xxiv 



INTRODUCTION. 



For Centigrade temperatures, 



The practical tables consist essentially of two mutually dependent 

 parts : — the first gives values of 2000 jh in a table of double entr}' of 

 which the altitude of tlie station and the mean temperature of the air 

 between the station and sea level are the arguments ; the second gives the 

 reduction to sea level in a table of double entr^- of which the arguments 

 are 2000 w and the observed barometric height corrected for temperature. 

 In addition, a subsidiary table gives the small correction for latitude to 

 be applied to the values of 2000 w. This correction, while of theoretical 

 interest, seldom becomes of practical importance, since its effect is in 

 general overshadowed by the relatively large uncertainties incident to the 

 determination of the true mean temperature. 



The mean temperature of the air column is to be obtained from the 

 obser^-ed temperature at the station by employing some assumption as to 

 the rate of change of temperature with altitude. In the discussion of 

 barometric obser\-ations made in the mountain and plateau regions of the 

 United States, it has beeu found that this rate of change is a climatic factor 

 which needs to be determined for ever>- station for different seasons of the 

 3'ear. and for different atmospheric conditions. When the results of such 

 investigations are embodied in tables for reduction to sea level, the tables and 

 the method of their use may be simplified and the labor of obtaining the 

 reduction gxeath* abridged ; but in the nature of the case, these special 

 methods can not be utilized in the construction of general tables which 

 are to be applicable to all phases of topography and climate. 



Whatever method be used for obtaining the mean temperature of the 

 air column {&) from the observed temperature at the station, the former 

 and hence the latter is subject to the important condition that it shall not 

 contain the diurnal fluctuation. Hence in reducing to sea level anv indi- 

 xadual obser\-ation of the barometer, the simultaneous obser\'ation of air 

 temperature used in obtaining 6 should be reduced to the daily mean by 

 a correction, or, better, the actual mean temperature of the preceding twenty- 

 four hours should be taken. 



