50 



101. This reading may next be corrected for temperature by apply- 

 ing the proper correction taken from Table I, corresponding to the 

 temperature shown by the attached thermometer, and further cor- 

 rected, if necessary, by the addition or subtraction, as the case may 

 be, of all other corrections known for the instrument, such as cor- 

 rection for capillarity, instrumental error, imperfect vacuum, grav- 

 ity, etc. 



102. Redwction to stmvdmd gravity. — The following will elucidate 

 the nature of the gravity correction as applied to barometric observa- 

 tions — an important matter that is often but indifferently considered 

 in the ordinary textbooks of meteorology : 



By the well-known principle of hydrostatics on which the action 

 of the mercurial barometer is based the pressure of the atmosphere 

 is equal to the pressure of the column of mercury that it will support. 

 But this latter pressure is only another name for the weight of the 

 mercury, and for columns of equal section the weight varies both 

 with the height of the column and with the force of gravity. 



The force of gravity varies with latitude and altitude; therefore 

 the height of the barometer, even when corrected for temperature 

 and instrumental error, does not give us a true measure of the atmos- 

 pheric pressure unless we first eliminate the small variations that 

 are due to gravity; that is, observations taken over a widely ex- 

 tended region to be strictly comparable must be reduced to a standard 

 force of gravity. 



The standard gravity adopted by physicists is that at the level of 

 the sea in latitude 45°. 



Tables of corrections for gravity are given in Section VIII.^ 



103. Reduction to sea level. — It was mentioned in paragraph 5 that 

 the atmospheric pressure was in the main nothing more than the 

 weight of a vertical column of air extending to the limits of the at- 

 mosphere. It naturally follows that as we go above the general sur- 

 face of the earth, whether in balloons or by ascending mountains, 

 the atmospheric pressure becomes less and less as we leave more and 

 more of the air beneath us. When it is desired to chart and com- 

 pare simultaneous observations of atmospheric pressure over extended 

 areas, and at various elevations above sea level, no inference can 

 easily be drawn from the actual pressures themselves, but each must 

 be reduced to some standard level. The sea level plane is most gen- 

 erally adopted, but the selection of a plane 2,000 to 5,O0D feet above 

 the sea offers more rational conditions in certain respects and is 

 sometimes advocated. For comparative purposes, therefore, baro- 

 metric observations from different stations requires a " reduction for 

 elevation." 



We can form a clear idea of what is wanted by confining our atten- 

 tion to the case of a barometer in a balloon at an elevated point 

 above the sea. The reduction for elevation is simply a measure, ex- 

 pressed in inches of the mercurial column, of the weight of the column 

 of air between the balloon — that is, the barometer cistern — and sea 



6 According to the formula for the force of gravity adopted by the International Bureaa 

 of Weights and Measures we have for the variations in gravity due to the latitude. 

 Gi=045 (1 — 0.00259 cos 2 -?). The variation in the force of gravity for different altitudes 

 i& small and is given by the ofrmula Gh=£ro( 1-0.0000000597 fe), li being the elevation 

 in feet. Neglecting this latter factor, the application of the gravity correction is equiva- 

 lent to multiplying the height of the bai'ometer by the factor 1 — 0.00259 cos S) . 



