PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. I. 415 



lasts considerably longer at the Jan Mayen station than at Bossekop. It seems evident from this that 

 the district of positive precipitation is moving westwards. 



This movement, which has so often been mentioned, and which has undoubtedly some connection 

 with the earth's rotation, is here very distinct, as the perturbations concerned are of longer duration 

 than usual, and perhaps also because they are at a time not very distant from the equinox. 



In Jan Mayen, however, forces soon appear which seem to counteract the effects of the positive 

 storm ; a negative system seems to encroach upon the positive for a short time, and once or twice cause 

 a reversal of the values of P/,. 



This negative storm is evidently the same that appears at Cape Thordsen, but here it is far more 

 powerful. The effects of the positive storm are slight. Before n h 20 the horizontal-intensity curve 

 at the latter station oscillates about the normal line, perhaps the result of the action of alternate slight 

 positive and negative precipitation. After n' 1 2o m , however, a very well defined negative polar storm 

 appears, which developes and reaches its maximum simultaneously with the positive storm in the south. 

 Simultaneous serrations are also frequently to be found, a circumstance which indicates the connection 

 which evidently exists between these cases of precipitation. 



A comparison of the horizontal-intensity curves for Cape Thordsen and Jan Mayen will give a 

 distinct impression that it is the negative storm that breaks in upon the positive, and produces the 

 peculiar phenomena found in Jan Mayen. That the positive storm is going on all the time seems to be 

 clearly evident, however, from the fact that simultaneously with the disappearance of the negative storm 

 at Cape Thordsen, the positive forces once more assert themselves, and the positive deflections then 

 diminish just as at Bossekop. It is also characteristic that at Bossekop too, the negative storm intrudes 

 and produces the peculiar curve that we find at about i6 u . 



The horizontal-intensity curve at Little Karmakul also shows clearly a condition exactly similar to 

 that in Jan Mayen, namely a long positive storm, upon which the somewhat shorter negative storm 

 intrudes. For a time too, the latter is the stronger, just before it reaches its greatest height. At i6 h , 

 however, positive forces once more appear, evidently the same strengthening of the positive system as 

 at Sodankyla. After that hour the curve oscillates about the normal line, thus indicating the supremacy 

 of the positive and negative forces alternately. In declination, however, the direction of the deflections 

 is nearly always the same, namely westward; but here too, the curve is exceedingly jagged and dis- 

 turbed in character. 



At Fort Conger, the last of the polar stations, it will be seen that the declination-curve very much 

 resembles that at Cape Thordsen, and we may therefore assume that the system continues westwards 

 through that station. 



At the southern stations, the deflections are evidently governed by the precipitation in the arctic 

 regions; and we sometimes find a very distinct resemblance between the various serrations. The deflec- 

 tions in the horizontal-intensity curves for Christiania and Pawlowsk are not constant in any part of the 

 period, but are sometimes in one direction and sometimes in another, although the negative deflections 

 predominate. Farther south, on the other hand, e. g. at Gottingen, we find negative deflections all the time. 



In declination we find the deflections for the most part directed westwards at Christiania and 

 Gottingen, whereas at Pawlowsk there are no very considerable forces in that component. In the vertical- 

 intensity curve at Pawlowsk a very distinct positive deflection appears. 



The conditions at Gottingen are exactly similar. The rise in the vertical-intensity curve at about 

 7 h , and the fall at about n h , are undoubtedly connected with the diurnal variation, while the last rise 

 with a maximum at about i6 h seems to be connected with the perturbations. 



These conditions, the distribution of the districts of positive and negative precipitation over the 

 various regions of the earth, and their intermingling, are thus in perfect accordance with our previous 



