Experiments Relating to Phlogiston, and the Seeming Conversion of Water into Air. By Joseph Priestley, LL. D. F. R. S.; Communicated by Sir Joseph Banks, Bart. P. R. S.

XXII. Experiments relating to Phlogiston, and the seeming Conversion of Water into Air. By Joseph Priestley, LL.D., F.R.S.; communicated by Sir Joseph Banks, Bart., P.R.S. Read June 26, 1783. TO SIR JOSEPH BANKS, BART., P.R.S. DEAR SIR, At the persuasion of my friends, I beg you would lay before the Royal Society my late observations on Phlogiston, and also on the seeming Conversion of water into air, though I have by no means done all that I have in view with respect to these subjects. The principal facts are, I think, sufficiently ascertained, though I do not presume to give any opinion with respect to the theory of them. I am, with the greatest respect, &c. Birmingham, April 21, 1783. Experiments relating to Phlogiston. There are few subjects, perhaps none, that have occasioned more perplexity to chemists than that of phlogiston, or, as it is sometimes called, the principle of inflammability. It was the great discovery of Stahl, that this principle, whatever it be, is transferrable from one substance to another, how different soever in their other properties, such as sulphur, wood, and all the metals, and therefore is the same thing in them all. But what has given an air of mystery to this subject, has been that it was imagined, that this principle, or substance, could not be exhibited except in combination with other substances, and could not be made to assume separately either a fluid or solid form. It was also asserted by some, that phlogiston was so far from adding to the weight of bodies, that the addition of it made them really lighter than they were before; on which account they chose to call it the principle of levity. This opinion had great patrons. Of late it has been the opinion of many celebrated chemists, Mr. Lavoisier among others, that the whole doctrine of phlogiston had been founded on mistake, and that in all cases in which it was thought that bodies parted with the principle of phlogiston, they in fact lost nothing, but on the contrary, acquired something; and in most cases an addition of some kind of air; that a metal, for instance, was not a combination of two things, viz. an earth and phlogiston, but was probably a simple substance in its metallic state; and that the calx is produced not by the loss of phlogiston, or of any thing else, but by the acquisition of air. The arguments in favour of this opinion, especially those which are drawn from the experiments of Mr. Lavoisier made on mercury, are so specious, that I own I was myself much inclined to adopt it. My friend Mr. Kirwan, indeed, always held that phlogiston was the same thing with inflammable air; and he has sufficiently proved this from many experiments and observations, my own as well as those of others. I did not, however, accede to it till I discovered it by direct experiments, made with general and indeterminate views, in order to ascertain something concerning a subject which had given myself and others so much trouble. I began with repeating the experiments in which I had found that inflammable air, made red-hot in flint glass tubes, gave them a black tinge, and was in a great measure absorbed, which I had discovered to be owing to the calx of lead in the glass attracting phlogiston from the inflammable air. As the quantity of air in these tubes was very small, though I gave it as my opinion, that the residuum in one of the processes was phlogisticated air, because I perceived no marks of ascension on presenting to it the flame of a small candle; I was not, on recollection, satisfied with this conclusion, and was desirous of repeating the experiment with more care, especially as, in one of the above-mentioned experiments, I found only a very small bubble of the inflammable air in the tube in which it had been heated. I found, however, great difficulties in repeating these experiments; and the quantity of inflammable air operated upon in them is necessarily so small, that the result is always liable to much uncertainty. I thought, therefore, that throwing the focus of a burning lens upon a quantity of pounded flint glass, surrounded with inflammable air, or rather on the calx of lead alone, in the same circumstances, would be a much easier experiment, and might bring me nearer to my object; and on making the experiment it immediately answered far beyond my expectation. For this purpose, I put upon a piece of a broken crucible (which could yield no air) a quantity of minium, out of which all air had been extracted; and placing it upon a convenient stand, introduced it into a large receiver, filled with inflammable air, confined by water. As soon as the minium was dry, by means of the heat thrown upon it, I observed that it became black, and then ran in the form of perfect lead, at the same time that the air diminished at a great rate, the water ascending within the receiver. I viewed this process with the most eager and pleasing expectation of the result, having at that time no fixed opinion on the subject; and therefore I could not tell, except by actual trial, whether the air was decomposing in the process, so that some other kind of air would be left, or whether it would be absorbed in toto. The former I thought the more probable, as if there was any such thing as phlogiston, inflammable air, I imagined, consisted of it, and something else. However, I was then satisfied that it would be in my power to determine, in a very satisfactory manner, whether the phlogiston in inflammable air had any base or not, and if it had, what that base was. For seeing the metal to be actually revived, and that in a considerable quantity, at the same time that the air was diminished, I could not doubt but that the calx was actually imbibing something from the air; and from its effects in making the calx into metal, it could be no other than that to which chemists had unanimously given the name of phlogiston. Before this first experiment was concluded, I perceived, that if the phlogiston in inflammable air had any base, it must be very inconsiderable: for the process went on till there was no more room to operate without endangering the receiver; and examining, with much anxiety, the air that remained, I found that it could not be distinguished from that in which I began the experiment, which was air extracted from iron by oil of vitriol. I was therefore pretty well satisfied that this inflammable air could not contain anything besides phlogiston; for at that time I reduced about 45 ounce measures of the air to five. In order to ascertain a fact of so much importance with the greatest care, I afterwards carefully expelled from a quantity of minium all the phlogiston, and every thing else that could have assumed the form of air, by giving it a red heat when mixed with spirit of nitre; and immediately using it in the manner mentioned above, I reduced 101 ounce measures of inflammable air to two. To judge of its degree of inflammability, I presented the flame of a small candle to the mouth of a phial filled with it, and observed, that it made thirteen separate explosions, though weak ones (stopping the mouth of the phial with my finger after each explosion), when fresh made inflammable air, in the same circumstances, made only fourteen explosions, though stronger ones. After this experiment I could not hesitate to conclude, that this inflammable air went totally, and without decomposition, into the lead which I formed at that time; and if the necessary circumstances of the experiment be considered, it will be thought extraordinary that, even admitting this, the result should be so decisively clear in favour of it: for, in the first place, the greatest care must be used to expel all air from the minium, and it must be used before it can have attracted any from the atmosphere; and in the next place, the water also (a considerable quantity of which must be used, and which will also be heated in the process) should be made as free