It goes without saying that in one case quick and accurate observation,
a retentive memory, and the association of causes and effects follow,
and that in the other they do not.

I advocate, therefore, the teaching of physical science in our
schools--_in all our schools_. Physical science taught by the
experimental method.

An experiment has been defined as a question put to Nature, a question
asked in _things_ rather than in _words_, and so conditioned that no
uncertain answer can be given. Nature says that all matter gravitates,
not in words, but in the swing of planets around the sun, and in the
leap of the avalanche. And men have devised ingenious machines through
which Nature may tell us the invariable laws of gravitation, and give
some hint as to why it is true.

There are two kinds of experiments, and two corresponding kinds of
investigators.

I. In original investigation there are the following elements:

1. The careful determination of all the conditions under which the
experiment takes place.

2. The observation of exactly what happens, with a painstaking
elimination of all previous notions as to what ought to happen.

3. The change of conditions, one at a time, with a comparison of the
results obtained with the changes made, in order to determine that each
condition has been given just its appropriate weight in the experiment.