subsists together very commonly in the same compound substance, they simultaneously repel each other on the decomposition of the substances with which they were united; and from this repulsion, and also the repulsion of oxygen and light of one another, he chiefly accounts for the combustion of oxygen gas.

On my principles above stated, we can explain why quicksilver of the temperature of 1000°, or more, of Fahrenheit's scale, cannot unite to the oxygen of oxygen gas; but why at this temperature caloric separates oxygen in the gas state from oxide of quicksilver, and why at the temperature of between 600° and 1000°, quicksilver does unite to the oxygen of oxygen gas, but at which temperature oxygen gas is not separable from the oxide of this metal : and again, why at below 600°, oxygen of oxygen gas can neither unite to quicksilver, nor be separated from its oxide.

These principles may be applied variously to the interpretation of the phenomena of combustion, and other cases of chemical combination, according to the state of aggregation of the substances which have a chemical attraction for one another. For example; the oxygen of oxygen gas cannot unite to the constituent substances of a wax or tallow candle, in a low temperature; because the cohesive attraction of the wax or tallow, as well as the chemical attraction between their constituent substances-hydrogen and carbon-counteract the chemical attraction between the oxygen of oxygen gas, and the ultimate particles of hydrogen and carbon of the candle. But at a pretty elevated temperature, when the wax and tallow are in the vapour state, the cohesive attraction no longer subsisting among the ultimate particles of these substances, and the motion excited by the ignited portion of wax or tallow bringing the ultimate particles of these substances within the sphere of chemical attraction of the particles of oxygen gas, these latter unite first to the carbon, by virtue of the stronger attraction between the oxygen and carbon, than that between the oxygen and hydrogen; but, in the instant of this disengagement of the carbon from the hydrogen, a portion of this hydrogen unites also to oxygen, and thus not only carbonic acid gas but water is produced; hence the blue flame, as appears on other occasions, in which there is combustion of hydro-carbonate gas. The remainder of the hydrogen of the decompounded wax or tallow, ascending in the gas state, it unites to the oxygen of oxygen gas; and, as in other cases of combustion of oxygen gas with hydrogen, a white or straw-coloured flame is produced. The wick answers the purpose, by means of capillary attraction, of applying the wax and tallow in such quantities as can be decompounded and combine with the oxygen of atmospherical air; hence the combustion is gradual and equal. The wick itself contains hydrogen and carbon; hence, in combining with oxygen of oxygen gas, it also produces a blue flame. Hydro-carbonate gas being specifically heavier than hydrogen gas, is another reason for the blue flame appearing distinct from the white, and at the inferior part of the frustum of a cone of flame

from a burning candle. It is scarcely necessary to say that hydrogen and carbon are constituent substances of wax and tallow; and that when they combine with oxygen, as above explained, the products are water and hydro-carbonate gas.

If I did not use the term demonstration in a more strict and precise sense than is usual, except in mathematics, I would venture to affirm that this theory is almost demonstrated by the agency of fire and carbon on black oxide of manganese. If colourless vitrifiable matter, and that oxide be melted together, by means of the white part of the flame of a candle, purple coloured glass will be produced; for then the hydrogen of this part of the flame carries off little or none of the oxygen from the oxide: but if these substances be melted together by the blue part of the flame, colourless glass will be produced; for then the carbon of this part of the flame carries off oxygen from the black oxide, and produces white oxide. By combining oxygen anew with colourless glass, containing manganese, it will become purple: and this is effected by melting such colourless glass with a little nitrate of potash, or by melting it in open vessels by the yellow flame of a candle. By separating oxygen from glass rendered purple by manganese, colourless glass is produced: and this is effected by melting such coloured glass with a little carbon, or by the blue flame of a candle. I have not thought it necessary to distinguish between the indigo and violet rays; or to notice the extremely thin film of violet flame, which an attentive observer may perceive surrounding the inferior part of the white, extending as high but scarcely higher than the wick, because the explanation is perfectly obvious from the preceding distinction of the two kinds of gases afforded by the candle. If I have explained more satisfactorily than former members of this society the above phenomena of combustion, I owe this advantage to some experiments of combustion of inflammable gases which I have made for some years past, in my chemical lectures; by which the colours are shown to be very different, and correspond to the above theory. I apprehend the theory of an ingenious member of this society cannot explain adequately the phenomena, and does not appear supported by any facts: for there is no evidence that the white flame is not equally the effect of immediate decomposition as well as the blue; or that the blue becomes white flame by ignition; and I presume that the experiments which I have mentioned show that these differences depend upon different decompounding substances contained in the candle.