from air as possible. In these circumstances, had I found the small residuum, of 2 ounce measures from 101, to have been phlogisticated or fixed air, I should not have been disappointed; and it would not have prevented my concluding that phlogiston was the same thing with inflammable air, contained in a combined state in metals, just as fixed air is contained in chalk and other calcareous substances; both being equally capable of being expelled again in the form of air. Afterwards using a calx of lead, which had been prepared in the same manner with the former, but which had remained some weeks exposed to the air, I found, that when by using it I had reduced 150 ounce measures of inflammable air to 10, this residuum was phlogisticated air. But examining this calx separately, I found that it gave, by heat in a glass vessel, a considerable quantity of phlogisticated air. I must observe, that the minium should not be reduced to a perfectly compact glass of lead; for then it would be too refractory to be easily revived by this process. Making use of some of it, I found that I could only melt it; but that a copious black fume came from it, and coated the inside of the receiver: an experiment which I shall repeat and re-consider. I must also observe, that the lead which I procured in the above-mentioned process was not to be distinguished from any other lead, and that the inflammable air was all procured from iron by oil of vitriol. When I made use of inflammable air from wood, I found, that though I was able to reduce minium with it, it was effected with more time and difficulty. Forty ounce measures of this kind of inflammable air I reduced to 25; when I found that the heat of the lens produced only glass of lead, and no metal. The air was still, however, inflammable; and there was a small mixture of fixed air in it. This kind of inflammable air, which burns with a lambent flame, I have some reason to think, consists of an intimate union of fixed air with that which is of the explosive kind extracted from metals. The result of those experiments which I made with that kind of inflammable air which is collected in the process for making phosphorus, and which burns with a lambent yellow flame, was similar to those which I made with inflammable air from wood, which burns with a lambent white flame. Having had this remarkable result with inflammable air, I immediately tried all the other kinds of air in the same manner; but in none of them did I procure any thing from the minium besides glass of lead, except in alkaline air, and vitriolic acid air. In fixed air, nitrous air, phlogisticated air, marine acid air, fluor acid air, as also in common and dephlogisticated air, I got no metal at all. In vitriolic acid air there was but a small quantity of lead produced, and I have observed that this kind of air imparts a certain portion of phlogiston to common air, rendering it in some measure phlogisticated, though by no means in so great a degree as nitrous air. Though nitrous air and phlogisticated air certainly contain phlogiston, they appear by these experiments to hold it too obstinately to part with it to minium in this process, though nitrous air quits it so readily to respirable air. I would observe, that there were some peculiar appearances in the experiments I made to revive the calx of lead in these kinds of air in which the attempt did not not succeed; but I must repeat the experiments, and note the appearances more accurately, before I report them. In alkaline air lead seems to be formed from the minium as readily as in inflammable air, and indeed I thought rather more so; and this is a remarkable confirmation and illustration of these experiments, in which, by taking the electric spark in a quantity of alkaline air, I converted it into three times as much pure inflammable air; an experiment which, on account of the extraordinary nature of it, I have repeated many times since I first published the account of it, and always with the same result. This experiment also throws some light upon those in which, by super-phlogisticating iron with nitrous air, I produced a strong smell of volatile alkali; an experiment which I have also frequently repeated with the same result. The reviving of lead in alkaline air may also help us to conceive how all acids should have an affinity both to phlogiston and to alkalies, which have hitherto appeared to be things so very different from each other; since, from these experiments, it is probable that one of them is some modification of the other, or a combination of something else with the other. To trace the connection between the alkaline and inflammable principles, is a curious subject; and from these hints it may, perhaps, not be very difficult to prosecute it to advantage. It is evident, however, from the following experiments, that alkaline air is the compound and inflammable air, or phlogiston, the more simple substance of the two. From $5\frac{1}{2}$ ounce measures of alkaline air I got, by means of litharge, 17 grains of lead, besides some that was dissolved in the mercury, by which the air was confined. There remained $2\frac{1}{2}$ ounce measures, which appeared to be phlogisticated air, and Dr. Priestley's Experiments relating to Phlogiston, and to have no fixed air in it. At another time, in eight ounce measures of alkaline air I got 15 grains of lead, besides what was dissolved in the mercury, which seemed to be a good deal in proportion to it. It was observable, that there remained in this process $3\frac{1}{2}$ ounce measures of phlogisticated without any mixture of fixed air in it, though the masticot which I used at this time gave by heat only a good deal of pretty pure fixed air. These experiments with alkaline air well deserve to be resumed, and I shall not fail to do it at a proper opportunity. Having thus produced lead in inflammable air, I proceeded in my attempts to revive other metals from their calces by the same means; and I succeeded very well with tin, bismuth, and silver; tolerably well with copper, iron, and regulus of cobalt; but not at all with regulus of antimony, regulus of arsenic, zinc, or the metal of manganese. I was desirous also of ascertaining by this means the quantity of phlogiston that enters into the composition of the several metals; but in this I found more difficulty than I had expected: and this arose chiefly from the allowance that was to be made for the inflammable air which entered into that part of the calx which was only partially revived; and it was not easy to revive the whole of any quantity of calx completely. After many trials, I think I may venture to say, that an ounce of lead absorbs 100 ounce measures of inflammable air, or perhaps something more; for in one result it seemed to have imbibed in the proportion of 108 ounce measures. An ounce of tin absorbs inflammable air in the proportion of 377 ounce measures to the ounce. An ounce of copper from verditer absorbed 403 ounce measures; from a solution of blue vitriol, precipitated by salt of tartar, and afterwards made red-hot with spirit of nitre, 640; but from blue vitriol itself 909 ounce ounce measures. In this case, however, much of the inflammable air went to the formation of the vitriolic acid air, the smell of which was very perceivable in the course of the experiment. The copper that I made in this way was brittle, and therefore seemed not to be perfectly metalized; but being fuzed with borax it became perfect copper, and, as I think, without any loss of weight. Bismuth absorbed inflammable air in the proportion of 185 ounce measures to the ounce. The calx I used was a precipitate from the solution of this metal in spirit of nitre. Iron I got from a precipitate of a solution of green vitriol by salt of tartar, moistened with spirit of nitre, and exposed to a red heat. This calx absorbed in the proportion of 890 ounce measures of the inflammable air to an ounce of iron, which was in the form of a black powder; but to all appearance as much attracted by the magnet as iron filings. But it could not be expected, that