON CHEMISTRY. 
423 
oxides, let us take for an example the peroxide of iron and alu¬ 
mina. In the analyses of crystallized minerals, these two bases 
are found to replace one another in various proportions, with¬ 
out altering the form of the crystal. The peroxide of iron is 
universally regarded as a compound of two atoms of iron with 
three of oxygen; hut the law of Mitscherlich requires that 
whatever number of atoms are removed by replacement, the 
same number must be contained in the body which replaces 
it; the combining proportions of iron and alumina must there¬ 
fore contain the same number of atoms. It is a further deduc¬ 
tion from the discoveries of Mitscherlich,—first pointed out 
by the isomorphism of the phosphoric and arsenic acids, and 
afterwards confirmed by numerous other examples,—that bi¬ 
nary compounds which replace each other contain not only the 
same absolute number of atoms, but also the atoms of the two 
elements in the same relative proportion. If therefore in perox¬ 
ide of iron the iron is to the oxygen in the atomic ratio of two 
to three, the atoms of aluminum and oxygen must in alumina 
have the same ratio,—or both bases must be sesqui-oxides. 
Similar reasoning leads to the same conclusion in regard to the 
peroxides of manganese and oxide of chromium, which form 
part of the same isomorphous group with alumina and peroxide 
of iron. 
Oxides of Chromium .—The compounds of chromium with 
oxygen afford a beautiful illustration of the application of the 
law in the way now adverted to. The oxygen in the oxide of 
chromium is to that in chromic acid as 1 :2. Now this is com¬ 
patible with, and indeed naturally suggests, the idea, that in the 
oxide one atom of the metal is combined with one of oxygen, 
and in the acid with two; but the oxide is isomorphous with 
the peroxide of iron, alumina, and the peroxide of manganese, 
while the acid is in like manner isomorphous with the sulphuric 
and selenic acids: the oxide therefore must be composed like 
the peroxide of iron, and the acid like sulphuric acid; that is 
to say, the former must contain two of base to three of oxygen, 
and the latter one of base to three of oxygen. Thus the single 
datum given by analysis,-—the ratio, namely, of the oxygen to 
the metal in the two compounds,—proves sufficient, when aided 
by the principle of isomorphism, to lead us with certainty to the 
true atomic constitution of both compounds. 
Silica .—Silica affords another striking example of an oxide 
in which a knowledge of its isomorphous relations would re¬ 
move the obscurity attending the relative proportions of its two 
elements. It is the only oxide we possess of the metal silicium; 
and though we can determine, by analysis, that the metal and 
