272 BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, igO2 1903. 



February, and the 27th and 3ist October (see the corresponding Plates), where the change, however, 

 was of an opposite kind, a more easterly negative storm seeming to encroach upon the westerly posi- 

 tive storm for a time. On the gth December, 1902 (PI. IX), there is an example of still more typical 

 conditions. At Dyrafjord and Kaafjord the arrows have strongly-marked easterly directions. The pro- 

 nounced westerly directions at Axeleen are, we are inclined from the above to think, a continuation of 

 a more easterly-situated negative polar storm. At Matotchkin Schar, on the other hand, we find that 

 now one storm, now the other, seems to be the stronger, so that the directions of the arrows are 

 always swinging round from west to east, or from east to west. These conditions, however, can be 

 better studied in the material from 188283, where we have at our disposal observations from a 

 larger number of polar stations. 



This sudden change may be illustrated by imagining the two systems like those in fig. 50 a & b, 

 moving together until they are lying close to each other, and imagining the rays to the east deflected 

 as in fig. 50 a, and those to the west as in fig. 50 b. If we imagine a system such as this displaced, 

 we shall obtain conditions at those places through which the boundary between the two kinds of 

 polar storms passes, similar to those found at Matotchkin Schar. 



The fifth polar storm, or second intermediate storm, shown in Charts VII X, also exhibits in its 

 main features the same peculiarities as the long storm. The explanation of the change we here see 

 should apparently be sought in a suddenly strengthened impulse in the polar system, whereby the 

 latter, in southern latitudes, acquires a greater effect. This causes the area of convergence here too, 

 to appear more distinct, the effect of the polar system being for a time greater than that of the equatorial 

 storm ; and we obtain current-arrows pointing eastwards (see Chart VIII). The area of divergence also 

 becomes stronger, and it thus appears that in this system too, there should be an impulse at the 

 same time. 



Finally, with regard to the sixth polar, or third intermediate storm (Chart XV), the conditions are 

 quite analogous. There is -an increased impulse in the polar systems, especially in the negative, an 

 increase which is only slight, although relatively strong, the perturbing forces now being very small. 

 The equatorial storm still seems to have an effect which acts in the very opposite direction in the 

 south of Asia, but in America in the same direction as the polar systems. 



In this way we have succeeded in explaining all the above phenomena in a manner that is exactly 

 analogous to that employed in the preceding perturbations, and based only upon our previously-discovered 

 simple elementary phenomena. 



THE PERTURBATIONS OF THE 23rd & 24th NOVEMBER, 1902. 



(PI. VIII). 



70. After the powerful storms at the end of October and the beginning of November have ceased, 

 conditions are fairly quiet, at any rate at the Norwegian stations; and the few perturbations that do 

 occur are of comparatively small strength. On the igth November, however, quite a powerful pertur- 

 bation appears rather suddenly. This forms the introduction to a series of powerful perturbations which 

 develope daily for rather more than a week, the last powerful storm being on the 26th. These storms 

 reach their maximum of strength between the 23rd and the 25th. The conditions recall those in October, 

 when there was a similar period of powerful storms. 



We remarked then that the position of the moon must have exercised an influence upon the 

 behaviour of the perturbations, as the maximum occurred just about the time of the new moon. On this 

 occasion too, we are in a period not far from the new moon; but the maximum does not coincide with 

 it in time. The most powerful storms occurred, as we have said, between the 23rd and the 25th November ; 



