of heat, did only the same quantity of aqueous vapour subsist in it as before sunrise; while on the contrary, its quantity ought to increase by a greater evaporation being produced on the ground, which dries when heated by the sun. Hence it appears, that there is some connexion between the increase of the quantity of electric fluid and the diminution of that of aqueous vapour in the atmosphere, during this period.

Art. IX. This points out, in the first place, a formation of electric fluid in the atmosphere, while the sun's rays pervade it. Light, the increase of which in the atmosphere is here the immediate cause, is certainly one of the component parts of the electric fluid; therefore, this fluid must be composed in some operations of nature on our globe. Now it is here already probable, that the sun's rays, in pervading the atmosphere, encounter in it the substances with which they compose the new quantity of electric fluid then manifested; and that, in general, they enter there into various combinations, is proved by their intensity being sensibly greater on the top of high mountains than in the lower parts of the atmosphere, as has been shown from experiments by M. de Saussure; which difference must proceed from their quantity being diminished in pervading the air.

Art. X. Some other experiments of M. de Saussure lead besides directly to this system concerning compositions and decompositions of electric fluid, as producing phenomena, the causes of which were unknown or mistaken. For instance, it has been found by experience that, when water is poured upon an isulated plate of hot iron connected with an electroscope, this plate becomes negative: whence it had been concluded that, when water is converted into vapour, it requires a greater capacity for electric fluid; and thus deprives of a certain quantity of this fluid the body on which it evaporates. But M. de Saussure having repeated the same process upon different heated bodies, found that some, in particular silver, became positive: whence he concluded very naturally, that during the evaporation of water on hot iron some electric fluid was decomposed, and some on the contrary composed when the same operation took place on silver. He has also surmised, what I have since found by direct experiments related in my work Idées sur la Météorologie, that in the discharge of the Leyden vial, and in my experiment of the magic picture, the spark produces some diminution of the quantity of electric fluid on these bodies; which cannot be but by decomposition. It will successively be seen in what manner these previous remarks on the electric fluid, and the experiments on the same subject contained in my former paper and the first parts of this, are connected with meteorology.

Art. XI. I have said that, as we ascend mountains, the hygrometer, successively falling, indicates less and less evaporated water in the air. We thus, however, attain the region in which clouds and rain are formed; and there it is, that the lessons of nature itself may guard us against the arbitrary dictates of imigination: I shall therefore relate what I have observed. At times when the atmosphere is

so clear, that distant objects are seen very distinctly, and that the hygrometer, according to the tables that M. de Saussure and I have made from direct and separate experiments, does not indicate above two or three grains in each cubic foot of that air, small clouds may be seen forming in all parts of the very stratum of the atmosphere in which we stand, with very little or no wind. Sometimes without any change in the temperature or moisture of the intermediate parts, these embryoes of clouds dissipate: but at other times they rapidly increase, unite together in the whole stratum in sight, and announce to the observer, that soon he will be enveloped by clouds. However, till the clouds, either moving towards him, or forming around him, occupy the very spot in which stand the hygrometer and the thermometer, he observes no sensible change in them but the instant that a cloud envelopes him the hygrometer arrives at its point of extreme moisture, and all the bodies are wet. Art. XII. These preliminaries of rain often remain a long time, with only some variations, and at last dissipate without effect; and as soon as the clouds disappear in one spot, the hygrometer indicates the same dryness, as if no cloud had been there. But at last; though without any perceptible difference in the preliminaries, because some other test of the state of the air, besides those we possess, is wanting, the clouds increase in extent and thickness, above and below the place of observation, and rain is produced in more or less abundance. If rain be lasting, and at the same time in a great extent of country, it may happen either in a calm air, or during some regular wind. But when rain is partial and in showers, sudden, and sometimes violent winds accompany these, arising from the expansion of the air, by its decomposition into aqueous vapour in some place, while a vacuum is produced in other parts by the resolution of that vapour into rain. Hence it is, that the direction of these winds is rapidly changing, and that they cease with the return of the transparency of the air. Lastly, in a stratum of air, which perhaps only half an hour before was calm and transparent, in which the hygrometer did not indicate any increase in the small quantity of evaporated water, and without any indication of increase of the quantity of electric fluid, some clouds, rapidly forming, produce lightning, thunder, hail, torrents of rain, and such violent winds, as tear up trees and overturn cottages on mountains.

