194 BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 1902 1903. 



storms, we must, if possible, consider as an effect of the polar storm the deviations from the conditions 

 before and after the period in which the polar storm occurs. In this way the conditions are certainly 

 elucidated, as will best be seen when we come to consider the field of force. If we look at the total 

 force as belonging to the polar storms, we here find a change in the direction of the force that has no 

 parallel farther north, where, as we shall see, it remains almost constant in direction throughout the 

 perturbation. 



At Christchurch too, there is a very considerable and well-defined perturbation, which is particularly 

 well developed in H, and exhibits a course that in the main resembles that at Dehra Dun, but has 

 perhaps a still greater resemblance to those in North America. 



In the western hemisphere we also find simultaneous considerable perturbations, which are especi- 

 ally powerful at Sitka, but also of no little strength in the United States; while even at Honolulu there 

 is a very considerable effect on that day. 



We will first consider the four northernmost stations. 



In the //-curve, in particular, the course of the perturbation exactly corresponds with that at our 

 Norwegian stations. It commences with some strength at about I9 h o m , increases rather rapidly to a 

 maximum, and remains fairly powerful for about an hour, after which it diminishes, but then once more 

 increases somewhat, and forms a new, secondary maximum at 21 h 3o m . We have then first a powerful 

 maximum and then a weaker one a condition we observed at all the Norwegian stations. In declination, 

 on the other hand, the conditions here are somewhat peculiar. A perturbation appears at the three 

 stations in the east of North America, at 17'' 56, answering to a deflection westwards, and remains, 

 excepting for a short interval when the polar storm is at its height, almost constant for several hours, 

 only ceasing at about 23'' o m . Whatever this deflection may be due to, we must assume that it cannot 

 be the effect of the system we are now considering, as this does not begin to act until more than an 

 hour later. 



At Honolulu a distinct variation is noticed especially in the //-curve, coinciding with the polar 

 storm; but on drawing the mean line, it appears that there are perturbations both before and after. 

 Before, H is greater than the normal, while after, it has a value that, is considerably below the normal. 



The field during the powerful storm is shown on ten charts. The first represents the conditions at 

 19'', the last at 22 h 30. In southern latitudes a decomposition of forces has been effected on the charts 

 from I9 h 15 to 21'' 30, but at the Norwegian stations and Sitka this has not been done. At the latter 

 places the powerful storm is so dominant that the total forces are principally conditioned by the powerful 

 polar storm. The field at these northermost stations remains, as we see, fairly constant in its -form 

 throughout. At the Norwegian stations the current-arrows on the whole are directed westwards along 

 the auroral zone. 



At Dyrafjord the current-arrows at first have the very usual direction, WSW (see the chart for 

 jgh i^m^ byj afterwards turn northwards, and remain almost the whole time pointing towards the west, 

 or even farther towards the north. The vertical component of the perturbing force is directed upwards 

 all the time. 



At Axeleen and Kaafjord we have the field that is typical of these storms. The current-arrows 

 are almost parallel except at about 19'* I5 m , and WSW in direction. The horizontal component of 

 the perturbing force is greatest at Axeleen; but on the other hand, the vertical component at Kaafjord 

 is greater throughout, and is directed upwards at this station, and downwards at the former. At 

 about i9 h 15 a peculiarity makes its appearance at Kaafjord, namely, that the horizontal component 

 becomes about 0, while at the same time the vertical is very powerful. To explain this, it is natural to 

 conclude that there is a local perturbation at Kaafjord of contrary effect. Sharp local deflections such 



