Table 99. II7 



REDUCTION OF BAROMETRIC HEIGHT TO STANDARD TEMPERATURE.* 



*The height of the barometer is affected by the relative thermal expansiotiof the mercury and 

 the glass, in the case of instruments graduated on the glass tube, and by the relative expansion of 

 the mercury and the metallic inclosing case, usually of brass, in the case of instruments graduated 

 on the brass case. This relative expansion is practically proportional to the first power of the tem- 

 perature. The above tables of values of the coefficient of relative expansion will be found to give 

 corrections almost identical with those given in the Ititernational Meteorological Tables. The 

 numbers tabulated under a are the values of a in the equation /// — Hi' — a (/' — /) where Ht is the 

 height at the standard temperature, Ht' the observed height at the temperature/', and a (/' — /) the 

 correction for temperature. The standard temperature is o° C. for the metric system and 2S°.5 F. 

 for the English system. The English barometer is correct for the temperature of melting ice at a 

 temperature of approximately 28°.5 P., because of the fact that the brass scale is graduated so as 

 to be standard at 62° F., while mercury has the standard density at 32° F. 



Example.— A barometer having a brass scale gave H = 765 mm. at 25° C. ; required, the cor- 

 responding reading at 0° C. Here the value of o is the mean of .1235 and .1251, or .1243 \ .• . a{t' — /) 

 = .1243X25 = 3.11. Hence //b = 765 — 3- ' i =761.89. 



N. B.— Although a is here given to three and sometimes to four significant figures, it is seldom 

 worth while to use more than the nearest two-figure number. In fact, all barometers have not the 

 same values for o, and when great accuracy is wanted the proper coefficients have to be deter« 

 mined by experiment. 



Smithsonian Tables. 



