68 AKSEL S. STEEN. TERRESTRIAL MAGNETISM. [NORW. POL. EXP. 



observations, and after their termination. By the aid of this formula, I have 

 drawn up a table of log y with ^ - expressed in scale divisions as argu- 

 ment. The table contains the value of log y to 5 places of decimals, for 

 every tenth of "T - . 



The correction for rate of chronometer is calculated according to the 

 usual formula: 



-1 -l- 8 

 i" 86400' 



where s is the number of seconds gained or lost by the chronometer in 

 24 hours. Log p has been placed in a table with s as argument. 



The chronometer used as observation-chronometer on the 1st August, 

 1893, was the Kutter; afterwards, until the end of 1894, the Haagensen 

 watch was used, being compared before and after the series of observations 

 with the Hohwu chronometer. It lost daily between 9 and 15 seconds. In 

 1895 and 1896, the Frodsham chronometer was constantly used, being regu- 

 lated according to sidereal time. Its daily acceleration in relation to mean 

 solar time, which varied between 231 '7 and 228'8 seconds, and its error on 

 local time for the observation-days in question, have been given me by 

 Professor GEELMUYDEN. 



As the time for the value of the horizontal intensity corresponding to 

 each separate calculated time of vibration, I have taken the mean of the 

 hour noted at the needle's first and last passage over the middle division of 

 the scale, reduced to local time. 



I have been unable to introduce any correction for the torsion force of 

 the suspending thread, as no observations for its determination were made. 



After three series of vibration-observations with magnet VI on August 

 1st, 1893, at Khabarova, none were made until August 18th, 1894, when 

 magnet V was used. After this the vibration time was determined regularly, 

 most frequently for magnet V, now and then for magnet VI. There are 

 altogether 82 series of vibrations for magnet V, and 19 for magnet VI. 

 A few of the series, however, were made during such great disturbance, that 

 the value of the time of vibration found must be considered very uncertain. 



