bottles, when inflated with air, weighed more upon a balance than when they were empty and flaccid. The like experiment may now be tried very cheaply on a blown bladder, which will be found sensibly heavier than when the air is expelled. In a proper vessel of copper we can weigh air of different densities, and compare them together; or we can exhaust the air, and weigh a vacuum, at least by comparison. By inany statical experiments, which have been made with the greatest circumspection, it is found that the weight of air, under a mean state of the atmosphere, is to the weight of water as 1 to 850. But the weight of air is very different at different times; and this difference proceeds from two causes. When air is more rarefied by heat, it is lighter; when it has more moisture, it is heavier. There is a difficulty in this part of the subject, which I could never clear up to my own satisfaction. All other matter, but that which is the cause of gravity, must gravitate; therefore the grosser air must gravitate. Yet the gravity of those matters which are incorporated with the air, interfere so much in our experiments, that it is hard to draw the line between the native and the adventitious gravity of air.

Air is 850 times lighter than water; so that water 850 times rarefied in vapour will have the same gravity as air: and if air should contain water about 1000 times rarefied, then the whole weight we attribute to air will nearly be due to the water contained in it. I have found cold moist air weigh heavier than the drier air of an apartment; and this under such circumstances, that I could not but impute the difference chiefly to the moisture. It therefore seems to me so difficult to adjust the gravity of simple air, that at present I am obliged to leave it for farther examination.

That air contains much water may be shewn by many experiments. The slacking of quick lime, by exposing it to the air, shews us that water enough is applied for this purpose by the air itself. Such alkaline salts as are of a more fiery nature than lime, will shew this sooner and better. Light clouds of water may be seen to separate from the air, when it becomes rarefied about or by exhaustion in the receiver of an air-pump; and if a glass be never so clean and dry when the air is exhausted from it, the air, when re-admitted, never fails to deposit a strong dew all over its internal surface. The air which

enters

enters first is instantly rarefied, and its water is separated from it, till the quantity which follows restores the equilibrium. Hence it happens, that we are obliged to wipe the inside of the glass after a few experiments, if it is required that the vessel should be pellucid. To shew this more sensibly, let a mixture of pounded ice and salt be put into any kind of vessel, and stirred about for some time. Though the air of the room be ever so pure and dry in appearance, a frozen dew will gather apace on the outside of the vessel, which, in a short time, will increase to a thick shining crust of hoar frost, with which the outside of the vessel will be in vested, as high as the freezing composition reaches within side. The reason here is the same as before; fresh clouds of moisture are presented every moment to the outside of the vessel, by reason of the fire rushing into. it to restore the equilibrium; and in passing it leaves the air and water behind at the surface, where fresh quantities of water are accumulated and fixed, as the freezing proceeds. The sweating of marbles and pave ments in warm moist weather is to be accounted for in the same manner. During a course of cold weather, the stone pavements

and walls of a house, with other solid objects, become very cold; but when the warmer air enters the house at a change of weather, its fire penetrating cold bodies, to reduce them to the general temper, deposits its vapour at the surface in its passage. Every fluid that is warm will sustain more of any foreign matter than the same fluid when cold; therefore warm air will sustain more water than cold air. This is partly owing to the greater agitation of bodies with a greater degree of heat. Light bodies will float about in disturbed air, which settle to the earth when the air is at rest: therefore, in certain cases, nothing more is requisite than a greater degree of coldness to make the air drop its moisture. In a frosty night, when the air abroad is colder than the air within, the dampness of the internal air settles on the glass panes of the windows, and is frozen within side into beautiful forms, of which we shall speak more particularly hereafter.

The rays of light from the sun, moon, and stars, suffer a refraction in passing to a spectator upon the earth, through the medium of the atmosphere, so that every object in the heavens appears above its place; or, in other terms, its apparent altitude, by refraction, is

greater

greater than its true altitude: This refraction is greatest of all at the horizon, and decreases very fast as the altitude increases; insomuch that the refraction at the horizon, differs from the refraction at one degree above the horizon, by about one-third part of the whole quantity. At the horizon, in this climate, it is found to be 33' *. In climates nearer to the equator, where the air is purer, refraction is less; and in the colder climates, nearer to the pole, it increases exceedingly, and is a happy provision for lengthening the appearance of the light to those regions so remote from the sun. Gassendus relates, that some Hollanders, who wintered in Nova Zembla, in lat. 75°, within 15 degrees of the pole, were surprised with a sight of the sun 17 days before they expected him in the horizon. This difference was owing to the greater refraction of the atmosphere in that latitude;

VOL. IX.

E E

A table of refractions, calculated by Sir Isaac Newton, and given to Dr. Halley, is to be found, together with some observations upon it, in Phil. Trans. Abr. vol. vi. P. 167. This table was afterwards much improved by Dr. Bradley. I have a table of refractions in manuscript, under the name of Mr. Flamstead, which is very nearly the same with that afterwards published by Dr. Halley. Dr. Bradley's table is calculated for that state of the air when the barometer is at 29.6, and the thermometer at 50° of Fahrenheit.