perfect iron, containing its full proportion of phlogiston, should be produced in this manner, since inflammable air may be expelled from perfect iron in this very process. Silver I evidently revived from a solution of it in spirit of nitre precipitated by salt of tartar, and also from luna cornea. A quantity of this last substance absorbed 23 ounce measures of inflammable air; but I could not get any calx of silver free from small grains of the perfect metal, which was easily discovered by a magnifier, and therefore I could not ascertain the quantity of inflammable air absorbed by it. Small grains of regulus cobalt I produced from zaffre, and inflammable air was absorbed; but I did not estimate the quantity. A quantity of manganese absorbed 7 ounce measures of inflammable air; but I could not perceive anything in it which had the appearance of metal. But I imagined I had not heat enough for the purpose, and mixing with it some calcined borax, I repeated the experiment, when there was again an evident absorption of air, and in the course of that experiment, I once thought that I did perceive a small globule of metal. Zinc and arsenic were only sublimed in this process. The same was the case with the glass of antimony; but the experiment was attended with this peculiar circumstance, that when the glass was melted in inflammable air it formed itself into needle-like crystals arranged in a very curious manner, though I could not produce that appearance in other kinds of air. Inflammable air being clearly imbibed by the calces of metals, and thereby reviving them, is a sufficient proof of its containing what has been called phlogiston; and its being absorbed by them in toto, without decomposition, is a proof of its being nothing besides phlogiston in the form of air, unless there should be something solid deposited from it at the same time that the proper phlogistic part of it was absorbed. With respect to this, I can only say that, in the course of the experiments, I did not perceive any thing of the kind: for though in some of the processes there was a black smoke produced, in others I could perceive nothing but part of the calx subliming, and clouding the glass. On this account, however, I could not pretend to ascertain the weight of the inflammable air in the calx, so as to prove that it had acquired an addition of weight by being metallized, which I often attempted. But were it possible to procure a perfect calx, no part of which should be sublimed and dispersed, by the heat necessarily to be made use of in the process, I should not doubt but that the quantity quantity of inflammable air imbibed by it would sufficiently add to its weight. Besides the formation of metals from their calces, I had other proofs, and of a nature sufficiently curious, of inflammable air containing phlogiston, though perhaps not sufficiently conclusive with respect to its being wholly and simply phlogiston itself. Thus, by means of it, I was able to make phosphorus, nitrous air, liver of sulphur, and sulphur itself, in all of which phlogiston is acknowledged to be a principal ingredient. Throwing the focus of the lens upon a quantity of that glassy matter which is made from calcined bones by oil of vitriol in inflammable air, some of it was absorbed, and all the inside of the receiver was covered with an orange-coloured substance, which had a strong smell of phosphorus. I then wanted sun-shine to continue the experiment; but I was satisfied that there was sufficient proof of phosphorus being actually formed in this manner. With alkaline air I succeeded much better. In 2½ ounce measures of this air I produced, from the glassy matter mentioned above, 2 grains of phosphorus in one mass, the vessel being only filled with white fumes during the process. One-fourth of the bulk of the air remained, and this was inflammable, burning with a yellow lambent flame, exactly like that which is produced in the process for making phosphorus. That nitrous air contains phlogiston is sufficiently evident, if there be any such thing as phlogiston: and I have farther proved, that it contains as much phlogiston, in proportion to its bulk, as inflammable air itself. I have now, however, the farther satisfaction to be able to make nitrous air from its two constituent principles, viz. nitrous vapour and inflammable air. The most easy process for this purpose is, to throw a stream Dr. Priestley's Experiments relating to Phlogiston, stream of nitrous vapour into a large phial previously filled with inflammable air. In this manner nitrous air is instantly formed, and in great quantities; but as this nitrous vapour is produced by the rapid solution of bismuth in spirit of nitre, which at the same time produces a quantity of nitrous air, the experiment is not quite unexceptionable. I therefore attempted the same thing in the following manner. Taking a quantity of what I have called a nitrated calx of lead, which I first produced by uniting nitrous vapour to muriatic acid (in consequence of which, from being a red and powdery substance, it became white, compact, and brittle), I placed it upon a stand, in a receiver filled with inflammable air, and throwing the focus of the lens upon it, there was a diminution of the inflammable air, which amounted to about two-thirds of the whole, and during this time lead was revived from the calx. After this there was no more diminution of the air, or revival of the calx: and then examining what remained of the air, I found it to be all strongly nitrous: and, from the circumstances in which it was produced, it must have been formed from the nitrous vapour contained in the calx, and the inflammable air in the receiver. In order to ascertain the purity of this nitrous air, I mixed it with an equal quantity of common air, and found that they occupied the space of 1,32 measures. Fresh nitrous air, made in the usual way, and mixed with common air in the same proportion, occupied the space of 1,26. This difference arose not from any impurity in the nitrous air, but from the mixture of the dephlogisticated air, which is also expelled from this calx by heat. Liver of sulphur was procured by throwing the focus of the lens upon vitriolated tartar in inflammable air, and it appeared to be perfectly well formed. Lastly, Lastly, to produce sulphur, I threw the focus of the lens on a quantity of oil of vitriol, contained in an hollow earthen vessel, and evaporated it to dryness in a receiver filled with inflammable air, in consequence of which the inside of the receiver acquired a whitish incrustation, which when warmed had a strong smell of sulphur; and repeating the process in the same receiver, I was able, this second time, to scrape off enough of the matter to put on a piece of hot iron, and to produce the genuine blue flame, as well as the peculiar smell of sulphur. I shall conclude these observations on phlogiston with two articles; one of which seems to contradict an established maxim among chemists; and the other a former opinion of my own. It is generally said, that charcoal is indestructible, except by a red heat in contact with air. But I find that it is perfectly destructible, or decomposed, in vacuo, and by the heat of a burning lens almost wholly converted into inflammable air; so that nothing remains besides an exceedingly small quantity of white ashes, which are seldom visible, except when, in very small particles, they happen to cross the sun-beam, as they fly about within the receiver. It would be impossible to collect or weigh them; but, according to appearance, the ashes thus produced from many pounds of wood could not be supposed to weigh a grain. The great weight of ashes produced by burning wood in the open air arises from what is attracted by them from the air. The air which I get in this manner is wholly inflammable, without the least particle of fixed air in it. But, in order to this, the charcoal must be perfectly well made, or with such a heat as would expel all the fixed air which the wood contains; and it must be continued till it yield inflammable air only, which, in