Dephlogijlicated Air. 241

was cool, it appeared that no air had beenproduced. The mercury rose and filledall the interstices of the sublimate. 1 thenput two ounces of this substance into agreen glass retort, which will bear a greaterdegree of heat than flint glass; and by de-grees covered it with live coals, but all thatfollowed was the sublimation of the matterinto the neck of the retort, and no pro-duction of air. Even the common air, thatcame over first, was not at all altered.

I was not able to make much more ofcommon salt. P'rom an ounce of it, in asstrong a red heat as I could give it in a glassretort with a long neck, and in sand (inwhich it may be made to bear more heatthan when surrounded with live coals) Igot about two ounce measures of air, thefirst part of which was fixed air, and thelast phlogisticated air, extinguishing acandle, and not affected by nitrous air. Ihave, however, at various times, repeatedthis experiment, and once with the heat ofa smiths fire; but getting little or no air, Irather suspect that the phlogisticated air inR the