PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. I. 329 



we sec, upon the boundary between these two regions, and at the hour in question has a current-arrow 

 directed southwards, which may be interpreted as a resultant of the effects of these two systems. 



The sun has now almost reached the meridian of the magnetic axis. 



We will now consider the further course of the perturbation upon the succeeding charts. 



Chart IV also represents three epochs, namely, 16'' jo'", 77* 20"' and 77* 40"'. 



The fields on this chart have, in the main, exactly the same appearance, the only difference being 

 that the strength of the perturbing forces at the various stations has undergone certain alterations. 



The positive storm now appears at first only at Bossekop, and then in the district about Bossekop 

 and Little Karmakul. The perturbing forces there are now very considerable, and at the same time 

 the forces arrange themselves at the southern stations in a manner that accords very well with what, 

 from our previous investigations in Part I, we should expect to find. This, at any rate, is the case at 

 the nearest stations, Sodankyla, Pawlowsk and Christiania. 



Between Bossekop and Sodankyla the forces diminish greatly, in accordance with the fact that the 

 point of divergence is being approached. At Pawlowsk this point has been passed, and the direction of 

 the current-arrow is the reverse of that at the two stations just mentioned. The forces at Christiania are 

 also what they would be if there were an area of divergence in that region; and at Gottingen also, the 



accordance is in a measure satisfactory. 



i 

 We have seen that the perturbing forces during this period first appeared with considerable power 



at Bossekop, and then at Little Karmakul. Whether this is a displacement of the positive system, or 

 only owing to an increase in the size of the area of precipitation, is a question about which there may 

 be some doubt. If we look, however, at the area at the stations situated a little farther south, the pro- 

 bability seems to be in favour of the first alternative. Unfortunately we have no observations from the 

 district in, or south of, the auroral zone west of Norway ; where there would undoubtedly have been 

 marked effects of the positive system of precipitation, which would have been of some assistance in 

 studying it. We must thus, in employing the more southerly stations, once more make use of the same 

 method of procedure as in Part I. In the present case, however, we have a station, of which the situ- 

 ation in this connection is of no small interest, and which was wanting in the former observations, na- 

 mely Christiania. This station, in connection with Pawlowsk, will be, as we shall see, of much service 

 in finding a kind of limit for the positive area of precipitation. 



At i6 h 5o m the arrow at Pawlowsk shows that this station is now in the eastern part of the area 

 of divergence, while Christiania at that time is probably not far off the transverse axis of the system. 



At I7 h 40 Pawlowsk is in the vicinity of the transverse axis, while Christiania is then evidently 

 in the western portion of the area. 



These circumstances thus appear to indicate that this is rather a movement of the system, than 

 an increase in the size of the precipitation-area of a system which does not change its position much. 



The conditions at Gottingen also to some extent agree fairly well with the above, although the 

 direction of the arrows there seems perhaps to be a little too southerly. 



The conditions in Jan Mayen are rather interesting too. They show that the positive system there 

 must lie to the south of the station. The inconsiderable forces occurring in a horizontal direction may 

 be naturally explained by assuming that the negative system to the north, and the positive system to 

 the south, neutralise one another's effect in a horizontal direction, but on the other hand act together 

 in a vertical direction, so that the aggregate effects are all the greater. 



As regards the vertical intensities at the other stations in the positive polar area, the conditions at 

 Bossekop show that the area of precipitation must be looked for somewhat to the north of that station. 

 This has also been the case in most of the previous instances of similar storms in Part I (see perturba- 

 tions of nth, 23rd and 313! October, and gth December, 1902, and 8th and 15th February, 1903). At 



