Galileo (see opposite p. 72) for an explanation, the latter replied that evidently "nature's horror of a vacuum did not extend beyond 32 feet." It is quite likely that Galileo suspected that the pressure of the air was responsible for the phenomenon, for he had himself proved before that air had weight; and, furthermore, he at once devised another experiment to test, as he said, the "power of a vacuum." He died in 1642 before the experiment was performed, but suggested to his pupil Torricelli that he continue the investigation.

37. Torricelli's experiment. Torricelli argued that if water would rise 32 feet, then mercury, which is about 13 times as heavy as water, ought to rise but as high. To. test this inference he performed, in 1643, the following famous experiment:

3

(2)

(1)

Let a tube about 4 ft. long, which is sealed at one end, be completely filled with mercury, as in Fig. 25, (1), then closed with the thumb and inverted, and the bottom immersed in a dish of mercury, as in Fig. 25, (2). When the thumb is removed from the bottom of the tube, the mercury will fall away from the upper end of the tube, in spite of the fact that in so doing it will leave a vacuum above it; and its upper surface will, in fact, stand about 1 of 32 ft., that is, between 29 and 30 in., above the mercury in the dish.

FIG. 25. Torricelli's
experiment

Torricelli concluded from this experiment that the rise of liquids in exhausted tubes is due to an outside pressure exerted by the atmosphere on the surface of the liquid, and not to any mysterious sucking power created by the vacuum as is popularly believed even to-day.

38. Further decisive tests. An unanswerable argument in favor of this conclusion will be furnished if the mercury in the tube falls as soon as the air is removed from above the surface of the mercury in the dish.

A

To test this point, let the dish and tube be placed on the table of an air pump, as in Fig. 26, the tube passing through a

tightly fitting rubber stopper A in the bell jar. As soon as the pump is started the mercury in the tube will, in fact, be seen to fall. As the pumping is continued it will fall nearer and nearer to the level in the dish, although it will not usually reach it, for the reason that an ordinary vacuum pump is not capable of producing as good a vacuum as that which exists in the top of the tube. As the air is allowed to return to the bell jar the mercury will rise in the tube to its former level.

FIG. 26. Barometer falls when air pressure on the mercury surface is reduced

h

b

с

39. Amount of the atmospheric pressure. Torricelli's experiment shows exactly how great the atmospheric pressure is, since this pressure is able to balance a column of mercury of definite length. As the pressures along the same level ac (Fig. 27) are equal, the downward pressure exerted by the atmosphere on the surface of the mercury at c is equal to the downward pressure of the column of mercury at a. But the downward pressure at this point within the tube is equal to hd, where d is the density of mercury and h is the depth below the surface b. Since the average height of this

FIG. 27. Air column to top of atmosphere balances the mercury column ab

column at sea level is found to be 76 centimeters, and since the density of mercury is 13.6 grams per cubic centimeter, the downward pressure inside the tube at a is equal to 76 times 13.6 grams, or 1033.6 grams per square centimeter. Hence the atmospheric pressure acting on the surface of the mercury at c is 1033.6 grams, or, roughly, 1 kilogram per square centimeter. The pressure of one atmosphere is, then, about 15 pounds per square inch.

40. Pascal's experiment. Pascal thought of another way of testing whether or not it were indeed the weight of the outside air which sustains the column of mercury in an exhausted tube. He reasoned that, since the pressure in a liquid diminishes on ascending toward the surface, atmospheric pressure ought also to diminish on passing from sea level to a mountain top. As there was no mountain near Paris, he carried Torricelli's apparatus to the top of a high tower and found, indeed, a slight fall in the height of the column of mercury. He then wrote to his brother-in-law, Perrier, who lived near Puy de Dôme, a mountain in the south of France, and asked him to try the experiment on a larger scale. Perrier wrote back that he was ravished with admiration and astonishment" when he found that on ascending 1000 meters the mercury sank about 8 centimeters in the tube. This was in 1648, five years after Torricelli's discovery.

At the present day geological parties actually ascertain differences in altitude by observing the change in the barometric pressure as they ascend or descend. A fall of 1 millimeter in the barometric height corresponds to an ascent of about 12 meters.

41. The barometer. The modern barometer (Fig. 28) is essentially nothing more nor less than Torricelli's tube. Taking a barometer reading consists simply in accurately measuring the height of the mercury column. This height varies from 73 to 76.5 centimeters in localities which are not far above

sea level, the reason being that disturbances in the atmosphere affect the pressure at the earth's surface in the same way in which eddies and high waves in a tank of water would affect the liquid pressure at the bottom of the tank.

The barometer does not directly foretell the weather, but it has been found that a low or rapidly falling pressure is usually accompanied, or soon followed, by stormy conditions. Hence the barometer, although not an infallible weather prophet, is nevertheless of considerable assistance in forecasting weather conditions some hours ahead. Further, by comparing at a central station the telegraphic reports of barometer readings made every few hours at stations all over the country, it is possible to determine in what direction the atmospheric eddies which cause barometer changes and stormy conditions are traveling and hence to forecast the weather even a day or two in advance.

42. The first barometers. Torricelli actually constructed a barometer not essentially different from that shown in Fig. 28 and used it for observing changes in the atmospheric pressure; but perhaps the most interesting of the early barometers was that set up about 1650 by Otto von Guericke of Magdeburg (1602-1686) (see opposite p. 32). He used for his barometer a water column the top of which passed through the roof of his house. A wooden image which floated on the upper surface of the water appeared above the housetop in fair weather but retired from sight in foul, a circumstance which led his neighbors to charge him with being in league with Satan.

FIG. 28. The Fortin barometer

43. The aneroid barometer. Since the mercurial barometer is somewhat long and inconvenient to carry, geological and surveying parties

commonly use an instrument called the aneroid barometer. It consists essentially of an air-tight cylindrical box the top of which is a metallic diaphragm which bends slightly under the influence of change in the atmospheric pressure. This motion of the top of the box is multiplied by a delicate system of levers and communicated to a hand which moves over a dial whose readings are made to correspond to the readings of a mercury barometer. These instruments are made so sensitive as to

indicate a change in pressure when they are moved no farther than from a table to the floor. In the self-recording aneroid barometer, or barograph, used by the United States Weather Bureau (Fig. 29), several of the air-tight boxes are superposed for greater sensitiveness, and the pressures are recorded in ink upon paper wound about a drum. Clockwork inside the drum makes it revolve once a week. A somewhat different form of the instrument is used by aviators to record altitude.

QUESTIONS AND PROBLEMS

1. Why does not the ink run out of a pneumatic inkstand like that shown in Fig. 30?

2. If a tumbler is filled, or partly filled, with water, and a piece of writing paper is placed over the top, it may be inverted, as in Fig. 31, without spilling the water. Explain. What is the function of the paper?