flate, and even when on the decay, feems to poffefs this quality in an equal degree.

The plants on which the Author made his experiments were confined in jars about 9 or 10 inches long, and from inch to 2 in diameter. One of them which had been confined about a fortnight had confumed three-fourths of the common air included with it. Another, in the space of a month, had abforbed feven-eighths of the common air to which it was expofed. The plant was quite yellow, and dead; but though it had been fo for fome time, it had ftill continued to abforb air: fo that the remaining air was afterwards reduced to one-tenth of the original quantity. The inflammable air, to which another plant had been expofed, was reduced to one-feventh of the whole. An accident prevented the Author from examining the quality of the refiduum; which, however, when only one-third part of the air had been confumed, was found, to all appearance, to be as inflammable as ever. Another plant confined in nitrous air, became yellow, and died; and had then confumed one-third of the air.

The mention of the Author's experiments relative to vegetation naturally leads us, as well as the fingularity and importance of the fubject itfelf, to take particular notice of a moft remarkable difcovery, to which thofe experiments gave occafion. This is nothing less than the fpontaneous emiffion of the pureft dephlogisticated air from common well water, in certain circumftances. We cannot more properly commence our account of this difcovery, than by prefixing to it the Author's own edifying exordium:

Few perfons, fays he, I believe, have met with so much unexpected good fuccefs as myself, in the course of my philofophical purfuits. My narrative will fhew that the first hints, at leaft, of almoft every thing that I have difcovered, of much importance, have occurred to me in this manner. In looking for one thing, I have generally found another, and sometimes a thing of much more value than that which I was in queft of. But none of these unexpected difcoveries appear to me to have been fo extraordinary as that which I am about to relate; and it may ferve to admonish all perfons who are engaged in fimilar purfuits, not to overlook any circumstance relating to an experiment; but to keep their eyes open to every new appearance, and to give due attention to it, how inconfiderable foever it may seem.'

The Author having obferved bubbles of air that seemed to iffue fpontaneously from the roots of feveral plants growing in water, was firft led to fufpect that this air had percolated through the plant; which had probably feized upon and retained the phlogiston of the air which it had imbibed, and was now emitting the abforbed air, deprived of that principle, and

M 3

con

confequently in a state of greater purity. In fact, on collecting and examining fome of this air, he found it fo pure, that one measure of it, and one of nitrous air, occupied the space of only one measure.

The Author, however, afterwards found that the plants had no fhare in this production of air: for on taking them out of the vials, the remaining water continued to emit air as plentifully as when the plants were growing in it. He observed too that the vials and other veffels, in which this pure air had been emitted from the water, had their bottoms and fides more or Jefs covered with a green matter, from which the air evidently feemed to proceed. It appeared to him that this green matter could neither be of an animal or vegetable nature; but that it was a fubftance fui generis, and that neither the external air or animalcules could have any thing to do in the formation of it: for it was produced in vials clofely corked, and in the middle veffel of Mr. Parker's apparatus *.

On filling a number of vials with different kinds of water, as river water, rain water, pump water which contained a confiderable quantity of fixed air, and water artificially impregnated with fixed air; he found, after they had flood fome time, that no green matter was depofited in any of them, except in those which contained the pump water. He afterwards however found that much of this green matter, as well as of the pure air that rifes from it, was produced from the water that had been ftrongly impregnated with fixed air.

Of the purity of the air emitted from pump water, under thefe circumftances, the Reader may form a general eftimate from one of the Author's experiments; in which he ufed a tall conical receiver, about 18 inches high, and 5 wide at the bottom. This had been employed in former experiments, and was coated with this green matter, which in time paffes gradually to a kind of yellow or rather orange colour. On the 17th of September, 1778, the Author had taken all the air from this receiver, and had re-filled it on the 14th of October following, he took from it about 9 ounce meafures of air-the pureft he had ever got in this method: for one measure of it,

From fome experiments that we have made on this fubject, we fee reafon to infer that this green matter will not be depofited in vials closely corked, unless fome air is included; and that the quantity of the depofit bears fome proportion to that of the air left in the vial. In open vials completely filled, and inverted in water, the water contained in the vials has an intermediate communication with the atmosphere; and the procef goes on, as is defcribed above: but if that communication be flopped, from the beginning, by inverting the vials in quickfilver (a fluid impermeable to air), no green matter, or pure air, is produced.

and two of nitrous air, occupied the fpace of only 0.44 of a meafure which is quite as pure as dephlogisticated air is at a

medium.'

We have hitherto kept out of fight a most remarkable cir-› cumstance attending the production of this green matter, and the emiffion of this pure air; and which the Author feems to have obferved too often to justify a fufpicion that he can have been mistaken with respect to the fact. This is, the inftrumentality of the Sun's Light, qua light, and independent of his mere heat, in the production of the green matter, and the confequent emiffion of dephlogifticated air.

After making fome obfervations on this green matter, and declaring that he never found it except in circumftances where the water had been expofed to the light; the Author goes on to fay, that he had not proceeded far in this inquiry till it was too late in the last fummer to make ufe of Sunshine; though he was affiduous enough to avail himself of the ftate of the weather, fuch as it was. He then draws these general conclusions from the whole that he had hitherto been able to observe :

That whatever air is naturally contained in water, or in fubftances diffolved in water, as calcareous matter, &c. becomes, after long standing, but especially when expofed to the fun, depu rated, fo as at length to become abfolutely dephlogifticated; and that this air being continually emitted by all water, expofed to the action of the fun's rays, muft contribute to the melioration of the ftate of the atmosphere in general.

