That atmospherical air is a heavy, compressible, and elastic substance, may be proved by many simple and direct experiments. A bladder filled with air is heavier than when it is in its flaccid state. When subjected to compression, it may be made to occupy a smaller space ; and when that pressure is removed, its elasticity enables it to resume its original size. Since the air is heavy, the lower strata of the atmo sphere are compressed by the weight of the super incumbent mass. The lowest stratum supporting the weight of almost the whole atmosphere, will be more dense than the rest ; and the superior strata will gradually become more rare, in proportion to the weight which they sustain. The air in the higher regions, therefore, will be extremely rare, on account of its elasticity, which is not checked by any super incumbent pressure. If the air were perfectly elastic, it is obvious that there would be no limit to its 'ex pansion, and that the whole atmosphere would be dis sipated through infinite space. The elasticity of the atmosphere must consequently diminish in a greater ratio than the weight which compresses it, and there must be a certain state of rarity at which its elasticity ceases. Upon the supposition that the rarity of the air is reciprocally proportional to its superincumbent weight, it may be demonstrated, that if the heights in the atmosphere be taken in arithmetical progression, the rarity of the air at these heights will be in geome trical progression, or, what is the same thing, the al titudes in the atmosphere are as the square roots of the corresponding rarities. Hence we have a method of measuring differences of altitude, by ascertaining with the barometer the rarity of the air at two places whose vertical distance is required. A full account of this method will be found under the article BARD M ETER.
The weight and pressure of the atmosphere may be ascertained by very simple experiments. If we im merse in water a glass tube open at both ends, the water included in the tube will be on the same level with the fluid which surrounds it. When we apply our mouth to the upper end of the tube, and draw out the air, the included water instantly ascends till the weight of the elevated column added to the elas ticity of the remaining air, exactly balances the pres sure of the atmosphere on the surrounding- fluid. If we now take a long tube, 40 feet long for example, shut at one end, and having filled it with water, plunge the open end into a vessel of water, the fluid will then descend in the tube till the weight of the column exactly equals the pressure of the atmosphere ; for the air is now excluded from the upper part of the tube, and the weight of the column of fluid is the only force which is left t6 balance the weight of the atmospherical column. By making this experiment,. it will be found that the water stands at from 34 to.
35 feet above the general level of the surrounding fluid, and therefOre the weight of a column of air reaching to the top of the atmosphere, is equal to the weight of a column of water, of the same babe, with the altitude of 31. feet, or about 2156 t.1 pounds •on a square foot, or 15 pounds on every square inch. This experiment may be more easily made by using quicksilver instead of water. The quicksilver will
rise to the height of 29 inches in the tube, and will thus measure the pressure of the atmosphere. Hence it follows, that the whole atmosphere ex erts the same pressure on the surface of the earth, as if the surface of the globe were covered with water to the depth of 34 feet, or with quicksilver to the depth of 29 inches. This pressure has been computed at 12,02.2,560,000,000,000,000 pounds, or as equivalent to that of a globe of lead 60 miles in diameter ; and if we suppose that a man's body exposes a surface of nearly 15 square feet, he will sustain a pressure of 323•.31 pounds, or 144 tons.
From the changes which take place in the atmo sphere, its pressure is liable to very considerable va riations. The column of quicksilver which we have shown to be a measure of that pressure, varies from 28 to 31 inches. The cause of these changes, which are yet but imperfectly known, will be considered tin der the porticle M•TEQROLOGT.
If the atmosphere were of uniform density, it would be easy to ascertain, with the utmost accuracy, the height to which it extends ; for the height of the at mosphere would obviously be to the height of the mercury in the barometer, as the specific gravity of common air is to the specific gravity of mercury. By making the calculation on this supposition, it will be found that the height of the atmosphere is a little more than 5 miles. As the air, however, gradually diminishes in density, the atmosphere must reach to a much greater distance from the earth than 5 miles: It appears from the duration of twilight, that at the height of 444. miles, the atmosphere is sufficiently dense to intercept the light of the sun, and reflect it to the earth. We are therefore entitled to conclude, that it extends to a much greater height.
When a ray of light enters the atmosphere, it is bent from its course by the same cause which re fracts the rays of light when they pass through any dense medium, such as glass or water. The refrac tion sustained by light at its first entrance into the atmosphere must he very small, from the extreme rarity of the air. The deviation, however, will gra dually increase as it penetrates the denser strata, and the ray will describe a path increasing in curvature as it approaches the earth. From this property of the atmosphere, the apparent altitude of the sun, moon, and stars, is greater than their real elevation, and they appear to be raised above the horizon when they are actually below it. The refraction of the atmosphere near the earth's surface is liable to very considerable anomalies. A very extraordinar(phe nomenon arising from this cause has been described by Mr Vince. The castle of Dover, concealed by the hill which lies between it and Ramsgate, appeared, on the 6th of August :806, as if it had been brought over and placed on the side of the hill next to Rams gate. This phenomenon must have arisen from some variation of density in the intermediate air. Pheno mena of the same class with the preceding have been illustrated experimentally by the ingenious Dr Wol laston. See Edinburgh Transactions, vol. vi. p. 215. ; and Phil. Trans. 1778, p. 857 ; 1798.