s results seemed to disprove the hypothesis,
for the atomic weights of many elements differed from whole numbers by
more, it was thought, than the limits of error of the experiments. It
was noteworthy, however, that the confidence of Dumas was not shaken,
though he was led to modify the hypothesis, and, in accordance with
previous suggestions of Clark and of Marignac, to recognize as the
primordial element, not hydrogen itself, but an atom half the weight,
or even one-fourth the weight, of that of hydrogen, of which primordial
atom the hydrogen atom itself is compounded. But even in this modified
form the hypothesis found great opposition from experimental observers.

In 1864, however, a novel relation between the weights of the elements
and their other characteristics was called to the attention of chemists
by Professor John A. R. Newlands, of London, who had noticed that if the
elements are arranged serially in the numerical order of their atomic
weights, there is a curious recurrence of similar properties at
intervals of eight elements This so-called "law of octaves" attracted
little immediate attention, but the facts it connotes soon came under
the observation of other chemists, notably of Professors Gustav Hinrichs
in America, Dmitri Mendeleeff in Russia, and Lothar Meyer in Germany.
Mendeleeff gave the discovery fullest expression, explicating it in
1869, under the title of "the periodic law."

Though this early exposition of what has since been admitted to be a
most important discovery was very fully outlined, the generality of
chemists gave it little heed till a decade or so later, when three new
elements, gallium, scandium, and germanium, were discovered, which, on
being analyzed, were quite unexpectedly found to fit into three gaps
which Mendeleeff had left in his periodic scale. In effect the periodic
law had enabled Mendeleeff to predicate the existence of the new
elements years before they were discovered. Surely a system that leads
to such results is no mere vagary. So very soon the periodic law took
its place as one of the most important generalizations of chemical
science.

This law of periodicity was put forward as an expression of observed
relations independent of hypothesis; but of course the theoretical
bearings of these facts could not be overlooked. As Professor J. H.
Gladstone has said, it forces upon us "the conviction that the elements
are not separate bodies created without reference to one a