I pointed out that, in an atmospheric burner, it was not the oxygen of
the air introduced combining with and burning up the hydrocarbons, and
so preventing the separation of incandescent carbon, which gave the
non-luminous flame, but the diluting action of the nitrogen, which
acted by increasing the temperature at which the hydrocarbons are
broken up, and carbon liberated, a fact which was proved by
observation that heating the mixture of gas and air again restored the
luminosity of the flame. This experiment clearly shows that
temperature is a most important factor in the illuminating value of a
flame, and this is still further shown by a study of the action of the
diluents present in coal gas, the non-combustible ones being far more
deleterious than the combustible, as they not only dilute, but
withdraw heat.

Anything which will increase the temperature of the flame will also
increase the illuminating power, provided, of course, that the
increase in temperature is not obtained at the expense of the too
rapid combustion of the hydrocarbons.

As has been shown in the experiments relating to the action of
diluents on flame, already quoted, oxygen, when added to coal gas,
increases its illuminating value to a marked and increasing degree,
until a certain percentage has been added, after which the
illuminating power is rapidly decreased, until the point is reached
when the mixture becomes explosive. This is due to the fact that the
added oxygen increases the temperature of the flame by doing the work
of the air, but without the cooling and diluting action of the
nitrogen; when, however, a certain proportion is added, it begins to
burn up the heavy hydrocarbons, and although the temperature goes on
increasing, the light-giving power is rapidly diminished by the