an earthen retort, is soon produced. Wood, Wood, or charcoal, is even perfectly destructible, that is, resolvable into inflammable air, in a good earthen retort, and a fire that would about melt iron. In these circumstances, after all the fixed air had come over, I have several times continued the process during a whole day, in all which time inflammable air has been produced equally, and without any appearance of a termination. Nor did I wonder at this, after seeing it wholly vanish into inflammable air \textit{in vacuo}. A quantity of charcoal made from oak, and weighing about an ounce, generally gave me about five ounce measures of inflammable air in twelve minutes. The second article that I shall now mention affords an indisputable proof of the generation of fixed air from dephlogisticated air and phlogiston, or inflammable air. I have several times given it as my opinion, that fixed air is a \textit{factitious substance}, and a modification of the nitrous and vitriolic acids, my former experiments greatly favouring that conclusion; but that it was composed of dephlogisticated air and phlogiston, though maintained by my friend Mr. Kirwan, I was far from being satisfied with, till I was forced to consent to his proof of it from my own former experiments, and gave him leave to mention it, as he has done in his late excellent paper on salts. But I have lately had two direct proofs of it by experiment. The first was when, in repeating a beautiful experiment first made by Dr. Ingen-Housz, but with some variation, I was firing some shavings of iron in dephlogisticated air confined by mercury, by means of a burning lens. In this way I quickly fired the iron, and it burned away in a very pleasing manner. But what struck me most was, that, of the air that remained, a considerable portion was fixed air, though in the receiver I had nothing but the purest dephlogisticated air, together with the iron, iron, which could only give inflammable air. I would observe, that the melted iron formed itself into large balls, which appeared to be a mere flag or glass, and was no longer iron. Afterwards, to put this hypothesis concerning the constituent principles of fixed air to a more direct proof, I mixed iron filings, which gave only inflammable air, with red precipitate, which I found to give nothing but the purest dephlogisticated; and when I heated them in a coated glass retort, they gave a great quantity of fixed air, in some portions of which nineteen-twentieths were absorbed by lime-water; but the residuum was inflammable. However, when I mixed with iron filings a quantity of powdered charcoal, which I had found to give only inflammable air, the fixed air produced from it was so pure, that only one-fortieth part of it remained unabsorbed by water; so that this fixed air was as pure as that which is generally procured from chalk by oil of vitriol. It appeared, in some of these experiments, that three ounce measures of dephlogisticated air go into the composition of two ounce measures of fixed air. For one ounce of this red precipitate gave 60 ounce measures of dephlogistated air; and when mixed with two ounces of iron filings, it gave about 40 ounce measures of fixed air that were actually absorbed by water, besides a residuum that was inflammable. I had the same proportion when I used half an ounce of each of the materials. But when I used one ounce of each, I got only 20 ounce measures of fixed air, including the residuum. At other times I had different proportions with different quantities of iron filings and charcoal. I cannot conclude these observations without taking notice, how very valuable an instrument in philosophy is a good burning lens. This must have been perceived in many of my former Dr. Priestley's Experiments relating to Phlogiston, former experiments, but more especially in these. By no other means can heat be given to substances in vacuo, or in any other kind of air besides atmospherical; and without some method of doing this, no such experiments as these can possibly be made. I therefore congratulate all the lovers of science on the successful attempt of Mr. Parker to execute so capital an instrument as he has done of this kind. Such spirited and generous exertions reflect honour on himself and on our country. It is only to be wished, that we could have lenses of a smaller size (viz. from 12 to 18 inches diameter) made tolerably cheap, so that they might be in more common use. All my experiments were made with one of 12 inches in diameter. Experiments relating to the seeming Conversion of Water into Air. SINCE many persons have expressed a wish to be acquainted with the experiments I have lately made, which at first seemed to favour the idea of a conversion of water into air, but which terminated in the discovery of a fact, in my opinion, still more extraordinary, I shall submit to the Royal Society the result of the observations I have already made; though, as yet, I have by no means been able to satisfy myself so fully as I could wish with respect to some particulars connected with the subject. All the facts which I shall state may be depended upon; but it is probable, that different persons may draw different conclusions from from them; and to mere opinions I have never shewn myself much attached. Having formerly observed several remarkable changes in fluid substances, in consequence of long exposure to heat in glass vessels hermetically sealed (of which an account may be seen in the fourth volume of my Philosophical Observations); I then formed a design of exposing all kinds of solid substances to great heats, in a similar state of confinement; and for that purpose provided myself with a cast-iron vessel, which I could close at one end, like a digester, and of such a length, that one of the ends might be red-hot, while the other was sufficiently cool to be handled. To this end there was a cock connected to a tube, by means of which I could let off steam, or air, in any period of the process. I imagined, that when substances consisting of parts so volatile as to fly off before they had attained any considerable degree of heat, in the usual pressure of the atmosphere, were compelled to bear great heats under a greater pressure, they might assume new forms, and undergo remarkable changes, similar to what we may suppose to be the case within the bowels of the earth, where, by means of subterraneous fires, various substances bear great heats under very great pressures. I have had this instrument some years; but it was so ill constructed, that I could not make the use of it that I had originally intended. I therefore lately fitted up some gun-barrels in the same manner, and made my first experiment with limestone; expecting, that when the fixed air, and other volatile matters, that might be contained in it, should be compelled to bear a red heat, without a possibility of making their escape, the substance itself might undergo some change; but I had no particular expectation concerning the nature of that change. Vol. LXXIII. I i i I had, I had, however, been so often favoured with valuable results from merely putting things into new situations, that I was encouraged to make the experiment; but I found an unexpected difficulty in getting a cock that would be air-tight and steam-tight under so great a pressure as I wished to apply. I was mentioning these ideas to Mr. Watt, in whose neighbourhood I have the happiness to be situated, when he mentioned a similar idea of his, viz. that of the possibility of the conversion of water, or steam, into permanent air; saying, that some appearances in the working of his fire-engine had led him to expect this. He thought that if steam could be made red-hot, so that all its latent heat should be converted into sensible heat, either this or some other change would probably take place in its constitution. The idea was new to me, and led me to attend more particularly to my former projects of a similar nature, and I began with lime-stone, wishing to try the effect of giving a red heat to lime in which water only