ted through
the glass before the switch was closed, was seen to rotate when the
magnetic field was produced by the flow of the current. A similar
rotation is now familiar in the well-known tests of sugars--laevulose
and dextrose--which rotate plane-polarized light to left and right,
respectively.

But in this first discovery of a relationship between light and
magnetism Faraday had not taken the more important step that he
coveted--to determine whether the vibration period of a light-emitting
particle is subject to change in a magnetic field. He attempted
this in 1862--the last experiment of his life. A sodium flame was
placed between the poles of a magnet, and the yellow lines were
watched in a spectroscope when the magnet was excited. No change
could be detected, and none was found by subsequent investigators
until Zeeman, of Leiden, with more powerful instruments made his
famous discovery, the twenty-fifth anniversary of which has recently
been celebrated.

[Illustration: Fig. 32. Pasadena Laboratory of the Mount Wilson
Observatory.

Showing the large magnet (on the left) and the spectroscopes used
for the study of the effect of magnetism on radiation. A single line
in the spectrum is split by the magnetic field into from three to
twenty-one components, as illustrated in Fig. 34. The corresponding
lines in the spectra of sun-spots are split up in precisely the
same way, thus indicating the presence of powerful magnetic fields
in the sun.]

His method of procedure was similar to Faraday's, but his magnet and
spectroscope were much more powerful, and a theory due to Lorentz,
predicting the nature of the change to be expected, was available
as a check on his results. When the current was applied the lines
were seen to widen. In a still more powerful magnetic field each
of them split into two components (when the observation was made
along the lines of force), and the light of the components of each
line was found to be circularly polarized in opposite directions.
Strictly in harmony with Lorentz's theory, this splitting and
polarization proved the presence in the luminous vapor of exactly such
negatively charged electrons as had been indicated there previously
by very different experimental methods.

In 1908 great cyclonic storms, or vortices, were discovered at
the Mount Wilson Observatory centring in sun-spots. Such whirling
masses of hot vapors, inferred from Sir Joseph Thomson's results
to contain electrica