We may be for ever ignorant of the causes of these wonderful phenomena, but those who are aware that fiction, in the operations of nature, may lead to great errors, will prefer ignorance to a false science. As for me, from my first observations of these operations of unconstrained nature, and with the addition of a remark of M. de Saussure, which I shall mention, I changed very essentially my former ideas on the atmospheric phenomena, as I have explained in my works, and shall repeat hereafter.

Art. XIII. In order to evade the general consequence which, in my works, I have deduced from these facts, namely, that rain and

the other concomitant phenomena are produced by different kinds of decompositions of the atmospheric air, which consequence is certainly the subversion of the new theory of chemistry, M. Fourcroy invented the hypothesis of a dry solution of water by air; supposing that this water could no longer affect the hygrometer, which, in consequence he discarded from the rank of a meteorological instrument, and having obtained the assent of many chemists who have not applied to meteorology any more than himself, this instrument, so much wished before by natural philosophers, is now hardly mentioned.

But this hypothesis, grafted on that of Le Roy, is in the first place absolutely gratuitous; no fact having been adduced in bringing it forward in chemistry against the positive facts contained in M. de Saussure's works and mine: and besides it is of no avail, since M. Fourcroy himself, and all those who have adopted it, have been obliged to suppose that this pretended solution of the water remains dependent on the temperature; which they are obliged to do, otherwise it would be nothing more than my system, with the appearance of refuting it. For, if the enormous quantity of water which sometimes falls in rain from a very limited stratum of air, be not submitted to precipitation by the diminution of heat, it must have been changed into a permanent or aeriform fluid; and in the atmosphere no sensible quantity of any fluid of this kind exists, but the atmospheric air. Besides, since for this reason it is supposed in that hypothesis, that the evaporated water remains dependent on temperature, very little knowledge in hygrology is required to conclude, that it cannot cease to affect the hygrometer in proportion to its quantity, as is evident from M. de Saussure's experiments and mine. Lastly, with respect to that fluid the decomposition of which produces rain, its nature is clearly determined by the following circumstance: when we remain in a stratum of air till the end of the operations by which a deluge of rain, even with lightning and thunder, has been produced, the residuum, according to all tests, is the same air as before. Such are the objections which I have made to M. Fourcroy himself, and which he has not answered, nor any chemist for him.

Art. XIV. These formations and modifications of clouds, when viewed only over head from the plain, have naturally inspired the idea, that by some cause the liberation and condensation of evaporated water now and then take place in a great extent of the upper region of the atmosphere, which water descends and accumulates in the stratum of air where clouds form and produce rain. But this idea proceeds from a want of previous knowledge in hygrology, and of observations on high mountains; for, in the first place, whenceever and from whatever cause that quantity of water may be supposed to proceed before any precipitation can take place, even in the first state of vesicular vapour which constitutes clouds, it must be preceded by extreme moisture in the still transparent air, since it is only the excess of that water, which is first precipitated in a mist;

and when this precipitation ceases, extreme moisture still subsists in the air, as M. de Saussure and myself have found in all our hygroscopical experiments. Now, I have said above, from observations on high mountains, that air is there dry till the moment before the formation of clouds, and that as soon as the clouds are dissipated, the hygrometer indicates the same dryness as before. This evidently shows, that the production of clouds and rain have their cause in the very stratum of air where they are manifested; and this cause cannot be any other than a decomposition of the air itself.

