"non-luminous." This effect can not be caused by cooling, for the greater supply of oxygen makes the flame hotter, nor by dilution, for nothing is diluted. Evidently it must be caused by the greater supply of oxygen. The air contains four volumes of nitrogen-gas for every one volume of oxygen. When a substance burns in air, the particles of oxygen are hindered in getting at the combustible by the particles of nitrogen, which are four times as numerous. When the substance burns in pure oxygen there is no trouble of this sort, and the combustion takes place much more rapidly and energetically. Such is the case in our experiment. The oxygen-particles are no longer hindered in getting at the carbon by the inert nitrogen, and therefore seize on it so promptly as to leave none to make the flame luminous. If we allow the carbonic-acid gas, resulting from the burning, to accumulate in the jar, this gas will perform the function of the nitrogen of the air. At first a very bright spot appears in the center of the flame where the least oxygen penetrates. It is brighter than a gas-flame in air, because the more rapid combustion makes the flame hotter, and consequently the little particles of carbon glow more intensely. As the carbonic acid gas accumulates, it becomes more and more difficult for the oxygen to get at the flame. When four fifths of the gas in the jar are carbonic acid, the flame burns as it does in the air. As the carbonic acid still goes on accumulating we get another effect, viz., that of dilution, and the flame becomes "non-luminous" again. In this experiment are illustrated all three of the causes which render luminous gas-flames "non-luminous." At first we observe that the combustion is so rapid that no carbon is separated. At the close of the experiment we see the flame become "non-luminous" again on account of the dilution of the air with inert gas and the attendant cooling of the flame.
If we try this same experiment with a flame which owes its luminosity to incombustible solid particles, or to vapors, we shall get only the effect of the greater heat of the flame. For example, hydrogen-gas containing a small proportion of the vapor of chromyl chloride burns with a luminous flame, the light being due to the separation in the flame of particles of solid oxide of chromium. If this flame be placed in pure oxygen it becomes of an almost dazzling brilliancy. The particles of oxide of chromium are incombustible, and therefore the greater heat of the flame only makes them glow the more brightly. This experiment shows that a flame becomes hotter in pure oxygen than in the diluted oxygen of the air, and furnishes an indirect proof that the reason why a gas-flame is made "non-luminous" in oxygen is because its carbon is burned up so quickly.
We distinguish, then, two effects produced on a luminous gas-flame by pure oxygen. First, it makes the flame hotter, and consequently tends to make it brighter; and, second, if the supply of oxygen is relatively large, it burns up the carbon of an ordinary flame at once, and thus renders the flame "non-luminous."
