r projects the entire flame, forming a horizontal, blue
cone of flame, which converges to a point at about an inch from the
wick, with a larger, longer, and more luminous flame enveloping it,
and terminating to a point beyond that of the blue flame.

To produce an efficient flame of oxidation, put the point of the
blowpipe into the flame about one third the diameter of the wick, and
about one twelfth of an inch above it. This, however, depends upon
the size of the flame used. Blow strong enough to keep the flame
straight and horizontal, using the largest orifice for the purpose.
Upon examining the flame thus produced, we will observe a long, blue
flame, _a b_, Fig. 3, which letters correspond with the same letters
in Fig. 2. But this flame has changed its form, and contains all the
combustible gases. It forms now a thin, blue cone, which converges to
a point about an inch from the wick. This point of the flame possesses
the highest intensity of temperature, for there the combustion of the
gases is the most complete. In the original flame, the hottest part
forms the external envelope, but here it is compressed more into a
point, forming the cone of the blue flame, and likewise an envelope of
flame surrounding the blue one, extending beyond it from _a_ to _c_,
and presenting a light bluish or brownish color. The external flame
has the highest temperature at _d_, but this decreases from _d_ to
_c_.

[Illustration: Fig. 3]

If there is a very high temperature, the oxidation is not effected so
readily in many cases, unless the substance is removed a little from
the flame; but if the heat be not too high, it is readily oxidized in
the flame, or near its cone. If the current of air is blown too
freely or violently into the flame, more air is forced there than is
sufficient to consume the gases. This superfluous air only acts
detrimentally, by cooling the flame.

In general the operation proceeds best when the substance is kept at a
dull red heat. The blue cone must be kept free from straggling rays of
the yellow or reduction flame. If the analysis be effected on
charcoal, the blast should not be too strong, as a part of the coal
would be converted into carbonic oxide, which would act
antagonistically to the oxidation. The oxidation flame requires a
steady current of air, for the purpose of keeping the blue cone
constantly of the same length. For the purpose of acquiring practice,
the following may be done: Melt a little m