540 BIRKELAND. THE NORWECilAN AURORA POLARIS EXPEDITION, 1902 1903. 



The positive and negative average storms have respectively their maxima at about the same time as the 

 P and TV type of storminess. The two types of storminess are merely another aspect of the existence of 

 the two types of polar storms. It is, however, by no means a matter of course that the maxima of 

 storminess should fall on the same hours as the maximum of perturbing force of the average storm; for 

 from the equation S^ = S'' H S' I H it follows that S d might be small even when both S p and S" are large. 

 The coincidence regarding the occurrence of maxima of TV and P storminess on the one side, and the 

 perturbing force of the positive and negative average storm on the other, is a consequence of the fact 

 that the occurrence of the two types of storms does not greatly overlap, but that each type is mainly 

 restricted to its own time of day. 



Some fields of the average storm are represented on the four charts, fig. 198. 



The first chart gives the field at 13'' o m (Gr. M. T.) corresponding to the beginning of the posit iv< 

 storm. We notice that it is breaking in from the north-east. It is strongest at Axeloen and Matotchkin- 

 Schar. The current-arrows are directed eastwards along the auroral zone. For Dyrafjord, Kaafjord and 

 Matotchkin-Schar, the vertical component of the perturbing force is directed downwards, but upwards 

 for Axeloen, showing that the current goes to the north of the three former stations, but to the south of 

 the latter. 



The second chart gives the field at 17** o m , when the negative storm is on the point of breaking in 

 from the east. At Dyrafjord only the effect of the positive storm can be noticed. At Kaafjord the arrow 

 is slightly turned, and at Matotchkin-Schar even more so. At Axeloen, however, it is almost completely 

 turned to the west. It looks as if the force at Axeloen should at this hour be mostly due to current- 

 systems different from those producing the effect at the other stations, and as we shall see later, this 

 is also the case. 



On the third chart for 2i h o m , the negative storm dominates at Axeloen and Matotchkin-Schar and 

 almost completely at Kaafjord; but the effect of positive storms is still most prominent at Dyrafjord. 



On tlie last chart for i h o m , the negative storm dominates at all four stations; but it is now 

 strongest at Dyrafjord. The vertical component is directed downwards for Axeloen and upwards for 

 the other three stations, showing that the currents on an average at this time are running above the earth's 

 surface, and between Kaafjord and Axeloen. 



Through the treatment of separate perturbations we were led to the assumption of two types of 

 polar storms, which we called the positive and the negative polar storms. The statistical treatment of the 

 whole material shows exactly the same two types. 



The average storm in the afternoon has the properties of a typical positive polar storm ; the midnight 

 average storm has the properties of a negative polar storm; and we see that the predominant part of the 

 storminess, at least at the three southern stations, is made up of these two types. 



The cause of the singular character of Sy and Sy for Dyrafjord and S\ for Axeloen, will now become 

 evident. The reversal of the conditions of storminess of Sy for Axeloen only means that the storm- 

 centres of the two types of polar storms, positive as well as negative, pass between Kaafjord and 

 Axeloen. The small amplitude in the SD curve for Dyrafjord shows that the current on an average is 

 nearly perpendicular to the magnetic meridian at this place. The storminess of the vertical intensity 

 at Dyrafjord shows that the current-systems usually pass near the zenith, usually somewhat to the north 

 of the station. 



