of the ice-cap that covers the South Polar lands (see par. 268).

THE ATMOSPHERE.

196. Elasticity of the Atmosphere.-The atmosphere (from Gr. atmos, smoke or vapour, and sphaira, a globe) is as much a part of the globe as the ocean is. It forms the uppermost stratum, as it were, and partakes of the annual and diurnal motions. The matter of which it is composed is highly elastic, as is distinctly felt in using a pop-gun. The lower strata are consequently denser than the higher; just as is the case when you make a pile of light fluffy matter like wool or straw. It results from this that at the top of Mont Blanc, only three miles above sea-level, half the weight of the atmosphere is left below, and a given space, say a cubic foot, contains only half the mass of air that it contains at sealevel. The rarefaction goes on increasing in geometrical ratio with the ascent, until the expansive force becomes so weak that it is balanced by the weight of the particles. At what height this takes place has not been exactly determined. A height of 7 miles has been reached in a balloon, and there the air was so rare that it was hardly possible to breathe; and a pigeon dropped from the car, fell downwards, like a stone in water, till it reached air dense enough to resist its wings.

197. Height of the Atmosphere.-After the sun sets, his rays continue to strike the upper part of the atmosphere, and the particles of the air reflect or send some of the rays earthwards. It is these faint rays that constitute twilight. Now, this twilight region is observed to terminate at a height of about 50 miles, and it is inferred that beyond this the air is too thin to reflect

any light. But that the atmosphere extends far above this height is clear from the appearance of meteors. Meteors or shooting-stars are dark bodies careering through space, and that burst into light when they chance to enter our atmosphere. Their velocity is such that the friction against the air-particles heats them to whiteness. Now, it has been found by calculation that meteors come into view usually at a height of 70 miles, but occasionally they are observed as high as 200 miles. There must therefore be air at this height, however rare. 198. Weight of the Atmosphere.-The substance of the air is invisible, but we feel it resisting us when we

Fig. 31.

move a flat surface rapidly against it; and when it itself is in motion, we feel it as wind. Air has weight like everything material. A cubic foot of air at the earth's surface weighs rather more than an ounce. The weight of the whole atmosphere is such that it presses with a force of 15 lbs. (more nearly 142) on every square inch. That we are quite insensible of this is owing to the pressure being the same on all sides.

199. How Measured.-The amount of the pressure of the atmosphere is thus determined: Take a glass tube not less than 33 inches long, and open at one end; fill it with mercury; close the open end firmly with the thumb, invert the tube, and place it vertically with the end immersed in a cistern of mercury. When the thumb is removed, the mercury in the tube sinks, however long the tube is, till it stands about 30 inches above the level of the mercury in the basin. The space above the mercury in the tube is a vacuum. The whole

Torricelli's Experiment.

weight of this column in the tube presses on the mercury in the basin, and according to the law of hydrostatics, that pressure is propagated in all directions, and would force up the substance of the mercury were it not balanced by an equal pressure downwards. That pressure can come only from the atmosphere. That it is so was proved by Pascal carrying the apparatus to the top of a mountain; the height of the column was found to become less and less, the greater the elevation, because more and more of the atmosphere was left below.

If the area of the tube is a square inch, the column of mercury weighs about 15 lbs., which must therefore be the weight of a column of air having a square inch base, and reaching to the top of the atmosphere. If water were used in the experiment instead of mercury, the column would stand at the height of about 34 feet, the weight of water being nearly 14 times less than that of mercury. It is easy to find the weight of the whole atmosphere in pounds by multiplying the number of square inches on the surface of the globe by 15; it has been calculated to amount to 77,670,000,000,000,000 lbs. An ocean of mercury of the uniform depth of 30 inches, or one of water 34 feet deep, covering the face of the globe, would have the same weight.

200. The Barometer.-In the experiment of Torricelli with the mercury and tube (see par. 199), we have the essential principle of the barometer (from two Greek words signifying weight and measure); and when the apparatus is fixed in a frame, and provided with a scale to mark the height of the column, it forms what is called the cistern barometer, which is still the best where accuracy is required.

201. The Aneroid Barometer.-In this form of barometer no liquid is employed, and hence the name

H

(Gr. a priv., neros, wet). A flat circular metal box is nearly exhausted of air, and soldered air-tight. The external air causes the flat sides of the box to bend inwards more or less, according to varying pressure of the atmosphere, and a system of levers connected with the sides, to make an index hand move over a graduated scale. The scale is graduated to represent the inches of the mercurial barometer, and the results correspond with wonderful accuracy; but as the instrument is liable to change, it is necessary to make frequent comparisons with the ordinary barometer.

DISTRIBUTION OF PRESSURE.

202. The pressure of the atmosphere varies considerably from place to place, and from time to time, and these variations are the chief cause of all fluctuations of wind and weather. A knowledge, therefore, of the distribution of pressure over the globe, and the laws of the changes, as indicated by the barometer, is the basis of meteorology. Fluctuations of pressure arise from two causes-differences of temperature, and differences in the quantity of water vapour.

(1) Pressure affected by Temperature.-When a portion of the earth's surface is heated more than the surrounding regions, the air expands, becomes lighter, and flows over into those regions. The column of air over the heated area thus loses part of its mass, and has less pressure. The opposite effect is produced by cold. Over a cold region the air contracts and sinks down, and additional air flows in above, thus increasing the weight of the column.

(2) Pressure affected by Vapour.-Vapour has less specific gravity than air; a mixture of air and vapour

therefore weighs less than the same bulk of dry air. When the air, then, contains a deal of vapour, the weight of the whole column-in other words, the barometric pressure must be less than when the air is comparatively dry.

203. The Movements thus caused-Cyclones.-Whenever a difference of pressure is thus established between two atmospheric columns from one or both of these causes, the column of high pressure must flow out at the base towards the region of low pressure. A current of air flowing into an area of low pressure, does not proceed straight to the centre, but (owing to a cause explained in par. 255) in a spiral direction, circling round. and round it, and always getting nearer. On account of this circling of the air within it, an area of low pressure, or to express it shortly, a depression, is called a Cyclone, and the movement, a cyclonic movement. An area of high pressure is called in contrast an anti-cyclone. It is within the areas of depression that storms and foul weather occur; within an anti-cyclone, calm weather or gentle breezes prevail.

204. Isobars.-The distribution of atmospheric pressure is best exhibited by means of isobarometric charts; that is, charts on which lines, called isobars, are drawn through all places where the height of the barometer is the same. Such a chart may represent the mean pressure for the year, or the mean for a portion of the year, such as a particular month; or the readings of the barometer at a specified moment, such as the beginning or middle of a storm. A complete series of charts representing by isobars the mean pressure over the globe for the year, and for every separate month, has been constructed by Mr Alexander Buchan, Secretary to the Scottish Meteorological Society.