414 BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, IQO2 1903. 



At the southern stations the conditions are evidently regulated by the negative polar storm. This 

 storm appears particularly clear, if we compare the horizontal-intensity curves. 



Except at Little Karmakul, no very distinct traces of positive polar storms are to be found during 

 this period, Uglaamie being the only station at which there appear to be any more or less evident 

 effects of such a storm. It will be seen that this station, at the time, is on the afternoon side of the globe. 



The fields of perturbation at 3'' ao m and at 4'' 2o m are represented on Chart II. 



At 3 h 20 ra we find the negative storm in the polar regions developed, to any extent, only in Jan 

 Mayen. The three southern stations indicate simultaneously by their current-arrows that they are in the 

 eastern part of the area of convergence. 



At 4 U ao m , however, the storm-centre has move<l or perhaps rather expanded eastwards, thus 

 bringing Cape Thordsen into the district of negative precipitation. At the same time the current-arrows 

 for the three southern stations turn clockwise through a considerable angle, just as our previous assump- 

 tions would lead us to expect. 



At Pawlowsk, at this time, we find negative values of P t , indicating the existence there of an area 

 of convergence. At Gottingen too, the direction of the perturbing force in the vertical intensity seems 

 to be the same; there is a distinct wave in the curve just at the time of the negative deflections at the 

 two arctic stations. 



At 5 h the positive storms in America are over, and negative storms begin everywhere, developing 

 subsequently to a considerable strength. 



At Godthaab the negative storm began to develope earlier. The negative perturbing forces here 

 must be regarded as continuations of the powerful eastern system. 



At Fort Rae too, the negative deflections become stronger, and at 6' 1 a fairly powerful negative 

 storm begins to develope, and continues until about 17''. There are two maxima here, separated by a 

 period during which the negative forces are considerably weaker, although the direction of the deflections 

 remains unchanged. 



At Uglaamie the stronger negative forces appear somewhat later than at Fort Rae; but a little before 

 8 h they begin to increase rapidly until they attain considerable strength. The negative deflections then 

 continue more or less constant in strength until about I7 h , after which they are small. 



At Kingua Fjord too, negative storms appear to be at work ; but we will reserve our description 

 of the conditions there until we come to the charts. 



We note that this transition from positive to negative storms in America takes place at the time 

 when these districts enter the night-side of the earth. At the same time the districts in Asia and Europe 

 move on to the day-side of the globe, and at the polar stations here, Cape Thordsen excepted, we also 

 find a transition, but from negative to positive systems, and thus the reverse of that in America. 



The change takes place earliest in the most easterly districts. At Little Karmakul, for instance, 

 there seem to be positive storms as early as 3 h . At a little before 6 h , however, they begin to be more 

 distinct, the positive deflections becoming larger and larger, until about I4 h there is a maximum for the 

 positive deflections. 



At Bossekop and Sodankyla the positive storm developes very characteristically; but the positive 

 deflections begin a little later. At about 5 h 2o m the negative storm at Bossekop is over, and from that 

 time until about io l1 , there are small deflections now to one side and now to the other. At io h the 

 positive deflections begin to increase with comparative rapidity, and reach their maximum at about 15'' 20 m , 

 when they decrease rapidly. The development of the storm at Sodankyla is very similar. 



If we go on to Jan Mayen, we still find, at the beginning of the period, effects of negative deflec- 

 tions. After io h , the positive storm there developes powerfully. Thus while the effects of the positive 

 storm appear more or less simultaneously at Bossekop and in Jan Mayen, the previous negative storm 



