444 HIKKKLAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 1902 1903. 



are at a certain place, Val Joyeux, considerably smaller than the calculated. The cause of this should, 

 we believe, be sought for principally in two circumstances. The first of these is that in our current- 

 system the horizontal portion of current may be more conspicuous than in reality it is. If a rather 

 different form of this had been chosen, e. g. if it had been assumed that the rays were more as if they 

 ran in towards the earth in a point, as shown diagrammaticaly in the figure 187, the agreement would 

 probably have been closer also as regards the vertical intensity. A system of this form would also probably 

 be more in accordance with the actual current-system. As, however, it is a question of a very rough 

 estimate, and the calculation of the first is considerably easier, we have employed this form. 



In the second place, a no inconsiderable part of the perturbing forces observed will certainly be 

 due to earth-currents. These, as every one is aware, will, when they have the effect of increasing 

 the horizontal forces due to an external current-system, have the effect of decreasing the vertical 

 component. The magnitude of the vertical intensity has been employed, ever since Gauss's time, tor 

 the purpose of determining how great a part of the magnetic effects observed must be ascribed to exter- 

 nal forces, and how great a part to internal. In the chapter on earth-currents we shall look more closely 

 into this proportion as regards the magnetic storms. We must, however, expressly draw attention to the 

 great uncertainty that attaches to the determination of the perturbing force in the vertical intensity. We 

 may, for instance, refer to Chart III for the storm of the I5th February, 1903. Here, while at lYdr 

 there is a comparatively powerful vertical arrow, at the surrounding places, Val Joyeux, Wilhelmshaven 

 and Munich they are too small to be measured. The values of P e , therefore, when small, must IK 

 considered as only approximately correct. 



There is another circumstance that we may point out. In the negative polar storms about mid- 

 night, Greenwich time, the horizontal portion of current will as a rule fall between Axeleen and Kaa- 

 fjord, that is to say north of the latter station. In the most powerful storm we have studied, however, 

 namely that of the 3ist October and ist November, 1902, the current seems to have moved to the south 

 of this station, as there are now positive deflections in vertical intensity. In lower latitudes, at Wil- 

 helmshaven and Pola, we also find at the same time positive perturbing forces in the vertical intensity, 

 which however, we think should be considered as the effects of the negative equatorial storm, of which 

 there are also distinct effects. 



The third of the principal forms of magnetic storms, is the positive polar storm, of which the follow- 

 ing are the chief peculiarities : 



The form of the perturbation-area is on the whole the same as that of the negative polar storms; 

 but all the forces in that area, both horizontal and vertical, are in the opposite direction. In the polar 

 regions, in this type of perturbation, there are positive deflections in the horizontal intensity. We have 

 here employed the same terms as in the negative polar storms storm-centre, principal axis, transverse 

 axis, and points of convergence and divergence. Whereas in the negative polar storms we found the 

 system's area of convergence to the south of the storm-centre when considering the conditions in the 

 northern hemisphere, in the positive storms we find the system's area of divergence in that region. As 

 a rule, the perturbing forces in the positive polar storms diminish just as rapidly in strength as those in 

 the negative polar storms; and we find here too a reversal in direction of the horizontal component <>t 

 the perturbing force at about the same distance from the storm-centre as in the negative polar storms. 

 Not infrequently, however, we meet with cases in which there are positive deflections comparatively far south. 



In the matter of strength, the positive polar storms are as a rule somewhat weaker than the nega- 

 tive, and the character of the curves in the polar regions is not quite so disturbed in the former type 

 of perturbation as in the latter. The field of a positive polar storm appears most distinctly in fig. 34, 

 Chart IV for the gth December. In fig. 83 charts for the 151)1 February, 1903 the form of the field 



