442 BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 19021903. 



Whereas during this type of perturbation, negative values of the perturbing forces in the horizontal 

 intensity were found in the polar regions, in lower latitudes at the same time positive values of this 

 component were found. Mere, then, we have a reversal of the component. The deflections in declination, 

 on the other hand, may at one time be easterly, at another westerly, according to circumstances. 



We find the typical perturbation-area of a negative polar storm in its most perfect and distinct 

 form, shown in figs. 41 and 42. Areas such as these are constantly met during the negative polar 

 storms. In the auroral zone there are very strong current-arrows directed westwards, while south of it 

 the current arrows point in the opposite direction. In fig. 40, p. 86, we have endeavoured to give a 

 diagrammatic representation of an ideal form for the typical perturbation-area that appears during the 

 negative polar perturbations. The large vertical arrow, A, is supposed to coincide with the direction 

 of the current arrows found in the most perturbed area, the so-called 'storm-centre'. The entire lines arc 

 the lines of force for the horizontal magnetic forces that occur upon earth ; while at right angles to them 

 are the dotted potential lines, or, as we have called them, current-lines. It is the right half of this figure 

 that should correspond with the field represented, for instance, in figs. 41 & 42. For the sake of the 

 general idea, we called the line that coincides with the arrow A in fig. 40, or along the current-arrow 

 in the storm-centre, the principal axis of the system, and the line at right angles to it the transverse axis. 

 It will be seen that we have supposed the area of perturbation upon the two sides of the principal axis to 

 be exactly symmetrical, but in reality this will never altogether be the case, but only approximately. 

 As a rule there are occurrences all over the polar regions of strong or slight polar precipitation, which 

 easely effaces the traces that we might expect to find of such a condition. When the storms are 

 particularly simple and well-defined, however, indications may to a certain extent be found of a pertur- 

 bation-area on the other side of the principal axis, that is more or less symmetrical with the first. We 

 believe we have found a condition such as this in the perturbation of the I4th & I5th February, 1883, 

 where there is a simple, well-defined negative polar storm with storm-centre in the north of Europe, of 

 which the principal axis lies more or less along the auroral zone in this district, there being no storms 

 of any marked strength at the same time at other places round the polar zone. We here have some 

 stations more or less symmetrically situated on both sides of the principal axis; and in the description 

 on pp. 363 & 364, some conditions are pointed out that, although possibly only slight, would seem to 

 confirm this assumption. In figs. 152 & 153, the current-arrows at Kingua Fjord and Godthaab indicate 

 that such an area actually exists in the regions to the north of the principal axis. 



On the transverse axis there are two characteristic points that are enclosed by the current-lines. 

 The horizontal components of the perturbing forces in the regions round one of these points, are directed 

 straight in towards the point, while in the other all the horizontal forces point straight out from it. In 

 the points themselves, the horizontal force is zero. 



The first of these points we have called the system's point of convergence, the second its point 

 of divergence. 



The storm-centre during a magnetic storm does not remain in the same place all the time. As a 

 rule, a more or less distinct movement of the various storm-centres can be traced. In the polar regions 

 this can best be seen from the horizontal-intensity curves, where a more or less distinct difference in 

 the time of the beginning, maximum, and conclusion of the deflections at the various stations is found. 

 We may also refer here to the perturbation of the 151)1 December, 1902, where this condition conies 

 out with unusual clearness when we look at the horizontal-intensity curves for Dyrafjord and Axeleen, 

 on PI. X. There seems no doubt that the storm-centre here was at first situated in the vicinity of 

 DyraQord, and afterwards moved eastwards along the zone, so that at the end of the perturbations, it 

 was situated nearest to Axeleen. This is also apparent on looking at corresponding charts. At first 

 the current-arrow at Dyrafjord is the strongest; but it then decreases, while the current-arrow at Axeleen 



