70 AKSEL S. STEEN. [Snd arc. exp. fram 



THE INDEX-ERROR. 



As the inclination needle, B, was also intended for determinations 

 of intensity, it was never remagnetised. In consequence of this, the 

 inclination values found had to be corrected for the index-error of the 

 needle. If we call the value of the inclination observed with the Fox 

 circle /', the true inclination /, and the index-error J, we obtain 



1=1' + J 



For the determination of J, I have on a former | occasion ^ educed 

 the followins; formula: 



'& 



p cos (T -j- a) cos T 



J? + p cos (7' -fa) sin 7' sin 1' ^^' 



in which H is the horizontal intensity, /' the observed inclination, and 

 p and a two constants. 



From the observations made in Hamburg in 1893, and Wilhelms- 

 haven in 1897, and during the first Fram Expedition in the winter of 

 1893-94, I found, for needle B, 



p = _61.5' and « = 209°33' 



Since then there has been no opportunity of verifying these values, 

 as no determinations of intensity were made with the apparatus during 

 the second Fram Expedition; nor was the needle B used by Captain 

 Amundsen in his inclination observations made at Potsdam with the Fox 

 circle in December, 1902, when a new needle made by Dover was used. 



I have therefore, as an experiment, employed the above value of p 

 and a, and calculated a mean value of J for each of the four stations, 

 inserting in formula (5) the mean values for H and T deduced from 

 the observations for each separate station. The result of this calcula- 

 tion was as follows: 



The values thus found for the absolute inclination I, can scarcely 

 be correct, however; for if we calculate the total intensity, W, for each 

 station according to the formula 



Nansen Expedition. T. M. p. 134. 



