PART I. ON MAGNETIC STORMS. CHAP. II. 155 



It is easy to follow the development from experiment i up to the last-mentioned, represented in fig. 47. 

 In experiment i we see distinctly three characteristic light-areas round the terrella. In the succeeding 

 experiments these light-areas undergo several important changes. First the strength of the light dimini- 

 shes in the middle of the areas, so that the edges come out more distinctly. Then the edges also partly 

 disappear, except in the polar regions, where the 

 light increases in intensity. 



The first figure in the three rows (i, 4 & 7) 

 shows the light reduced to two patches, the lower 

 of which, however, has coincided with a descent of 

 rays upon the screen, indicating rays that have been 

 deflected and have turned back before they reached 

 the terrella (see fig. 39, third example). 



The second figure in the three rows illustrates 

 clearly the development mentioned above. 



The third figure, as photograph 9 shows, changes 

 into polar bands that have possibly been produced 

 by the covering over of more light-areas than the 

 three mentioned. These zones of light are best seen 

 in fig. 47. Other light-phenomena are also seen in 

 photographs 3, 6 & 9, fig. 68, about the magnetic 

 north pole and on the screen. 



These consecutive light-areas round the terrella 

 have some resemblance to other light-phenomena 

 observed by me during the study of the trajectories 

 of cathode rays under the influence of one magnetic 

 pole('). With one magnetic pole, the consecutive 

 figures became constantly smaller and smaller, while 

 here they are all nearly of the same size. 



From these experiments we shall draw comparisons both now, while discussing the cyclo-median 

 perturbations, and subsequently in the treatment of the observations from 1882 83, Vol. I, Part II, where 

 the question of districts of precipitation in the polar regions for magnetic storm-centres is discussed, and 

 lastly in the treatment of the observations of aurora and of cirrus clouds (Vol. II). 



The experiments described in connection with figs. 47 and 68, are of fundamental importance to 

 our theory of magnetic disturbances. Concluding by analogy from these, we should never expect to 

 have purely elementary magnetic perturbations upon the earth, as, among other things, the experiments 

 show that there are several districts of precipitation at the same time upon the earth for the electric 

 rays from the sun. In the preceding pages also, it has frequently been indicated that the magnetic cur- 

 rents are never purely elementary, like, for instance, the idealised polar form represented in fig. 40. 



As regards polar storms, we have only been able to study those with the district of precipitation 

 in the neighbourhood of the four Norwegian stations. 



In order to obtain a clear understanding of the circumstances, we ought to have simultaneous ob- 

 servations from stations right round the auroral zone, and if possible also from the antarctic regions. A 

 year's simultaneous observations from all the acting magnetic observatories in the world, and from, for 

 instance, 10 stations in a zone round the terrestrial-magnetic north pole, and from as many as far south 



(') Archives des Sciences Physiques et Naturelles. Quatrieme periode, t. VI. Geneva, Sept., 1898. 



