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



should be northerly. As we have no observations of horizontal intensity here, we are unable to verify 

 this; but it may perhaps be worth while to point out an interesting harmony with the conditions 

 in the area of convergence. We see from the chart that Fort Conger and Pawlowsk are situated more 

 or less symmetrically one on each side of the principal axis of the system. The tangents to the mag- 

 netic meridian at the two places are moreover more or less parallel. (The declination at Pawlowsk is 

 very near 0, and we see that the line magnetic N S, which is drawn on the chart through Fort 

 Conger, is very nearly parallel with the meridian of 30 east longitude.) 



As the forces on the two sides of the principal axis would probably be more or less symmetrical 

 in arrangement, we might perhaps expect to find a certain amount of symmetry in the declination- 

 deflections at the two places. When the declination-curve at Pawlowsk swings out to the west, the curve 

 at Fort Conger should swing out to the east, and vice versa. This will be immediately apparent if we 

 imagine the polar elementary field (fig. 40, p. 86, Part I) placed with its principal axis along the auroral 

 zone in the north of Europe. If we here imagine the storm-centre to move from time to time, and as 

 a consequence the current-arrow at Pawlowsk to turn clockwise, the current-arrow at Fort Conger will 

 turn through a corresponding angle counter-clockwise, and vice versa. 



It will be seen from Charts I IV, that we now have before us considerable oscillations of the 

 current-arrow at Pawlowsk, and it would therefore be another reason for now being able to find a corre- 

 sponding movement at Fort Conger. If we compare the declination-curves at about o 1 ', we do actually 

 find a similarity in form, which at first glance may seem unimportant, but which nevertheless is quite 

 characteristic. It is at this time, too, that the negative storm is most strong and the area of precipitation 

 so far concentrated, that one might expect to find- similar conditions as mentioned. 



The reason why the normal line is situated differently at the two stations, may only be that the 

 situation of the stations in respectively the areas of convergence and divergence, is a little different. It 

 is the form of the curve that gives the change in the force's strength and direction from time to time, 

 and the normal line that gives the absolute values of the force. In comparing the curves, it must of 

 course be remembered that the scale at Fort Conger is considerably larger than that at Pawlowsk, so that 

 the variations in the perturbing forces at work are somewhat similar in magnitude. 



In the interval between Chart II and Chart IV, the current-arrow at Pawlowsk, as we sec, makes 

 a considerable turn clockwise. During the same period, Pd at Fort Conger changes from east to west, 

 which means that the current-arrow, if assumed to have a component in a northerly direction, turns a 

 certain angle counter-clockwise. In the interval from Chart I to Chart II, in which the movement at 

 Pawlowsk is certainly distinct, but slight, nothing can be decided, as we do not know P/, at the other 

 station, and there is little variation in Pj. 



We must, of course, be careful not to attach too much importance to this circumstance, and the 

 apparent harmony between the actual perturbation-conditions and theory; but on the other hand, this has 

 a special interest, as it is one of the very few cases in which we seem able to trace the areas of both 

 convergence and divergence of the same polar elementary storm. 



This movement of the current-arrows, which we see, at any rate, distinctly in the area of conver- 

 gence, should therefore indicate that the storm-centre was moving eastwards during the perturbation. 

 The conditions at Little Karmakul, however, do not seem to indicate any such movement; on the con- 

 trary, the perturbing force diminishes here rather rapidly, and then, from Chart IV, changes. The field 

 in the first three charts does not, however, present any difficulties, as we only need to assume that the 

 district of precipitation to the east of the European stations is rather more northerly in situation than it 

 is in these regions. This is not at all at variance with what we have seen before, for even in Part I we 

 have drawn attention to the fact that the negative areas of precipitation on the day-side would be situated 

 a little farther north than those on the night-side. 



