-_|2 I'.IKKKI.A.M). [UK Me >R WFe ,1 A.\ ATKoRA I'UI.ARiS KXI'l-'lHTle >N, I gO2 1903. 



I he first important conclusion which we can draw from the curves is tin. 1 following: 



Tin' ini'lli i'iirn : i//s us inanifrsletl l>v tin' galvanometer ilrflcctinns if/trim; point- star/us <v//;t7i IMI;- 

 t/n'/r Cfii/ri's in ////' I'irinilv <>/ llif station, cannot explain the main part <>/ the perturbing force. 



The justilioatioii of this conclusion will he immediately apparent on looking at the curves; fi,' r 

 while the magnetometers can maintain a largo deflection in a certain direction lor hours, the galvano- 

 meters will change direction of deflection relative to the normal line usually a great many times duriii" 

 the same period. Indeed the galvanometer curves have often the appearance of oscillations round tin- 

 normal line (see I. i. the curves tor Nov. 2 and Fe.br. 12). 



Now we saw in the previous article that the earth-currents produce- magnetic variations according 

 to rules given in tahle CX11I. 



In consequence we. always find that, superposed on the main wave of the magnetometer curve, which 

 is probably due to extra-terrestrial currents, there arc a number of waves and oscillations which, as regards 

 occurrence and form, coincide with the galvanometer oscillations, a phenomenon, that is well illustrated in 

 Series 1 and II, Plates XXX XXXIII, and even better in Series 111, Plates XXXIX' and XXXV, giving the 

 copies ot a number of rapicl-registerings. hrom the coincidence in form and phase I think we may safely 

 conclude that these; synchronous and similar rapid magnetic changes are direct effects of earth-currents flowing 

 underneath the magnets. 'I his conclusion is also confirmed by the fact that the curves of vertical intensity 

 run more smoothly than those' of the horizontal elements; for if the rapid changes are mainly due to 

 earth-currents spread over a considerable area, such currents would produce verv little effect in the 

 vertical din ction. 



The rapid synchronous oscillations in the two sets of curves will always occur with greatest strength 

 simultaneous with the strongest magnetic disturbances, and from this it is evident, that these briefer 

 variations must be due to the same primary cause as the magnetic storms themselves, i. e. according to 

 our assumption to an extra-terrestrial corpuscular current-svstem. The most natural way of explaining the 

 connection between the outer current system and the earth-currents is that the latter are induced by 

 variations in the former. From this, however, we cannot draw the conclusion that we always must find 

 such a simple connection between the two sets of curves as that expressed bv the rule of Blavier. On 

 regarding the curves one would also see, that such a connection in far the most cases does not exist. 



From the relations given in Table (.'XIII we should be able; to subtract from the magnetometer-records 

 the effect of the earth-currents. But even this corrected curve would hardly be competent to explain from 

 the rule of Blavie-r the manv oscillations of the earth-current curve. Looking at the curves for Kaafjord 

 we shall often find that the corrected curve for this place will apparently run rather smoothly compared 

 with that of the earth-currents. 



This circumstance is easily explained when we consider how the perturbation-conditions develope 

 in the polar regions. 



\Ye have become acquainted with the typical arrangement of the polar svstems. On the afternoon 

 side in latitudes that are not too high, we me-e't with the positive' polar storms, on the night side with 

 the negative. Wo have- seen that the positive system answers to the effect of ravs that descend towards 

 the earth, arc ele-llected westwards, and again leave the earth; the negative: to rays that are deflected 

 eastwards. Both these current-systems presumably lie at a comparatively great height above the auroral 

 /one, and their smaller and more rapid changes will therefore be' less evident, and the curves in consequence 

 are- characterised bv comparative smoothness. Among these svstems, however, rays are met with, which 

 descenel dire-ctly earthwards te> within e'omparativelv small heights above the surface of the earth. Here 

 the.- magnetic e'tirves are- exceedingly serrated. 1 hero are verv rapiel and comparatively strong variations, 

 some of wlue'h are due- to displacement of the districts of positive' and negative precipitation, and some 



