Accordingly, I made the following experiment:

21. Fifth Experiment.—(a.) I pressed upon the bottom of the dish A (Fig. 2) a tough mass, of the thickness of two fingers, made of wax, resin, and turpentine metal together; in the middle I fastened a thick iron wire which reached to the middle of the flask B; upon the point of this wire C, I stuck a small wax candle, whose wick I had twisted together out of three slender threads. I then lighted the candle, and at the same time placed over it the inverted flask B, which I then pressed very deep into the mass. As soon as this was done, I filled the dish with water. After the flame was extinguished and everything had become quite cold, I opened the flask in the same position under the water, when 2 ounces of water entered; the flask held 160 ounces of water. Accordingly, there is wanting here so much air as occupies

the space of 2 ounces of water. Has this air been absorbed by the inflammable substance, or has the heat of the small flame driven it out even before I could press the flask into the tough mass? The latter seems to have taken place in this case, as I conclude from the following:—I took a small flask capable of holding 20 ounces of water; in this I caused a candle to burn as in the preceding; after everything had become cold, I opened this flask likewise under water, whereupon similarly nearly 2 ounces entered. Had the former 2 ounces measure of air been absorbed, then there should have been only 2 drachms measure absorbed in this experiment.

(b.) I repeated the preceding experiment with the large flask in exactly the same way, except that I employed spirit of wine in place of the candle. I fastened three iron wires, which were of equal length and reached up to the middle of the flask, into the tough mass which was firmly pressed on to the bottom of the dish. Upon these wires I laid a four-cornered plate of metal, and upon this I placed a small vessel into which spirit of wine was poured. I set fire to this and placed the flask over it. After cooling, I observed that 3 ounces measure of air had been driven out by the heat of the flame.

(c.) Upon the same stand I placed a few small glowing coals, and allowed then go out in the same way under the flask. I found after cooling that the heat of the coals had driven out three and a half ounces measure of air.

The experiments seem to prove that the transference of phlogiston to the air does not always diminish its bulk, which, however, the experiments mentioned in §§ 8.16 shew distinctly. But the following will shew that that portion of the air which unites with the inflammable substance, and is at the same time absorbed by it, is replaced by the newly formed aerial acid.

22. Sixth Experiment.—After the fire had gone out and everything had become cold in the experiments mentioned above (§ 21. a. b. c.), I poured into each flask 6 ounces of milk of lime (lime water which has in it more unslaked lime than the water can dissolve); I then placed my hand firmly on the mouth of the flask and swung it several times up and down; then I held the flask inverted under water and drew my hand a little to one side, so that a small orifice might be made. Water immediately rose into the flask. Then I shut the mouth again very tightly with my hand under water, and afterwards shook it several times up and down. I opened it again under water; this operation I repeated twice more until no more water would rise into the flask, or until no more aerial acid was present in it. I then perceived that in each experiment between 7 and 8 ounces of water rose into the flasks, consequently the nineteenth part of the air has been lost. This was indeed something, but since in the combustion of phosphorus (§ 17) nearly the third part of the air was lost, there must be another reason besides, why as much is not absorbed in this case also. It is known that one part of aerial acid mixed with 10 parts of ordinary air extinguishes fire; and there are here in addition, expanded by the heat of the flame and surrounding the latter, the watery vapours produced by the destruction of these oily substances. It is these two elastic fluids, separating themselves from such a flame, which present no small hindrance to the fire which would otherwise certainly burn much longer, especially since there is here no current of air by means of which they can be driven away from the flame. When the aerial acid is separated from this air by milk of lime, then a candle can burn in it again, although only for a very short time.

23. Seventh Experiment.—I placed upon the

stand (§ 21. b.) a small crucible which was filled with sulphur; I set fire to it and placed the flask over it. After the sulphur was extinguished and everything had become cold, I found that out of 160 parts of air, 2 parts were driven out of the flask by the heat of the flame. I next poured 6 ounces of clear lime water into the flask and dealt with it by shaking, as already explained, and observed that the sixth part of all the air had been lost in consequence of the combustion. The lime water was not in the least precipitated in this case, an indication that sulphur gives out no aerial acid during its combustion, but another substance somewhat resembling air; this is the volatile acid of sulphur, which occupies again the empty space produced by the union of the inflammable substance with air. It is not, as may be seen, a trifling circumstance that phlogiston, whether it separates itself from substances and enters into union with air, with or without a fiery motion, still in every case diminishes the air so considerably in its external bulk.

24. Experiments which prove that ordinary air, consisting of two kinds of elastic fluids, can be compounded again after these have been separated from each other by means of phlogiston.

I have already stated in § 16 that I was not able to find again the lost air. One might indeed object, that the lost air still remains in the residual air which can no more unite with phlogiston; for, since I have found that it is lighter than ordinary air, it might be believed that the phlogiston united with this air makes it lighter, as appears to be known already from other experiments. But since phlogiston is a substance, which always presupposes some weight, I much doubt whether such hypothesis has any foundation....

25. How often must not chemists have distilled the

fuming acid of nitre from oil of vitriol and nitre, when it is impossible that they should not have observed how this acid went over red in the beginning, white and colourless in the middle of the distillation, but at the end red again; and indeed so dark-red that one could not see through the receiver? It is to be noticed here that if the heat is permitted to increase too much at the end of the distillation, the whole mixture enters into such frothing that everything goes over into the receiver; and, what is of the greatest importance, a kind of air goes over during this frothing which deserves no small attention. If one takes for such distillation a very black oil of vitriol, not only does the acid go over at the beginning of a far darker red than when one takes a white oil of vitriol, but further, when one introduces a burning candle into the receiver after about an ounce has gone over, this goes out immediately. On the other hand, when one places a burning candle in the receiver filled with blood-red vapours, towards the end of the distillation when, as has been said,