1904- No. 2. THE DIURNAL VARIATION OF TERRESTRIAL MAGNETISM. 13 



The observations of vertical intensity taken at the polar stations 

 during the polar year, are not, as we know, to be relied upon, and I 

 have therefore not ventured to make use of them in this investigation, 

 which will thus, as regards all the stations, comprise only the variations 

 in the horizontal component of terrestrial magnetism. In a separate final 

 section, however, I shall deal with the variations of the vertical intensity 

 at the few stations from which the observations may be regarded as 

 most reliable. 



As the period examined embraces 2 days, there are for each hour 

 2 observations of declination and horizontal intensity, which are combined 

 in a mean, from which again is calculated the corresponding value for 

 A* and Y'. As regards the stations whose obser\-ations are given after 

 local time, I have, in order to make the material as uniform as possible, 

 by interpolation determined A' and Y for even,' whole hour, Göttingen 

 time. From the 24 values for X and i' I have calculated the daily 

 mean A^ and Ye, and thereafter ^A*=A" — A'^. and J i"= Y — A*^ for 

 every hour expressed in whole and tenths of /, these values being 



_ J, + 2J^ -f-^3 



finallv smoothed according to the formula J^ = 



4 



It should be remarked that ^A' is reckoned positive northwards, and J Y 

 positive eastwards. 



With this material I have calculated the following quantities for 

 each hour: 



1. A = '^jX^-\-jY^ is the horizontal component of the deflecting 

 force, the radius vector in the horizontal vector-diagram. 



2. The angle a, which A forms with the meridian-line, reckoned 



positive from the north point eastwards ; thus tan a = — ^ . 



3. The horizontal projection of the electric current above the earth's 

 surface, which, according to Ampere's law, may be assumed 

 to have produced the direction and the force of the vector A. 

 The direction of the current, 5, is indicated according to the 

 point of the compass from which it comes. 



The material worked up is brought together in Table 2, and arranged 

 according to Göttingen mean time, G, the corresponding hour in local 

 mean time, L, being also added. The order of the stations is the same 

 as in Table i. 



