be desired that the experiments should be repeated at still higher
temperatures and with more exact instruments, in order to determine
the limit of temperature at which heated gases undoubtedly become
self-incandescent. The fact, however, that gases, at a temperature of
more than 1,500 deg. C, are not yet luminous, proves that the incandescence
of the flame is not to be explained as a self-incandescence of the
products of combustion. This is confirmed by the circumstance that, with
rapid mixture of the burning gases, the flame becomes shorter because
the combustion process goes on more quickly, and hotter because less
cold air has access. Further, the flame also becomes shorter and hotter
if the gases are strongly heated previous to combustion. As the rising
products of combustion still retain for a time the temperature of the
flame, the reverse must occur if the gases were self-luminous. The
luminosity of the flame, however, ceases at a sharp line of demarkation,
and evidently coincides with completion of the chemical action. The
latter, itself, therefore, and not the heating of the combustion
products, which is due to it, must be the cause of the luminosity. If
we suppose that the gas-molecules are surrounded by an ether-envelope,
then, in chemical combination of two or several such molecules, there
must occur a changed position of the ether-envelopes. The motion of
ether-particles thus caused may be represented by vibrations, which form
the starting-point of light and heat-waves.

In quite a similar manner we may also, according to Dr. Siemens,
represent the light-phenomenon occurring when an electric current
is sent through gases, which always takes place when the maximum of
polarization belonging to them is exceeded. As the passage of the
current through the gas seems to be always connected with chemical
action, the phenomenon of glow may be explained in the same way as in
flame, by oscillating transposition of the ether envelopes, by which the