Lastly, in these very clouds, which, being themselves a conducting mass, lean besides against mountains, it happens sometimes that lightning and thunder are produced; and this, as I have said before, without any previous sign of an uncommon quantity of electric fluid in them. This also points out some operation taking place in these clouds, by some modification in the cause which commonly produces a simple rain. The electric fluid thus suddenly disengaged must have been before in some chemical combination in the air itself, which prevented its manifestation, and is then destroyed. When we are above the clouds, we may see (as it has happened to me) lightning darting upwards, as it is commonly seen darting downwards when we are under the clouds; and even in this last case, we may judge that lightning is darted upwards, when we see only a great sudden light in the clouds, without any flash, followed however by thunder.

The above are leading facts in the maze of atmospheric phenomena, certainly indicating the existence in the atmosphere of subtle fluids besides those which have hitherto been discovered. This is the general object which I am going to examine.

I shall here begin by explaining one of the results of my long labours in the pursuit of the measurement of heights by the barometer, of which all the steps are described in my work Researches sur les Modifications de l'Atmosphere, published in 1772. My experiments and observations were first directed towards these two points. 1. To obtain, by a great number of observations at different measured heights, on mountains and towers, a coefficient expressive of height, to the determined law of the densities of air correspondent to the difference of pressure in a given temperature of the air. 2. An equation for the differences of actual temperatures with that fixed point. By dint of trials, I arrived so far in these determinations, that the method of measuring the height of mountains by the barometer has been found preferable to the geometrical operations, on account of the impossibility of determining a law in the terrestrial refractions; besides the difficulty of finding proper bases for the triangles. But though the measurement of heights had been my first view in this undertaking, other modifications of the atmosphere became soon predominant in my pursuit, as I shall now explain.

In order to ascertain the degree of exactness, which could be ob

tained in the determination of the two parts of the formula above defined, I had measured the heights of fourteen points above one another on Mount Salève, near Geneva; the whole height above the level of the base that I had chosen being above 3,000 feet; and at each of these determined elevations I had made a great number of observations of the barometer in different temperatures, both in the same days and in different seasons. I had taken all possible precautions to ascertain the height of each of these points, in verifying the trigonometrical operation by levelling the whole mountain in passing by these points, in order that the formula might be applicable to other places but had there been some inaccuracy in that respect, it could not effect the coefficient of the two laws, as applied to the same places; for if these laws had been sufficient, the formula would have assigned to them the same height by every observation. Now, no coefficient to the differences of heights in the barometer, associated with any equation for the differences of the thermometer, could bring the formula to express the same differences of height between the same points: a proof that the two conditions, with which alone the formula corresponds, do not embrace all the causes of variation in the density of air. Having, however, no other data, I fixed these two parts of the formula in the manner the most correspondent to the whole of my observations, amounting to near 600; so that the sums of anomalies in plus and minus comparatively to the measured heights were equal; after which the causes of these anomalies became the object of my researches.

One of the principal means by which I had considerably reduced the former great anomalies in this measurement, which had appeared unconquerable, had been by introducing an equation for the differences of expansion of air produced by heat in the atmosphere. Considering therefore this effect of heat, which by its increase diminishes the pressure of columns of air of a determined height, under the same pressure of superincumbent air, as indicated by the height of the barometer at the upper station; and connecting this circumstance with the idea, that the cause of heat is an expansible fluid, namely free fire, which occupies in air a space without any sensible addition to its mass; I concluded that some other fluid, for which we had not yet a test, as we have by the thermometer for the former, might be the cause of the above remaining anomalies.

This general conclusion brought into my mind the aqueous vapour, of which I knew that the specific gravity was much less than that of air; and supposing at that time, as was commonly thought, that its accumulation in the atmosphere was the cause of rain, I conceived that the difference of its quantity in different times must be very great, and that this might be the cause, or at least one of the causes of the anomalies I had in view.

The same consideration led me also to a system concerning the remarkable, though not constant, correspondence of the variations in the sedentary barometer with rain and fine weather; as the same