132 AKSEL S. STEEN. TERRESTRIAL MAGNETISM. [NORW. POL. KXP. 



about the same value on both the days mentioned. The same assumption 

 is also possible in the case of the last two observation-days, February 10th 

 and March 30th, 1894. There are thus two groups of observation-data, from 

 which may be calculated, supposing that R 2 has remained constant during 

 the period of a few weeks which each group embraces. We then obtain 



.. _ sin i/> 2 sin t// 2 ' 

 t sin T/> 2 ' f sin t// 2 ' 



when i// 2 and i// 2 ', t and f indicate the angle of deflection and the tempe- 

 rature observed respectively on the first and the second of the two days of 

 the groups in question. The calculation gave the two following values 

 for f : 







1893. Oct. 16 and Dec. 2 .... - 0-0001979 



1894. Feb. 10 and March 30 ... - 0'002435 



Mean - 0-00132 



Although this result cannot naturally lay claim to any great degree of 

 accuracy, I have thought it possible, in the absence of anything better, to 

 make use of it, and have therefore, with = 0'00132, calculated E 2 by 

 the observations in Hamburg in 1893, and in Wilhelmshaven in 1897, and 

 have found, 



E, 



for June 9, 1893 . . . 0-41198 

 April 20, 1897. . . 0-41054. 



Starting with the supposition that R z has decreased proportionally with 

 time, I have been able, by graphic interpolation, to deduce the following 

 values for the above-mentioned 4 days, viz. 



B, 

 in 1893, Oct. 16 .... 0-4118 



Dec. 2 .... 0-4118 

 in 1894, Feb. 10 .... 0-4117 

 March 30 ... 0'4117. 



By substituting these values in formula (3), we obtain the total intensity 

 W, which, together with the corresponding values of the horizontal intensity 

 specified on page 131, give the actual inclination /, according to formula (2). 



