erhaps the most curious effect of these timed
impulses ever described was that observed by a watchmaker, named
Ellicott, in the year 1741. He left two clocks leaning against the
same rail; one of them, which we may call A, was set going; the other,
B, not. Some time afterwards he found, to his surprise, that B was
ticking also. The pendulums being of the same length, the shocks
imparted by the ticking of A to the rail against which both clocks
rested were propagated to B, and were so timed as to set B going.
Other curious effects were at the same time observed. When, the
pendulums differed from each other a certain amount, set B going, but
the reaction of B stopped A. Then B set A going, and the re-action of
A stopped B. When the periods of oscillation were close to each
other, but still not quite alike, the clocks mutually controlled each
other, and by a kind of compromise they ticked in perfect unison.

But what has all this to do with our present subject? The varied
actions of the universe are all modes of motion; and the vibration of
a ray claims strict brotherhood with the vibrations of our pendulum.
Suppose aethereal waves striking upon atoms which oscillate in the
same periods as the waves, the motion of the waves will be absorbed by
the atoms; suppose we send our beam of white light through a sodium
flame, the atoms of that flame will be chiefly affected by those
undulations which are synchronous with their own periods of vibration.
There will be on the part of those particular rays a transference of
motion from the agitated aether to the atoms of the volatilised metal,
which, as already defined, is absorption.

The experiment justifying this conclusion is now for the first time to
be made before a public audience. I pass a beam through our two
prisms, and the spectrum spreads its colours upon the screen. Between
the lamp and the prism I interpose a snapdragon light. Alcohol and
water are here mixed with common salt, and the metal dish that holds
them is heated by a spirit-lamp. The vapour from the mixture ignites
and we have a monochromatic flame. Through this flame the beam from
the lamp is now passing; and observe the result upon the spectrum. You
see a shady band cut out of the yellow,--not very dark, but
sufficiently so to be seen by everybody present.

But let me exalt this effect. Placing in front of the electric lamp
the intense flame of a large Bunsen's burner, a platinum capsule
containi