With regard to the evidence afforded by the foregoing experiments concerning the composition of water and of hydrogen and oxygen gas: these substances are now accounted for in two ways only; namely, 1. By saying that these two gases consist of water and imponderable matter; and that during combustion the water is precipitated. 2. By saying that the two gases consist of a peculiar basis, one of which is named oxygen and the other is hydrogen, each of which is rendered into the gas state by uniting to caloric,

and perhaps also to light; and that during combustion these bases unite with one another, thus compounding water and discharging caloric and light. If complete demonstration could be given, there would not be two opinions; for its proofs, if understood, command universal assent: but the case being otherwise, that opinion must be adopted on the side of which the evidence preponderates according to the laws of reasoning in physical science. Now with regard to the former of these opinions, I can perceive but two facts in support of it. The first of these is supposed to afford a sort of synthetical proof: it is the instance of water being required to obtain, by fire, the whole of the carbonic acid from carbonate of baryt. Here the fact, if admitted, is true only of carbonic acid; but even in this case it has not been shewn that the water enters into the composition of carbonic acid; and until that is proved, it is warrantable to suppose that the water serves to disengage the carbonic acid, by attracting and uniting with the baryt; especially as this acid gas may be wholly separated from other substances without the intermediation of water, and as it can be compounded by uniting the driest carbon to the driest oxygen gas. Further, although carbonic acid gas be very apt to contain water, it may be obtained in so dry a state as not to indicate moisture to the most delicate tests.

The next fact affords a sort of analytical proof; which is the existence of water in air or gas in general, as shown by muriate of lime, acetite of potash, sulphuric acid, quicklime, potash, &c. But the absorption of water by these substances only shews that it may be suspended or dissolved in gases; but not that it enters into their composition. Gas may be obtained or rendered quite free from water, as just said; and as the compounds by the combination of the gases with various different substances are quite different from the compounds by the combination of water with the same various different substances, there does not appear to be in this case any admissible evidence of water being an essential constituent substance in the composition of gases.

The experiments of separating oxygen and nitrogen gas from water by boiling, and by exposing it in vacuo and applying caloric, only shew how much more tenaciously, and in how much greater quantity, these gases are contained in water than is commonly supposed; for however long a given quantity of water yielded gas, so that there seemed no limit to the quantity, the water which afforded it was still very considerably greater than the weight or quantity of gas.

With respect to the other opinion, that hydrogen and oxygen gas are compounded of a peculiar basis and caloric, and that water is compounded of these two bases-1. It has been shewn by an experiment, the accuracy of which is not questionable, that when these two gases produce water by combustion, the amount of the weight of the water is exactly the weight of the gases consumed, and that no ponderable matter but water is produced when nothing but the

