each branch, and in front of each is a movable index, which may be raised or depressed until it comes to the free surface of the mercury in each branch. By means of these scales the difference of level in the two branches may be measured. This difference is the height of the barometric column.

To prevent violent oscillations when the instrument is moved from place to place, the two branches communicate through

a

a fine, almost capillary tube. This
arrangement also prevents the pos-
sibility of a bubble of air penetrating
from the shorter to the longer branch,
when the instrument is inclined.

174. The Wheel Barometer. This is a form of the SIPHON BAROMETER in which the rise and fall of the mercury are shown by the movements of an index around a graduated circle. The manner in which it acts is shown in Fig. 122.

The index is attached to an axis which bears a pulley. Passing over this pulley is a fine wire, at one end of which is attached an iron weight, a, which rises when the height of the mercury diminishes, and falls when this height increases. At the second extremity is a counterpoise, b, which keeps the wire tense, and causes the wheel to turn as the weights rise and fall.

Fig. 122.

Fig. 123.

Fig. 123 shows its external appearance with a thermometer attached.

It will be seen that a slight change in the level of the mercury in the tube will produce a considerable movement of the index.

Notwithstanding this advantage, this form of barometer is of little value when accurate observation is required. The iron weight, a,

is somewhat heavier than the counterpoise, b, and thus there is a slight force in addition to the pressure of the air, which acts to sustain the column of mercury. Again, when the mercury in the shorter branch tends to rise, it must overcome the excess of weight in a, and consequently very minute changes of pressure are not recorded by this instrument.

175. The Aneroid Barometer. The action of this curious instrument depends upon the effect produced by atmospheric pressure upon a metallic box from which the air has been partially exhausted. Its appearance and construction are shown in Fig. 124.

An increased atmospheric pres-x sure tends to force the cover inward; but when the atmospheric pressure diminishes it is pressed outward by its own elastic force, aided by a spring in the interior. The movements of the cover, transmitted by a combination of delicate levers, cause an index to move over a graduated scale.

Fig. 124.

Being very easily portable, this form of barometer has lately come into extensive use, especially for measuring the heights of mountains.

Instruments of this kind are now made that may be carried in the pocket like a watch, and they are so sensitive to slight changes of pressure that they will indicate a change of level of not more than three or four feet.

176. Causes of Barometric Fluctuations.. Since the mercury in the barometer is sustained by the weight of the column of air above it, changes in the weight of this column of air will produce changes in the height of the mercurial column. Such changes are constantly going on, and consequently the barometer is continually fluctuating.*

* The atmosphere surrounds the earth like an immense ocean, nearly fifty miles in depth. It is never at rest, but has its great currents and tides; and, like the ocean of water beneath, it is agitated by storms, and

Certain very slight changes occur regularly; thus, there is a daily variation by which the mercury stands highest at ten o'clock morning and evening, and lowest at four o'clock afternoon and morning. These changes are greatest at the equator, where the variation is about one tenth of an inch. In latitude 40° it is 0.05 inch, and in lat. 70° only 0.003 inch.

There is also an annual inequality dependent on the seasons. In this country it is scarcely perceptible, but in China, and throughout a large part of Asia, the average height of the mercury is three fourths of an inch greater in January than in July.

The greater changes in the weight of the atmosphere are not periodical, but depend upon changes of temperature. When the temperature at any place is elevated, the air expands and rises until its lateral tension is greater than that of the surrounding air, when it flows away to the neighboring regions. When, on the contrary, the temperature is diminished, the air contracts and an additional quantity flows in from the neighboring regions.

The barometer, then, falls where there is a dilatation, and rises where there is a contraction of the air.

177. The Barometer as a Weather-Indicator.—The barometer is often called a weather-glass, and the scale of the instrument is sometimes inscribed with words intended to indicate the weather that may be expected when the top of the column stands opposite them. This, however, conveys an incorrect notion, for a change in weather is not indicated by the absolute height of the mercury at any given time.

Moreover, there are other conditions besides the weight of the atmosphere, which are quite as important as this, for the prediction of the weather. The temperature, the amount of moisture in the atmosphere, and the force and direction of the wind, are all to be considered as elements of the problem.

It is true, however, that changes in the heat, the moisture, moves in immense waves. When the crest of one of these waves is over the barometer the mercury rises, and it falls again as the depression follows the crest of the waves.

or the movements of the air, are almost always accompanied, or immediately followed, by changes in the height of the barometer. Hence the changes in the height of the mercurial column may, to a certain extent, be relied on for predicting the weather. The following rules are generally reliable :

1. The rising of the mercury indicates the approach of fair weather; the falling of the mercury shows the approach of foul weather.

2. A great and sudden fall of the inercury precedes a violent storm of short duration.

3. If, during fair weather, the mercury falls continually for several days, a long succession of foul weather will probably follow; and, again, if during foul weather which continues for a long time, the mercury gradually rises, fair weather may be expected to follow and continue for several days.

4. A fluctuating and unsettled state in the mercurial column indicates unsettled weather.

178. Measure of Mountain Heights. One of the most important applications of the barometer is to the measurement of the height of any place above the level of the sea.

As we ascend above the level of the sea, the pressure of the air diminishes, and the barometer falls. Formulas have been deduced, by means of which the difference of level between any two places can be found, when we have the heights of the mercurial columns at the two places, together with the temperatures of the air and mercury at these places.*

* The exact rule for finding the height of a mountain by this method is rather complicated. Allowance must be made for temperature and for the latitude of each station; and other minor corrections are to be made. The following rule is given by Todhunter as nearly accurate for heights of not more than 3000 feet:

Observe the height of the barometer at the bottom and at the top of the mountain; divide the difference of the heights by their sum, and multiply the result by 52,428; this will give the height of the mountain in feet.

The above rule assumes that the temperature at each station is 32° Fahrenheit. The result is made more accurate by adding a thousandth part for every degree in the sum of the temperatures above 64°. Thus, if the temperature at the lower station is 60°, and at the higher 45°, the sum is 1050, which exceeds 64° by 41°. Therefore the result obtained by the rule should be increased by the 1 part of itself.

The following table shows the height of the barometer at different altitudes where observations have been made:

179. Pressure on the Human Body. - The pressure

Fig. 125.

on each square inch of the body is fifteen pounds; hence, on the whole body the pressure is enormous. If we take the surface of the human body equal to 2000 square inches, which is not far from the average in the case of an adult, the pressure amounts to 30,000 pounds, or fifteen tons.

If it be asked why the body is not crushed by this enormous pressure, the answer is, because it is uniformly distributed over the whole surface, and is resisted by the elastic force of air, and other gases, distributed through the tissues of the body.

The following experiment will show that the tissues of the human