• When I have kept water a long time in the shade, it has not generally yielded any other kind of air than it would have yielded at the firft; and though, when it has been kept in an open veffel the air has been better, it has never been fo good as the air in the fame kind of water that has been expofed a much lefs time to the fun.

No degree of warmth will fupply the place of the fun's light; and though, when the water is once prepared by expofure. to the fun, warmth will fuffice to expel that air; yet, in this cafe, the air has never been fo pure, as that which has been yielded fpontaneoufly, without additional heat. The reafon of this may be that, befides the air already depurated, and on that account ready to quit its union with the water, heat expels, together with it, the air that was phlogifticated, and held in a clofer union with the water; which air, the action of light, whatever that be, would in time have depurated alio.

The quantity of air, yielded by water fpontaneouflv, far exceeds that which can be expelled from it by heat. Inc.ed, I have frequently obferved, that whatever circumftance depraves air, leffens alfo the quantity of it; fince it requires a large quantity of dephlogifticated air to make a finall quantity of

M 4

phlogif

phlogisticated air, or even of common air, which is air partially phlogisticated.

If the water naturally contains fixed air, yet, in confequence of this exposure to the light, it is all diffipated, and the natural refiduum of it becomes pure dephlogisticated air. For no fixed air at all, but only the pureft dephlogifticated air, is at length procured from it; and water impregnated with fixed air yields, after this expofure, the greatest quantity of dephlogisticated air.'

In confirmation of thefe conclufions, the Author recites fuch of his experiments as appear fufficient to establish every thing that is of importance in them. Recommending the perufal of thefe to the philofophical reader, we fhall tranfcribe the Author's lateft obfervation relative to this subject, inserted at the end of this yolume:

On the Effect of Light on Water.

My obfervation that Light difpofes water, containing calcareous and other fubftances, to make a depofit of a greenish or brownish matter, and then to yield dephlogisticated air, seems to be confirmed by the following experiment.

On the 19th of February, 1779, I placed two jars of pump water, each containing about 170 ounces, in the same south window; one of them nearly covered from the fun with brown paper, and the other quite uncovered. In about ten days, the water in the uncovered jar had yielded about four ounce meafures of air, and the covered jar only a few bubbles. Taking a journey I could make no farther obfervations on these jars till my return; but on the 2d of April I found that the uncovered jar had yielded 10 ounce measures of air, fo pure that one measure of it, and one of nitrous air, occupied the space of .84 measures; whereas the covered jar had very little more than one ounce measure, and with this the meafures of the teft were 1.55 measures; i. e. by no means fo pure as the former. Alfo the uncovered jar had a fediment larger than the other in about the fame proportion, viz. of 10 to 1. Oil of vitriol expelled from this fediment a very great quantity of fixed air. N. B. The lowest part of the jar was not covered with the paper, left being moistened with the water, in the difh in which the jar tood inverted, it should imbibe the water, and caufe it to evaporate too foon *.'

We

It may not be amifs, on an obfervation fo very fingular and curious, to add our teftimony, fo far as it goes, to that of the Author. Two quart glafs retorts, and two 12 ounce vials, filled, at the fame time, with the fame well water (which naturally contained a mo. derate quantity of fixed air), were expofed to the fun, in July latt, inverted in a balon of the fame well water, ftanding on a table placed

We must not omit to take notice of an objection, which Dr. Priestley here anticipates and anfwers, that may be made to the refults which he deduced from fome of his former experiments on air. From thefe he inferred that vegetation was one of the means employed by nature in purifying the atmosphere, by depriving it of that noxious principle with which it is charged by animal refpiration, combuftion of inflammable fubftances, putrefaction, &c. From the foregoing experiments a fufpicion may arife that, as his trials were made with plants growing in water, the obferved melioration of the included air might be principally effected by the emiffion of dephlogifticated air from the water in which they grew. We fhall give what he says on this fubject:

It will probably be imagined that the refult of the experiments recited in this fection throws fome uncertainty on the refult of former experiments, from which I have concluded that air is meliorated by the vegetation of plants; especially as the water by which they were confined was expofed to the open air, and the fun, in a garden. To this I can only fay, that I was not then aware of the effect of these circumstances, and that I have reprefented the naked facts, as I obferved them; and having no great attachment to any particular hypothefts, I am very willing that my reader should draw his own conclufions for himself.

I muft inform him, however, that my experiments at Leeds were made in a north-east window of the house, where the infuence of the light on the water could not be very confiderable; that fome of the proceffes were completed in two days, and generally in about a week; and that the water within the jars was fo fmall, in proportion to the quantity of air, that I do not at prefent imagine that the melioration of the air at that time could have been owing to it. Befides, as I have obferved, I frequently kept air in the fame expofure, with refpect to water, light, and every other circumftance that occurred to me to attend to, and the fame fpace of time, but without any plant vegetating in it, when there was no fenfible melioration of it."

In an Appendix to this volume are contained feveral papers communicated to the Author by his Correfpondents. No and 2, contain observations, by Sir William Lee, on the efficacy of

in the middle of a garden. One retort, and one vial, were covered with black filk; which, at the fame time that it excluded the fun's light, communicated more heat to the water than was tranfmitted to the other retort and vial. After they had ftood a month, a large quantity of pure air, amounting to about one-ninth part of the bulk of the water in each, had been feparated in the naked retort and vial whereas on taking off the filk from the other two veffels, only a bubble of air, not exceeding the fize of a fmall nut, was found to have been feparated from the water contained in them.

water