should be previously combined, thinking it might possibly have the same effect with making the water itself red-hot. Accordingly, I took a quantity of well calcined lime, and mixing with it a little water, out of which all air had been carefully boiled, I exposed it gradually to a strong heat in an earthen retort, such as I had been usually supplied with by Mr. Wedgwood (who is as much distinguished by his love and generous encouragement of science, as he is by his improvements in his own curious art), not imagining that it could make any difference whether the lime, so prepared, should receive its heat in an earthen retort, or in a vessel of iron or glass. Proceeding, however, in this manner, I found that nothing came over in the form of steam, but that there was a great quantity of air, several hundred times more than the bulk of the water, and at that that time there was in it a considerable proportion of fixed air, which I imagined might either be that which had not been sufficiently expelled before, or might be composed of some phlogistic matter contained in the lime, and the purer air that was yielded by the water: for I own I then concluded, that the air which I got (and which, when the fixed air was extracted from it, was such as a candle would just burn in) came from the water, especially as in some of the processes, the weight of the air I caught was very nearly, if not quite equal to that of the water, and interposing a large glass balloon between the retort and the recipient for the air, I observed that it remained perfectly cool and dry during the whole process; and several hours afterwards there was not the least moisture condensed in it. I also received a quantity of another produce of air made in this manner in mercury, and having viewed it with the greatest attention, observed that, after several days, it never deposited the least moisture. I then calcined a quantity of natural lime-stone with this glass balloon, interposed in the same manner, and found no water, but only air to come from it, though the stone is generally supposed to contain water. But when I used much more than half an ounce of water to the quantity of whiting or lime above-mentioned, I always had some water come over, though very little in proportion to the quantity made use of. I did not fail to examine whether there had been any loss in the weight of the lime, or whiting, in order to determine whether any part of these solid substances had entered into the composition of the air; but I found much difficulty in weighing them with exactness, after shaking them out of an earthen retort, into which I could not see, and to which part of these earthy matters often adhered, so that I could not obtain much satisfaction even when I broke the retort. Besides, there was always some loss of the earth in the cloudiness of the air, whenever the production of it was rapid. In a future process I had abundant proof that the air did not come from any earthy matter with which the water had been combined. Hitherto I had no idea but that all that was necessary to the conversion, as I concluded it to be, of water into air, was to give it a red heat, without which it would not quit the calcareous earth; and I imagined by this means the matter or principle of heat was so intimately combined with it, as not to be separated from it by cooling, as in the case of steam. But I, as well as all my friends, was a long time utterly disconcerted upon finding that when I put the whiting and water into a coated glass retort, the water came over in the form of steam, and little or no air was produced. The result was also the same when I made the process in a gun-barrel, in a porcelain retort, or even in an earthen retort glazed in the inside. That the earth had not lost its property of doing its part in the business, I found by putting more water to the same whiting which had failed in the glass retort, and which had been used no less than four times before, and then heating it in an earthen retort; when again it gave air only, and no water, the same as before. And at this time I observed, that part of the air was hardly to be distinguished from that of the atmosphere. I cannot express my surprise at my unexpected failure with the glass retort; and my speculations on the subject were various, but at that time altogether ineffectual. Among other things it occurred to me that, possibly, some phlogiston, either contained in the earthen retort, or coming through it (though I could not tell how, or on what principle) from the fire, might be necessary to water, and all other substances, assuming the form of air. But when, with this idea, I put spirit of wine, oil, oil, or iron-filings to the lime, I got nothing from these mixtures in glass retorts besides steam and inflammable air, from the decomposition of these substances containing phlogiston. That there was nothing in the materials of which the earthen retort was made that necessarily produced the air, was evident from my not succeeding when I pounded a broken retort, and heated it, mixed with water, in one of glass. Being satisfied that the production of air depended very much upon the retort itself, I thought of using the retort only with water, but without any lime, or earthy substance; and I found it succeed far beyond my expectation. For when I put a small quantity of water into one of these retorts, and endeavoured to distil it gently, I never failed to procure about an hundred ounce measures of air; and this I could do as often as I pleased, with the same retort, and without its losing any weight; and the air produced in this manner had never any portion of fixed air in it, and was always but very little inferior to that of the atmosphere. In all these processes I observed, that very little of this air was procured till all the water that could be poured out of the retort was evaporated, for the difference in the produce was very little, whether I exposed the retort to the fire quite full of water, or with only about an ounce measure of water in it, or even after letting it remain full for a short time, and then pouring out all that I could from it; so that it was only that water which was entangled, as it were, in the pores of the retort, and which had been in some measure united to the substance of it, that had contributed to this production of air. These retorts (which Mr. Wedgewood informs me are made of a mixture of fresh and of burnt Devonshire pipe clay) are pervious to water, though not to air; so that while the air is produced produced from that water which has entered the pores, the rest is sometimes visibly making its escape in the form of a copious smoke on the outside. It was evidently impossible, however, and contrary to all the laws of hydrostatics, that air should enter by the same pores by which the water or steam was escaping, and at the same time that its endeavour to force its way out of the retort was such that it overcame a considerable resistance from the column of water, at the mouth of my recipients. Air might have escaped through any unobserved pores in the retort, but none could have entered that way: and if there was the least sensible crack in any part of the retort, I was never able to collect any air at all. But the following experiments may, perhaps, shew that it is sufficient for the production of air that steam come into contact with clay sufficiently heated. Between a copper still and the glass tube communicating with my recipient for air, I introduced the stem of a tobacco-pipe; and by means of a small furnace, I kept about three inches of the middle part of it moderately red-hot. In this state, making the water boil, I uniformly received air, though mixed with steam, at the rate of five ounce measures in twelve minutes for more than an hour; but when I let the pipe cool, nothing but steam was delivered by it without any air at all. There was no fixed air in this produce, and it was all such as a candle would hardly have burned it. It might, I thought, have been better and also more in quantity if I had not used the stem