molecule at once suggests a new question: How can an aggregation of
atoms, having all their affinities satisfied, take any further part in
chemical reactions? Seemingly such a molecule, whatever its physical
properties, must be chemically inert, incapable of any atomic
readjustments. And so in point of fact it is, so long as its component
atoms cling to one another unremittingly. But this, it appears, is
precisely what the atoms are little prone to do. It seems that they are
fickle to the last degree in their individual attachments, and are as
prone to break away from bondage as they are to enter into it. Thus the
oxygen atom which has just flung itself into the circuit of two
hydrogen atoms, the next moment flings itself free again and seeks
new companions. It is for all the world like the incessant change
of partners in a rollicking dance. This incessant dissolution and
reformation of molecules in a substance which as a whole remains
apparently unchanged was first fully appreciated by Ste.-Claire Deville,
and by him named dissociation. It is a process which goes on much more
actively in some compounds than in others, and very much more actively
under some physical conditions (such as increase of temperature) than
under others. But apparently no substances at ordinary temperatures,
and no temperature above the absolute zero, are absolutely free from its
disturbing influence. Hence it is that molecules having all the
valency of their atoms fully satisfied do not lose their chemical
activity--since each atom is momentarily free in the exchange of
partners, and may seize upon different atoms from its former partners,
if those it prefers are at hand.

While, however, an appreciation of this ceaseless activity of the atom
is essential to a proper understanding of its chemical efficiency,
yet from another point of view the "saturated" molecule--that is, the
molecule whose atoms have their valency all satisfied--may be thought of
as a relatively fixed or stable organism. Even though it may presently
be torn down, it is for the time being a completed structure; and a
consideration of the valency of its atoms gives the best clew that has
hitherto been obtainable as to the character of its architecture.
How important this matter of architecture of the molecule--of space
relations of the atoms--may be--was demonstrated as long ago as 1823,
when Liebig and Wohler proved, to the utter bewilderment of the
chemical world, that two su