It is very singular, indeed; but I think rather too powerful to be pleasing. And is not this sound accounted for? MRS. B. That the percussion of glass, by a rapid stream of gas, should produce a sound, is not extraordinary: but the sound here is so peculiar, that no other gas has a similar effect. Perhaps it is owing to a brisk vibratory motion of the glass, occasioned by the successive formation and condensation of small drops of water on the sides of the glass tube, and the air rushing in to replace the vacuum formed.* [Footnote *: This ingenious explanation was first suggested by Dr. Delarive. --See Journals of the Royal Institution, vol. i. p. 259.] CAROLINE. How very much this flame resembles the burning of a candle. MRS. B. The burning of a candle is produced by much the same means. A great deal of hydrogen is contained in candles, whether of tallow or wax. This hydrogen being converted into gas by the heat of the candle, combines with the oxygen of the atmosphere, and flame and water result from this combination. So that, in fact, the flame of a candle is owing to the combustion of hydrogen gas. An elevation of temperature, such as is produced by a lighted match or taper, is required to give the first impulse to the combustion; but afterwards it goes on of itself, because the candle finds a supply of caloric in the successive quantities of heat which results from the union of the two electricities given out by the gases during their combustion. But there are other circumstances connected with the combustion of candles and lamps, which I cannot explain to you till you are acquainted with _carbon_, which is one of their constituent parts. In general, however, whenever you see flame, you may infer that it is owing to the formation and burning of hydrogen gas*; for flame is the peculiar mode of burning hydrogen gas, which, with only one or two apparent exceptions, does not belong to any other combustible. [Footnote *: Or rather, _hydro-carbonat_, a gas composed of hydrogen and carbon, which will be noticed under the head _Carbon_.] EMILY. You astonish me! I understood that flame was the caloric produced by the union of the two electricities, in all combustions whatever? MRS. B. Your error proceeded from your vague and incorrect idea of flame; you have confounded it with light and caloric in general. Flame always implies caloric, since it is produced by the combustion of hydrogen gas; but all caloric does not imply flame. Many bodies burn with intense heat without producing flame. Coals, for instance, burn with flame until all the hydrogen which they contain is evaporated; but when they afterwards become red hot, much more caloric is disengaged than when they produce flame. CAROLINE. But the iron wire, which you burnt in oxygen gas, appeared to me to emit flame; yet, as it was a simple metal, it could contain no hydrogen? MRS. B. It produced a sparkling dazzling blaze of light, but no real flame.