- 
2 MITSCHERLICH ON CHEMICAL REACTIONS 
to it, the sugar, without combining with the sulphuric acid, fixes 
some water and is converted into glucose. The decomposition 
of ammoniacal gas by incandescent copper is one of the few ex- 
amples in which decomposition of a gaseous body is induced 
by a solid body; but there are on the other hand numerous ex- 
amples of this class of decompositions between liquid and solid 
bodies which themselves undergo no change in the reaction ; 
such for instance are the decomposition of the peroxide of hy- 
drogen by various bodies, the decomposition of chlorate of potash 
by the oxide of copper and other fixt bases. 
Before it is possible to ascertain why certain chemical reac- 
tions are effected under the influence of bodies which act only 
by their presence, we should seek to form a clear notion of the 
manner in which bodies that do not combine chemically behave 
when brought into close contact. The attraction which a solid 
exercises on a gaseous body may be shown with the greatest 
facility, by employing the former in such a state that it offers the 
greatest possible surface in the smallest volume, in extremely 
thin leaves, or, better still, in impalpable powder. Carbon and 
several other substances scarcely fusible, such as platinum, which 
may be obtained in a state of extreme division, are well suited 
for such inquiries. 
In the first volume of my ‘ Manual of Chemistry,’ I have caleu- 
lated the surface which would be represented by a cubic inch 
divided into small cubes of z;1,5th of an inch laterally by two 
series of perpendicular sections. It equals, neglecting the size 
of the cells, 100 square feet. If any substance be prepared in 
such a state of fineness that its division may be supposed to have 
stopt short only at its atom, or at least at a degree the approxi- 
mation of which is known, the surface which it would represent 
might still be calculated. Mathematics can-furnish the value of 
the greatest diameter of an atom of a chemical compound, when 
this can be obtained in the state of thin laminz or of vesicles; 
thus the diameter of an atom of water must be at the furthest 
10.d00,000th of an inch, from the colour which the thinnest film 
of a soap bubble presents. 
By reducing a dilute solution of chloride of platinum by means 
of carbonate of soda and formic or tartaric acid, or by decom- 
posing a very weak solution of the sulphate of platinum by weak 
alcohol, the metallic platinum obtained in both cases must be in 
the molecular state, since the molecules at the moment of the 
