made,
were simple expressions of empirical facts, independent of any theory;
but gradually it became more and more plain that these facts all
harmonize with the atomic theory of Dalton. So by common consent the
proportionate combining weights of the elements came to be known as
atomic weights--the name Dalton had given them from the first--and
the tangible conception of the chemical atom as a body of definite
constitution and weight gained steadily in favor.

From the outset the idea had had the utmost tangibility in the mind of
Dalton. He had all along represented the different atoms by geometrical
symbols--as a circle for oxygen, a circle enclosing a dot for hydrogen,
and the like--and had represented compounds by placing these symbols of
the elements in juxtaposition. Berzelius proposed to improve upon this
method by substituting for the geometrical symbol the initial of the
Latin name of the element represented--O for oxygen, H for hydrogen, and
so on--a numerical coefficient to follow the letter as an indication of
the number of atoms present in any given compound. This simple system
soon gained general acceptance, and with slight modifications it is
still universally employed. Every school-boy now is aware that H2O is
the chemical way of expressing the union of two atoms of hydrogen with
one of oxygen to form a molecule of water. But such a formula would have
had no meaning for the wisest chemist before the day of Berzelius.

The universal fame of the great Swedish authority served to give general
currency to his symbols and atomic weights, and the new point of view
thus developed led presently to two important discoveries which removed
the last lingering doubts as to the validity of the atomic theory. In
1819 two French physicists, Dulong and Petit, while experimenting with
heat, discovered that the specific heats of solids (that is to say, the
amount of heat required to raise the temperature of a given mass to a
given degree) vary inversely as their atomic weights. In the same year
Eilhard Mitscherlich, a German investigator, observed that compounds
having the same number of atoms to the molecule are disposed to form the
same angles of crystallization--a property which he called isomorphism.

Here, then, were two utterly novel and independent sets of empirical
facts which harmonize strangely with the supposition that substances are
composed of chemical atoms of a determinate weight. This surely could
not be c