XXX INTRODUCTION. 



Stations respectively). The first entrj' of Table 25 with the argument B 

 gives an approximate value of the elevation of the upper station above 

 sea level, and the second entr}^ with the argument B^ gives an approximate 

 value of the elevation of the lower station. 



Table 26 gives the temperature correction : 0.00367 6 C. x Z. 



The side argument is the approximate difference of elevation Z and the 

 top argument is the mean temperature of the air column. The values of Z 

 vary b}^ intervals of 100 m. from 100 to 4000 metres and the temperature 

 varies bj'- intervals of 1° from 1° C to 10° C. with additional columns for 20°, 

 30°, and 40° C. Attention is called to the fact that the formula is linear 

 with respect to 6, and hence that the correction, for example, for 27° equals 

 the correction for 20° plus the correction for 7." When the table is used for 

 temperatures below 0° C, the tabular correction must be subtracted from, 

 instead of added to, the approximate value of Z. 



Table 27 (pp. 1 12 and 113) gives the correction for humidity resulting 



from the factor 0.378 j x Z= ^ Z. 



Page 112 gives the value of 0.378- multiplied by loooo. The side 



argument is the mean pressure of aqueous vapor, <?, which serves to repre- 

 sent the mean state of humidity of the air between the two stations. 

 ^ = i (/+/o) (/and/^ being the vapor pressures observed at the two stations) 

 has been written at the head of the table, but the value to be assigned to c 

 is in reality left to the observer, independently of all hypothesis. The top 

 argument is the mean barometric pressure 7 (-^ X i?J. 



The vapor pressure varies by millimetres from i to 40, and the mean 

 barometric pressure varies by intervals of 20 mm. from 500 mm. to 760 mm. 



The tabular values represent the humidity factor /3 or 0.378-T, multiplied by 

 I 0000. 



Page 113 gives the correction for humidity, with Z and loooo x 0.378 7- 



b 

 (derived from page 112) as arguments. 



The approximate difference of altitude is given by intervals of 100 

 metres from 100 to 4000 metres, and the values of ioooo/3 vary by intervals 

 of 25 from 25 to 300. The tabular values are given in tenths of metres to 

 facilitate and increase the accuracy of interpolation. 



Table 28 gives the correction for latitude, and for the variation of 

 gravity with altitude in its effect on the weight of the mercurial column. 

 When altitudes are determined with aneroid barometers, the latter factor 

 does not enter the formula. In this case the effect of the latitude factor 

 can be obtained by subtracting the tabular value for latitude 45° from the 

 tabular value for the latitude in question. 



The side argument is the approximate difference of elevation Z, varying 

 by intervals of 100 metres from 100 to 4000. The top argument is the 

 latitude varying by intervals of 5° from 0° to 75° 