of a foul pipe. But when I used a clean pipe in the same manner, I did not find the air much, if at all improved. Suspecting this to arise from the near contact of the fuel, I inclosed the tobacco-pipe in an earthen tube, and then I had air as good as I had generally got. got in the earthen retort, and not much worse than that of the atmosphere. Another circumstance I observed was, that if the outside of the vessel which contained the water or steam, through which it passed, when the requisite heat was applied to it, was not dry, or perhaps surrounded with good air (for in those circumstances the following experiment differs from the preceding ones) the experiment did not succeed. When I put the ball of an earthen retort, filled with moist clay, into an iron digester, and applied heat to it, I get only a very little fixed air, which was probably composed of a small quantity of air beginning to be produced from the materials and inflammable air from the vessel. All that come over besides was steam, and at last inflammable air, from the vessel itself. Being now able to procure air by means of water in this most simple method, viz. by water only in the earthen retort, I had an opportunity of ascertaining, with great ease and exactness, several circumstances relating to the process, and of obviating, as I thought, some objections to the conclusion that I had drawn from it. Among other things I fully satisfied myself that the earth of the retort contributed nothing at all to this production of air, but the water only: for having used the same retort till I had got from it nearly an ounce weight of air, or 800 ounce measures, I found that it had not lost so much as a single grain in weight. After the first process it weighed just three grains more than it did at first, and it continued to weigh the same till after the last process. This small addition of weight might easily have come from a little of the water having been imbibed by the neck of the retort, where the heat of the fire could not reach it. When all the processes were over, I kept the whole retort in a red heat for several hours, and then found that, besides losing those three grains, it weighed eight grains less than it did at first. Before this I had found, that the calcined whiting which I had used in the first experiment could not, as some supposed, attract from the atmosphere any considerable part of the air which I got from it, after combining water with it: for two ounces of the whiting (which was the quantity which I generally made use of) did not attract more than eight grains of anything when it was exposed a whole day in an open dish, though it had lost more than half its weight in calcination. It has been imagined by some, that the air which I got in these earthen retorts was that which had been attracted from the atmosphere by the inside surface of them. But, besides that no air could ever be produced without water, to obviate this objection more particularly, when one of these retorts was giving its last portion of air, I immersed the mouth of it in a basin of water; and letting it cool in that situation, filled it again without admitting any access of air to the inside; and yet, on repeating the process with it, the air was produced just as freely as before. This operation I repeated several times. If it be said, that the outside of the retort attracted the air, still the inside, being composed of the same materials, must have attracted air also; and it would have appeared by the ascent of the water from the basin, the retort being sufficiently impervious to air. By some it was imagined, that either the air itself that I procured, or at least the power of the retort to contribute to the production of it, was owing to something that was transmitted from the burning coals, but which could not pass through glass or metals. To determine this, I took an earthen tube, tube, of the same composition with the retort, and putting a little water in it, placed it, surrounded with sand, in a glass vessel, and this again, surrounded also with sand, in an iron one; and yet the heat transmitted through all these substances enabled the earthen tube to give air, in the same proportion, and of the same quality, as it would have done if it had been exposed to the naked fire. Having now procured air, by means of a water, in a very simple and, as I thought, an unexceptionable manner, I wished to make it in greater quantities in proportion to the water employed; and for this purpose I first thought of increasing the size or the thickness of the porous retorts; but I thought it might answer as well if I put into the retort, in powder, the materials of which they were made, or other substances of the same kind. Accordingly, by mixing ground flint and clay in various proportions, I presently increased the quantity of air much beyond my expectation. In the first trials, in which I had much flint and a little clay, I never failed to get 200 ounce measures of air from one of water. Then, using more clay and less flint, I had still more air; and at last, leaving out the flint altogether, and using clay only, I never failed to get much more than 400, and generally between 500 and 600, ounce measures of air from one of water, which was about three-fourths of the weight of the water; and in one particular process I procured very little less than nine-tenths of the weight of the water in air, and this air was never much less pure than that of the atmosphere. Sometimes it could not be distinguished at all from it at all by the test of nitrous air; and once or twice I thought it even purer than that of the atmosphere. I must here observe, that I found it not convenient to put so much water to any quantity of clay as would make it cohere in one mass, but only so much as that it should remain in the Dr. Priestley's Experiments relating to Phlogiston, form of powder. By this means it would easily pour out of the retort when the experiment was over. The weight of the water expended in this production of air I ascertained, in the most unexceptionable manner, by weigh- ing the retort, with all its contents, before and after the pro- cess. I shall explain this by the result of two of the processes. In one of them, the retort and moistened clay together lost in weight 1 oz. 4 dwts. 12 grs. after yielding 741 ounce measures of air, which (in the proportion of six grains to one ounce measure) would have weighed 18 dwts. 12 grs. and conse- quently three-fourths of the weight of the water. In the other process the loss of weight was 15 dwts. 18 grs. after yielding 556 ounce measures of air, which would have weighed 13 dwts. 21 grs. The proportion, therefore, between the weight of the air and that of the water was 111 to 116, or nearly nine to ten. I also found now, that so much heat as I had hitherto ap- plied was neither necessary nor useful. In the last mentioned process the retort was constantly suspended about six inches above a moderate charcoal fire; at another time more than twelve or fifteen inches above it, where a Fahrenheit's thermometer did not shew more than 210°. With this moderate heat I got 465 ounce measures in the course of about twelve hours. When the retort was suspended within six inches of the fire, the air was generally produced at the rate of 30 ounce measures in five minutes. But a thermome- ter, the bulb of which was immersed in the clay, was still only at the heat of boiling water. In all these processes, however, there was evidently some loss of water; for, excepting the first experiment with the lime, I never got the whole weight of the water in air; and it might be said that I only expelled the air before contained in the water, though from these experiments it appeared to con- tain much more air than it had been thought capable of con- taining. To obviate this objection, I contrived to catch all the water that escaped through the pores of the retort in the fol- lowing manner. Having put the moistened clay in an earthen tube, to which I had fitted a cock and a long glass tube (by means of which I could collect