TART 



ON MAGNETIC .STORMS. CHAPT. I. 



5 1 



The table shows that the scale-values for H and V are not constant; in the case of e e in particular 

 there is a considerable increase, and in the determination of the 3rd March, 1903, the balance was almost 

 immovable. This abnormal circumstance seems, however, to have been only of a temporary nature, as 

 will be seen from the curves before and after. We have not employed any smoothing formula here 

 for /,, but have found the scale-values by interpolating between two successive observations. 



We have employed the following formula for : 



= 4.76 -f 0.0285 *> 

 t indicating the number of days reckoned from the ist October. 



TABLE 111. 



Scale-values for Bossekop. 

 r<i= 1740 mm. 



HO indicates the magnetic force, to which the declinometer-needle is actually subjected. During 

 the first determination of sensibility it is only terrestrial magnetism that is acting. 



The force acting on the declinometer-needle, during the 2nd observation, may be determined in 

 two ways. We can either use the deflection by the torsion, having the same thread and the same 

 position in both cases; or we can employ the deflection with the deflecting magnet, the magnet having 

 been placed at the same distance on the deflection-rod in both cases. The two methods give about 

 the same result. The value given is the mean value. For the period from the 25th to the 2gth March, 

 we have no scale-value, a determination of sensibility that was made on the 2yth having been unsuccessful. 



DETERMINATIONS OF SENSIBILITY FOR DYRAFJORD. 



19. The registering at Dyrafjord was begun on the 25th November, 1902, and was continued 

 almost without interruption until the I5th April, 1903. 



During that time, neither the declinometer nor the variometer for horizontal intensity underwent any 

 change, except that the torsion-head of the variometer for the horizontal intensity was a little twisted on 

 the ist December, 1902. 



As we shall presently notice more fully, the variometer for the vertical intensity, in the course of 

 the above-mentioned period, underwent a few small changes, which, however, have had no perceptible 

 influence upon the scale-value. 



As the torsion in the thread of the declinometer is slight, /. will be small. The torsion has there- 

 fore only been determined 3 times, namely on the 28th November, and 8th December, 1902, and the 

 1 6th January, 1903. As the mean of these three, it is found that 



y. 0.00164. 



