the ancient Greeks and Romans knew that water would rise in a tube when the pressure within the tube was reduced, and hence they tried to obtain water from wells in this fashion, but the water could never be raised higher than 34 feet.  Let us see why water could rise 34 feet and no more.  If an empty pipe is placed in a cistern of water, the water in the pipe does not rise above the level of the water in the cistern.  If, however, the pressure in the tube is removed, the water in the tube will rise to a height of 34 feet approximately.  If now the air pressure in the tube is restored, the water in the tube sinks again to the level of that in the cistern.  The air pressing on the liquid in the cistern tends to push some liquid up the tube, but the air pressing on the water in the tube pushes downwards, and tends to keep the liquid from rising, and these two pressures balance each other.  When, however, the pressure within the tube is reduced, the liquid rises because of the unbalanced pressure which acts on the water in the cistern.

[Illustration:  FIG. 42.—­Water rises in the tube when the air is withdrawn.]

[Illustration:  FIG. 43.—­The air supports a column of mercury 30 inches high.]

The column of water which can be raised this way is approximately 34 feet, sometimes a trifle more, sometimes a trifle less.  If water were twice as heavy, just half as high a column could be supported by the atmosphere.  Mercury is about thirteen times as heavy as water and, therefore, the column of mercury supported by the atmosphere is about one thirteenth as high as the column of water supported by the atmosphere.  This can easily be demonstrated.  Fill a glass tube about a yard long with mercury, close the open end with a finger, and quickly insert the end of the inverted tube in a dish of mercury (Fig. 43).  When the finger is removed, the mercury falls somewhat, leaving an empty space in the top of the tube.  If we measure the column in the tube, we find its height is about one thirteenth of 34 feet or 30 inches, exactly what we should expect.  Since there is no air pressure within the tube, the atmospheric pressure on the mercury in the dish is balanced solely by the mercury within the tube, that is, by a column of mercury 30 inches high.  The shortness of the mercury column as compared with that of water makes the mercury more convenient for both experimental and practical purposes. (See Laboratory Manual.)

78.  The Barometer.  Since the pressure of the air changes from time to time, the height of the mercury will change from day to day, and hour to hour.  When the air pressure is heavy, the mercury will tend to be high; when the air pressure is low, the mercury will show a shorter column; and by reading the level of the mercury one can learn the pressure of the atmosphere.  If a glass tube and dish of mercury are attached to a board and the dish of mercury is inclosed in a case for protection from moisture and dirt, and further if a scale of inches or centimeters is made on the upper portion of the board, we have a mercurial barometer (Fig. 44).