18 H. MOHN. METEOROLOGY. [norw. pol. exp. 



was never noted until this was reached. The hygrometer in the screen 

 served as a check on the other, when the latter was set out after being cleaned 

 or verified. 



The comparisons made between the psychrometer and the hygrometer 

 give a mean error of comparison of from 1 to 2 %. For the winter, the error 

 of a single adopted value of the relative humidity is estimated to be from 3 % 

 to 50/0. 



The vapour-tension (or absolute humidity) has been computed 



(1) from the psychrometer-observations and Jelinek's Tables, when the 

 temperature of the wet-bulb thermometer was above — 10°; from 0° to 

 — 10° by Ekholm's rule. 



(2) from the corrected readings of the hair-hygrometer. 



If e,„ be the maximum tension of vapour at the temperature of the air, 

 e the vapour-tension sought, and 

 r the relative humidity, 



r 



we have e = e» X '^^ ' 



The values for e^ were taken from Jelinek's Tables I a, Spannungstaiel. 

 For temperatures from — 40° to — 55° the following auxiliary table was 

 computed by the formula^ 



fir 



B = A.IO = e,n 



where A = 457 mm, ^ = 0-03134, «„ == 0-003667, and t = temperature 

 of the air. 



Vapour-tensions below O'OS mm. are given in the tables as 0-0 mm. 



The values given for the relative humidity may be uncertain to the 

 extent of ± 5 %. The influence of such an error upon the deduced value of 

 the vapour-tension is trifling, and practically of no importance. The follow- 



' 0. J. Broch. Tension de la vapeur d'eau p. A. 30. 



