PART III. EARTH CURRENTS AND EARTH MAGNETISM. CHAP. I. 74 l 



In Series II we have collected a number of powerful storms. Here the external perturbing forces 

 terfere largely, so that the effects of the earth-currents only appear, as a rule, as secondary waves on 

 main dellections. Here too, however, they are generally very distinct. 

 We will indicate a few. 



Dec. 26, time about 2o 3 /4 h 



Jan. 23, I? 2 /; 



l/ 8 h 



Feb. 7, I7 h 18'' 



and 2i h 22'' 

 9, about 18'' 

 10 & IT, 23'' i 1 '. 



The last two sets of curves in Series II are from Bossekop. We will examine them a little more 

 isely later on. 



On looking at the reproduced curves from Kaafjord, and especially the intervals mentioned above, 

 \- notice in the first place that the two earth-circuits exactly correspond in every detail; and as the 

 ble CXII shows, the relation between the deflections in the two components for one and the same day 

 very nearly constant, whereas it varies somewhat from day to day. 



In the next place, the resemblance between the earth-current curves and the declination curves is 



considerably greater than between the former and the horizontal intensity. 



These facts are, I think, accounted for by the small sentiveness of the H, /.-magnetometer com- 

 ed with that of the declinometer, and further by the fact that the direct effect of extra-terrestrial 

 rent-systems is much more pronounced in H than in D. 



Owing to the smallness of the oscillations and the difficulty of identification P h is only found in 

 i ntively few cases. 



If we could put (/! = q.,, we should expect to find that the relation P h jje EVi . would equal P d jje NS 

 II we compare the mean figures, we also find that such is the case; but while the relation 

 / _A' VS . remains nearly constant all the time, relation P h /Jc f . w varies very considerably. As long as 

 tH shunt is kept unaltered, however, the relation is fairly constant. 



Before January 12, in the NS line, galvanometer B was employed, after that date galvanometer 

 h With the change, a very distinct leap in the values of the relation Pijde Kg is observable. A simi- 

 la leap is observable at the change from shunt-resistance 0.2 Q to o.i J2. 



In the last case the relations are reduced to very nearly half the value, which probably indicates 

 th: contact-resistances have here played a decisive part, and they must be assumed to occur in the 

 slint-circuit itself. An explanation may also be found for the discontinuity here found in the conditions 

 c.i i changing the galvanometers on January 12, merely by assuming that an influence is exerted by 

 cc tact-resistances. In such case it must be assumed to occur at the points where the cable is connected 

 w i the galvanometer and shunt circuits. Its effect will be the same as if the resistance of the cable 

 increased by a corresponding amount. 



It is therefore doubtful whether any great importance can be attached to the agreement between 

 th mean figures. 



ON THE CONNECTION BETWEEN POLAR STORMS AND EARTH CURRENTS. 



147. As above mentioned the second series of plates, PI. XXI XXIII, contains a number of 

 sii iltaneous records of earth-currents and the magnetic elements during a number of comparatively 

 -nt storms. The conditions during polar storms are also given in greater detail in some of the rapid 

 : MI\!S contained in the third series of plates, e.g. PI. XXXIV and XXXV. 



Birkeland. The Norwegian Aurora Polaris Expedition, 19021903. 94 