all the air that came from it), I put this within an iron tube, which was closed at the end next the fire, but open at the other end, and so long that I could easily keep this open and quite cool while the other was in the fire; conse- quently, whatever water escaped through the pores of the earthen tube, it would be condensed in the cool part of the iron one. This water I carefully collected, and always found that the weight of it, together with that of the air produced in the experiment, was nearly that of the original weight of the water, estimated by the loss of weight in the earthen tube and its contents. I also found, that the water so collected served for the production of more air, just as well as any other water whatever, so that there had been no decomposition of the water in the case. In the last process that I went through of this kind, the loss of weight in the earthen tube, or rather of the water con- tained in it, was 12 dwts. 4 grs.; the air collected was 173 ounce measures, which would have weighed 4 dwts. 3 grs. and the water which escaped through the pores of the earthen tube, and which I collected, was nearly 8 dwts. 3 grs.; so that the air and this water together weighed 12 dwts. 14 grs. or ten grains more than the original water. But as I estimated the weight of the water only by the space which it occupied in a cylindrical glass tube, divided according to ounces and parts of ounces of water, it was not easy to avoid an error of a few grains. At other times there was an error of a small magnitude on the other side. But it will appear hereafter, that more steam must have escaped invisibly at the open mouth of the iron tube than I was aware of. That nothing could enter by the pores of the retort at the same time that the water was making its escape out of them, I thought I ascertained pretty satisfactorily by immersing the bulb of it in mercury, contained in an iron vessel. In these circumstances I obtained air as usual, only the produce was not so rapid. In this way, however, I procured above an hundred ounce measures of air from moistened clay; and I discontinued the process without perceiving any termination of it. But the moment the retort was raised out of the mercury, it gave air three times as fast as it had done before. The quality of the air was the same in both cases, viz. a little worse than that of the atmosphere. I even collected thirty ounce measures of air when the bulb of the same retort was immersed in hot linseed oil, but the production of air gradually ceased, and the next day I found the retort almost full of the oil, which had soaked through it. Distilling this oil I get 300 ounce measures of air, wholly inflammable, except a very few ounce measures at the last, which were only phlogisticated. Still hearing of many objections to the conversion of water into air, I now gave particular attention to an experiment of Mr. Cavendish's, concerning the re-conversion of air into water, by decomposing it in conjunction with inflammable air. And in the first place, in order to be sure that the water I might find in the air was really a constituent part of it it, and not what it might have imbibed after its formation, I made a quantity of both dephlogisticated and inflammable air in such a manner as that neither of them should ever come into contact with water, receiving them as they were produced in mercury; the former from nitre, and in the middle of the process (long after the water of crystallization was come over), and the latter from perfectly-made charcoal. The two kinds of air thus produced I decomposed by firing them together by the electric explosion, and found a manifest deposition of water, and to appearance in the same quantity as if both the kinds of air had been previously confined by water. In order to judge more accurately of the quantity of water so deposited, and to compare it with the weight of the air decomposed, I carefully weighed a piece of filtering paper, and then having wiped with it all the inside of the glass vessel in which the air had been decomposed, weighed it again, and I always found, as nearly as I could judge, the weight of the decomposed air in the moisture acquired by the paper. As there is a source of deception in this experiment, in the small globules of mercury, which are apt to adhere to the inside of the glass vessel, and to be taken up by the paper with which it is wiped, I sometimes weighed the paper with the moisture and the mercury adhering to it; and then exposing it in a warm place, where the water would evaporate, but not the mercury, weighed it again, and still found, as nearly as I could pretend to weigh so small a matter, a loss of weight equal to that of the air. I wished, however, to have had a nicer balance for the purpose: the result was such as to afford a strong presumption that the air was re-converted into water, and therefore that the origin of it had been water. Another presumption in favour of the generation of our atmosphere from water was, that the purity of the air that I produced from it is so very nearly the same with that of the atmosphere. And the degree of heat requisite to produce it is no greater than may be given by the rays of the sun in certain circumstances. Subterraneous fires, however, would be abundantly sufficient for the purpose, as it appears to be sufficient for the conversion of water into respirable air, that it come into contact with clay, and perhaps many other earthy substances in the form of vapour. I must, however, observe, that when I threw the focus of a burning lens upon a quantity of moist clay, either in vacuo, or in common air, I got no air from it. I made this experiment both with the clay exposed in an open dish, and also confined in a short earthen tube. Had I then proceeded to repeat this last process with a communication between the inside of the earthen tube and the external air, as I then proposed to do, but was prevented, I should much sooner have discovered what I did afterwards, viz. that there was no real conversion of water into air in this process. In favour of which, however, it may not be amiss to observe, that the great difficulty Mr. de Luc and others have found in expelling all air from water, is best accounted for on the supposition of the generation of air from water, though in other circumstances than those that I have observed. I have the pleasure to add, that Mr. de Luc himself concurs with me in this opinion. The difficulty that strikes many persons the most forcibly, is the want of analogy between the conversion of water into air with any other known facts in philosophy or in nature. But admitting that this conversion is effected by the intimate union of what is called the principle of heat with the water, it appears to me to be sufficiently analogous to other changes, or rather combinations of substances. Is not the acid of nitre, and also that of vitriol, a thing as unlike to air as water is, their properties being as remarkably different? And yet it is demonstrable, that the acid of nitre is convertible into the purest respirable air, and probably by the union of the same principle of heat. It is true, that steam is a thing very different from air, and I find that it is not able to decompose nitrous air; but then, though it has acquired sensible heat, it has got no latent heat so intimately combined with it as it is with air; and for the same reason, perhaps, the vapour of nitrous acid is not dephlogisticated air. By the same process by which respirable air is made by means of water, inflammable air may be made from liquid substances containing phlogiston. Making spirit of wine to boil in a glass retort, I made the vapour pass through the stem of a hot tobacco-pipe, and found that it was all seemingly converted into inflammable air, and it was of that kind which burns with a lambent white flame. But when I let the pipe cool no air was produced, but only vapour, which was instantly condensed in the water. Being now master of a new and easy process, I was willing to extend it to other liquid substances; and