bases of these two gases unite; but that when other substances, such as nitrous acid and carbonic acid, were produced, as well as water, then nitrogen and carbon were present. 2. It has been shewn, that when oxygen in a concrete state, as in metallic oxides, unites to the hydrogen of hydrogen gas, the water thus compounded is equal in weight to that of the hydrogen gas consumed, and the deficient weight of the oxide. 3. It appears that when hydrogen in a concrete state, as in alcohol, is united to the oxygen of oxygen gas, the water compounded is equal in weight to the weight of the alcohol and oxygen gas consumed, provided the addition be made of the carbonic acid also produced by the carbon of the alcohol. The carbon of alcohol can be shewn by analysis; and its existence in carbonic acid can be shewn both by analysis and synthesis. 4. When water is applied to certain substances, nothing but hydrogen gas is obtained, and new compounds, the constituent parts of which are oxygen and the substance applied; for these new compounds have the same properties as the compounds of the same substances so applied to water, and of the oxygen of oxygen gas. And in these cases the additional weight derived from the water, together with the weight of the hydrogen gas separated from the water, is equal to the water destroyed. 5. It has been rendered at least very probable, that when oxygen and hydrogen are rendered into the gas state, they absorb or unite with a large quantity of caloric, or of both caloric and light. I might add, that besides the positive evidence just stated, there does not appear to be any evidence to contravene the conclusion that water is compounded of hydrogen and oxygen; and that these gases are compounded of a peculiar basis, called hydrogen and oxygen, united to caloric for it would be easy to show, but it would be digressing too far, that the remaining partizans of the phlogistic sect have advanced erroneous evidence in some cases, and in others have neglected the consideration of opposing but well authenticated evidence.

In the experiments contained in this paper no support can be found for the former opinion concerning water, and hydrogen and oxygen gas; but they confirm the latter opinion. For, 1. the combustion of these gases, rendered perfectly dry, afforded water: Exp. IV. 2. The evidence from the process above described is peculiar; for no one other process affords oxygen and hydrogen gas from water. In all other processes for decompounding water, either the decompounding concrete substance receives something from the water applied, and at the same time hydrogen gas is produced; as in the instance of passing water through a red hot iron tube or the decompounding concrete substance receives something from the water; and this compound unites to that in water, which produces hydrogen gas with caloric; as in the instance of applying water to red hot carbon; or two decompounding substances being applied, each of them receives a different thing from the water, namely, hydrogen and oxygen. Hence, by the process above

:

described, the objection is removed, that the hydrogen gas might be separated from the decompounding substance itself, by water taking its place. 3. The production of hydrogen and oxygen gas, and the production of water by the combustion of them, in the above process, afford an additional evidence of the decomposition and composition of water, and of hydrogen and oxygen gas; as the mode of their production by this process is perfectly consistent with the rationale of their production in all other cases.

Thus it appears that the grandest discovery ever made in chemical philosophy, by an illustrious member of this Society, in 1781, has been confirmed by a number of subsequent experiments. The body of evidence is indeed so numerous, and of such a nature, that, in the minds of those who understand its import, and who rely on the accuracy of the weights and measures employed, it produces as much conviction concerning the composition of water as can be obtained by the evidence of almost any other case of composition. I must, however, beg leave to protest against those able philosophers, who have maintained, that the composition of water, and several gases, has received full and complete demonstration; by such unwarrantable pretensions their adversaries have obtained over them some advantages. For in the chain of causes and effects there are some links which cannot be explained by the direct evidence of sense; and there, in so far as we admit hypotheses, although consistent with the phenomena, we may be said to quit day-light, notwithstanding we at the next step emerge into light derived from the perceptions of sense. For instance, if I combine a certain known weight of hydrogen and oxygen by combustion of their gases, and produce water equal in weight to that of these gases; and if I again resolve this water, by means of the electric fire, or red hot iron, into the original quantity of hydrogen gas and oxygen, I cannot give the full and complete demonstration of the composition water and these gases; for, as I proceed in the interpretation, I at length come to demonstrate the mode of agency of the particles of the hydrogen and oxygen gas on one another when they produce water, caloric, and light. But here I must call in the aid of the imagination; accordingly I imagine that the gases consist of hydrogen and oxygen, which are ponderable, united to caloric, and perhaps light, which are imponderable; that these ponderable particles unite with one another, and their caloric and light are set at liberty. Now here I have not any evidence of sense: for I cannot perceive, by the senses, the existence of the composition of the gases just stated, nor of their decomposition, and union of their ponderable parts. This being the case, other kinds of imponderable may unite, or escape, besides caloric and light; consequently I cannot give the full and complete demonstration in these instances. But the same objections may be made to the pretensions to demonstrate fully and completely the composition of sulphate of soda, sulphate of pot-ash, or any double salt, perhaps, whatever: for caloric is separated, and possibly kk