the course of the barometer is generally in the opposite direction to that of the thermometer; that is, that when the temperature rises the barometer falls, and vice versâ; which indicates that the barometric variations at any given place are produced by the expansion or contraction of the air, and therefore by its change in density. If the temperature were the same throughout the whole extent of the atmosphere, no currents would be produced, and, at the same height, the atmospheric pressure would be everywhere the same. But when any portion of the atmosphere becomes warmer than the neighbouring parts, its specific gravity is diminished, and it rises and passes away through the upper regions of the atmosphere; whence it follows that the pressure is diminished, and the barometer falls. If any portion of the atmosphere retains its temperature, while the neighbouring parts become cooler, the same effect is produced; for in this case, too, the density of the first-mentioned portion is less than that of the others. Hence, also, it usually happens, that an extraordinary fall of the barometer at one place is counterbalanced by an extraordinary rise at another place. The daily variations appear to result from the expansions and contractions which are periodically produced in the atmosphere by the heat of the sun during the rotation of the earth.

126. Relation of barometric variations to the state of the weather. It has been observed that, in our climate, the barometer in fine weather is generally above 30 inches, and is below this point when there is rain, snow, wind, or storm, and also, that for any given number of days at which the barometer stands 30 inches, there are as many fine as rainy days. From this coincidence between the height of the barometer and the state of the weather, the following indications have been marked on the barometer, counting by thirds of an inch above and below 30 inches:

In using the barometer as an indicator of the state of the weather, we must not forget that it really only serves to measure the weight

-127]

Wheel Barometer.

113

of the atmosphere and that it only rises or falls as this weight increases or diminishes; and although a change of weather frequently coincides with a change in the pressure, they are not necessarily connected. This coincidence arises from meteorological conditions peculiar to our climate, and does not always occur. That a fall in the barometer usually precedes rain in our latitudes, is caused by the position of Europe. The south-west winds, which are hot, and consequently light, make the barometer sink; but at the same time as they become charged with aqueous vapour in crossing the ocean, they bring us rain. The winds of the north and north-east, on the contrary, being colder and denser, make the barometer rise; and, as they only reach us after having passed over vast continents, they are generally dry.

When the barometer rises or sinks slowly, that is, for two or three days, towards fine weather or towards rain, it has been found, from a great number of observations, that the indications are then extremely probable. Sudden variations in either direction indicate bad weather or wind.

127. Wheel barometer.-The wheel barometer, which was invented by Hooke, is a syphon barometer, and is especially intended to indicate good and bad weather (fig. 101). In the shorter leg of the syphon there is a float a, which rises and falls with the mercury (fig. 102). A string attached to this float passes round a pulley, and at the other end there is another and somewhat lighter float. A needle fixed to the pulley moves round a graduated circle, on which is marked variable, rain, fine weather, etc. When the pressure varies the float sinks or rises, and moves the needle round to the corresponding points on the scale.

The barometers ordinarily met with in houses, and which are called weather glasses, are of this kind. They are, however, of little use, for two reasons. The first is, that they are neither very delicate nor precise in their indications. The second, which applies equally to all barometers, is, that those commonly in use in this country are made in London, and the indications, if they are of any value, are only so for a place of the same level and of the same climatic conditions as London. Thus a barometer standing at a certain height in London would indicate a certain state of weather, but if removed to Shooter's Hill it would stand half an inch lower, and would indicate a different state of weather. As the pressure differs with the level and with geographical conditions, it is necessary to take these into account if exact data are wanted.

I

128. Determination of the heights of places by the barometer. One of the most important of the uses of the barometer has been its application to the measurement of the heights of places above the sea level. For, if we suppose the atmosphere

divided into horizontal layers of equal thickness, a hundred, for instance, a barometer at the sea level would support the weight of a hundred of these layers; and, as we have seen (116), would be at rest when its height was thirty inches. If it were raised in the atmo-,

-129]

Height of the Atmosphere.

115

sphere to the height of ten such layers, it would now only support the weight of ninety such layers, and the mercury would therefore necessarily sink. It would sink still further if it were raised to the twentieth layer, and so on to the limit of the atmosphere if that were possible. There it would be under no pressure, and the level of the mercury in the tube and in the cistern would be the same.

As the mercury sinks in proportion as we rise in the atmosphere, we might, from the amount by which it is lower, deduce the height above the sea level. If air had everywhere the same density up to the extreme limit of the atmosphere, the calculation would be very simple; for as mercury is about 10,500 times as heavy as air, an inch of the barometer would correspond to a column of air about 875 feet; hence, in ascending a mountain, a diminution of an inch in the height of the barometer would correspond to an ascent of about 875 feet. But the density of the air decreases as we ascend, for the layers of air necessarily support a less weight; hence, the measurement of the heights by the barometer is not so simple as we have supposed. Very complete tables have, however, been constructed, by which the difference in height between any two places may be readily ascertained, if we know the corresponding height of the barometer. For small elevations we may assume that an ascent of 900 feet produces a depression of an inch in the height of the barometer.

129. Height of the atmosphere. In virtue of the expansive force of the air, it might be supposed that the molecules would expand indefinitely into the planetary spaces. But, in proportion as the air expands, its expansive force decreases, and is further weakened by the low temperature of the upper regions of the atmosphere, so that, at a certain height, an equilibrium is established between the expansive force which separates the molecules, and the action of gravity which draws them towards the centre of the earth. It is therefore concluded that the atmosphere is limited. From the weight of the atmosphere, and its decrease in density, and from the observation of certain phenomena of twilight, its height has been estimated at from 30 to 40 miles. Above that height the air is extremely rarefied, and at a height of 60 miles it is assumed that there is a perfect vacuum. From certain observations recently made in the tropical zone, and particularly at Rio Janeiro, on the twilight arc, M. Liais estimates the height of the atmosphere at between 198 and 212 miles, considerably higher, therefore, than what has hitherto been believed.

ILLUSTRATIONS OF ATMOSPHERIC PRESSURE.

130. The pressure of the atmosphere is transmitted in all directions.-The atmosphere, like any other mass of fluid (75) must necessarily transmit its pressure in all directions, upwards and laterally as well as downwards. We have already seen a striking instance of this in the Magdeburg hemispheres, (115) and the following experiment furnishes another illustration of this point.

A tumbler full of water is carefully covered with a sheet of paper, which is kept in position by one hand, while with the other the tumbler is inverted. Removing then the hand which held the paper, the water does not fall out, both water and paper being kept in position by the upward pressure (fig. 103). The object of the paper is to present a flat surface of water, for otherwise the water would divide and would allow air to enter, and then the experiment would fail.

The use of the wine-tester also depends on the pressure of the

atmosphere. It consists of a tin tube (fig. 104), terminating at the bottom in a small cone, the end of which, o, is open; at the top there is a small aperture, which is closed by the thumb. The two ends being open, the tube is immersed in the liquid to be tested; closing then the upper end by the thumb, as shown in the figure, the tube is withdrawn, and remains filled in consequence of the pressure at o. But if the thumb be withdrawn the pressure is transmitted both upwards and downwards, and the liquid flows out in obedience to the action of gravity.

131. Pressure supported by the human body. The surface of the body of a man of middle size is about 16 square feet; the pressure, therefore, which a man supports on the surface of his