I presently found, as I then imagined, that, by this means, I could give a permanent aerial form to any liquid substance that had been previously thrown into the form of vapour. When I made the vapour of spirit of nitre, heated in a glass retort, pass through the stem of the hot tobacco-pipe, I got as pure dephlogisticated air as ever I have procured from nitre; nitre; though the cork, by which the retort was connected with the pipe, was dissolved, and must have contributed to contaminate it, and give it a slight mixture of fixed air. With oil of vitriol I got air considerably phlogificated, so that a candle would not have burned in it; but this I also attribute to the cork, which was dissolved in the process. The result was nearly the same when I used water impregnated with vitriolic acid air, though the cork was not dissolved. But this acid is known to contain much phlogiston. Spirit of salt gave air no purer than the best atmospherical air. But as by this process I never got air so pure as this from water only, I concluded, that even this acid, as well as the nitrous and vitriolic, is capable of being turned into dephlogisticated air. When I used water impregnated with fixed air, this air was expelled by the heat, and came over without any change that I could perceive, except that the residuum was larger, from the water that came along with it. The air I got afterwards was only that from the water, and of the same quality as if it had not been impregnated with fixed air. Water impregnated with alkaline air gave neither fixed nor inflammable air, which I had rather expected, but only air considerably phlogificated; though some of it was so pure that a candle would have burned in it. N. B. In all these experiments with the tobacco-pipe all the air was remarkably turbid, like milk, and even the common air in the retort before the process properly began. In this state of the experiments I think I may venture to say, that no person could have seen them without concluding that there was a real conversion of water into air, there being no known principle or fact in philosophy, that could have led any person to suspect a fallacy in the case. In this, therefore, I must have acquiesced, as indeed did all my acquaintance, even those who had been the most incredulous on the subject, after they had themselves seen the experiments. But I was led to the farther prosecution of this business, in consequence of having observed that the purity of the air which I procured depended upon the state of that which was immediately contiguous to the earthen retort, or tube, in which I supposed the conversion to have been made; and that some communication with the atmosphere was necessary to the production of any air, as in the experiment with the digester, and those with the clay and the burning lens. And since pure external air was necessary in order to procure good air, it was concluded by several of my friends, and especially Mr. Watt, that the operation of the earthen retort was, to transmit phlogiston from the water contained in the clay to the external air; and that the water, thus dephlogisticated, was capable of being converted into respirable air by the intimate union of the principle of heat. In order to ascertain what the influence of the external air in this case really was, I inclosed an earthen retort filled with moistened clay in a large glass receiver, open at both ends, through the upper orifice of which (being narrow) I thrust the neck of the retort, luting it so as to be perfectly air-tight; and placing the receiver in a basin of water, by which the air within was cut off from all communication with the external air, I fitted to the mouth of the retort a glass tube, through which I could receive whatever was produced in the process. In this situation I heated the retort by means of Mr. Parker's excellent burning lens, when air was received through the tube communicating with the inside of the retort as usual; but at the same time the water rose within the receiver. This effect might be owing to a phlogistication of the air within the receiver; but it was soon diminished far beyond the utmost limit of that process, so that very little of it remained; and examining this air, I found it to be but very little worse than that of the atmosphere, as that which came from the retort was a little better. This experiment made it probable, that the air on the outside of the receiver had actually passed through it, only a little purified in its passage; and yet it was contrary to all the known principles of hydrostatics, and even anything hitherto known in chemistry, that air should be transmitted through a vessel of this kind, and in a direction contrary to that in which it would have been forced by the pressure of the atmosphere; while the water, with which the clay was moistened, went the other way. For had the retort been pervious to air, as the inside had a free communication with the atmosphere, the water could not have risen within the receiver. This, however, appeared to be the case by the following decisive experiments. Having filled the earthen retort with the moistened clay as before, I made the inside of the receiver perfectly dry, and placed it in a basin of mercury; when, upon heating the retort as before, the receiver was all covered with dew, which collecting into drops trickled down the inside of the receiver, and remained upon the mercury, which rose within the receiver, while air was received from the retort as usual. I had no doubt, therefore, but that all the water within the retort would have got through into the receiver. Spirit of wine, or something that had the smell of it, was transmitted from the clay through the retort in the same manner. I then filled the receiver with inflammable air, and upon heating the retort it was all drawn through it, and delivered as strongly inflammable as ever by the tube communicating with the inside of it, while the water rose within the receiver, and even covered the retort, which was fixed at the very top of it, so that hardly any of the inflammable air remained within it. In like manner nitrous air passed through the retort unchanged. From these experiments it is impossible not to infer, that the clay of the earthen retort, being thus heated, destroys for a time the aërial form of whatever air is exposed to the outside of it; which aërial form it recovers after it has been transmitted in combination from one part of the clay to another, till it has reached the inside of the retort, while the water is drawn through it in the contrary direction. Had this hypothesis been proposed *a priori*, it would, I doubt not, have been thought more extraordinary than the conversion of water into air. I propose to make many other experiments in the prosecution of these; but till I have an opportunity of doing this, I shall not trouble the Society with any conjectures that have occurred to me on the subject. The great difficulty with respect to the experiment with the lens is, that the water should pass through the retort one way, and the air the other, and yet that the air should not be able to pass without the water. It is also not a little extraordinary, that the weight of the air and that of the water should be so nearly equal. In the last place I must observe, that there is nothing in this experiment that contradicts the idea of the conversion of water into air, though it does not prove it: for still the experiment with the tobacco-pipe, in which the steam is made red-hot (whereas in that with the lens it is only of a boiling heat) cannot be explained so well on any other hypothesis any more than Mr. Cavendish's experiment on finding water on the decomposition of air. I cannot conclude this account without acknowledging my obligation to Mr. Parker for the use of his incomparable lens, and his obligingly assisting me in the management of it. Indeed, without this very instrument, or one of greater power than my own, I do not know that the last mentioned experiments could have been made at all; certainly not to so much satisfaction.