NO. 17.] TEMPERATURE OF THE AIR. 401 



THE TEMPERATURE OF THE AIR. 



THE DIURNAL PERIOD. 



For the determination of the diurnal period of the temperature of the air, I have 

 taken the regular readings of the thermometer made each second or fourth hour given 

 in the "observations", and the registrations of the thermograph. The latter were con- 

 trolled by, and reduced to, the former. In some cases the registered temperatures could 

 serve to correct apparently erroneous readings of the thermometer. 



The hours of the thermograph were controlled and verified by simultaneous compar- 

 isons between the position of the pen of the thermograph and the reading of the 

 chronometer Hohwii 639 (Vol. II, No. 6, XIX), from which the exact local hour was 

 computed by Prof. Geelmuyden's astronomical tables in Vol. II, No. 6, H., p. 86. The 

 said comparisons were noted at the beginning and the end of the thermograph-sheets for 

 each week, and also in the meteorological journal almost every day. By these means 

 it has been possible to find the exact time for each registered temperature. 



On a tracing-paper, laid over the thermograph-sheet, the corrected temperatures 

 observed by the mercury or spirit thermometers in the screen were marked with fine 

 dots in their true place with reference to time and value. The tracing-paper was then 

 moved upon the thermograph-sheet so that the thermograph-curve could interpolate the 

 temperatures corresponding to the intermediate hours between the fixed 2- or 4-hourly 

 thermometric observations. 



The thermograph was kept working from the middle of October, 1893, to June 

 27 th , 1896. From that day up to the end of July, 1896, the temperature at the odd 

 hours was interpolated from the readings of the thermometers at the even hours. (For 

 July, 1896, every 4 th hour.) 



In some cases, when the screwing of the ice made the registering unreliable, or 

 the thermograph could not be kept going, the intermediate hourly temperatures have been 

 interpolated graphically between the regularly observed temperatures. Such values have 

 been printed in special type. 



The hourly temperatures thus found are given in the following tables for the time 

 during which the Fram was drifting in the ice, from the middle of October, 1893, to 

 the end of July, 1896. 



For each day are also given the lowest and the highest temperature. The day is 

 reckoned from midnight to midnight. The minimum and maximum temperatures given 



in the tables are mostly those observed with a minimum or maximum thermometer. 



51 



