522 FilRKELAND. THK NORWEGIAN AURORA POLARIS EXPEDITION, igO2 1903. 



Let us suppose, that the position of the source is such that the pencils strike the earth and produce 

 a perturbation. Let at the moment considered the heliographic longitude of the centre of the sun's 

 disc be io, and that of the source L To explain the observations we must assume that the angle 

 A. /. cannot vary greatly for the various pencils of rays which strike the earth; for if the pencils 

 could start from the source in the most varied directions in relation to the surface of the sun, the effect 

 would be the broadning out of the maxima, or the causing of an enormous variation in the interval 

 between successive maxima. As long as the maxima keep their pointed form, and occur at fairly 

 constant intervals we are justified in assuming that the final directions of the pencils relative to the 

 sun's surface are, roughly speaking, the same. 



As the only singular direction from a plane is its normal, and as there is only one predominant 

 direction of the pencils, I think there can be little doubt that the radiation starts almost perpendicular to 

 the surface of the sun. If, after starting, the rays were not exposed to any deflecting field of force 

 from the sun, A /.<, would be nearly equal to zero. 



It has been found by many observers, that there is a lag, or interval between the passage of a 

 sun-spot across the central meridian, and the occurrence of the magnetic storms. 



On the assumption that sun-spots act directly as a source, and the velocity of propagation of the 

 radiation is at least as great as that of ordinary cathode rays, the lag would mean that /. /. had a 

 positive value. In order, then, that the radiation, starting normally, shall reach the earth, the existemv 

 of deflecting forces is necessary. Assuming that the deflection is due to a magnetic field, and knowing, 

 from other considerations, the stiffness of the rays, I have recently (') calculated the intensity of magne- 

 tisation of the sun, that would account for the observed lag. 



The active area must be comparatively limited, for it is very seldom that a storm lasts for more 

 than 24 hours. Very often several storms occur in succession at the time of the maximum, which indi- 

 cates that the active area is more like a group of active spots. 



The theory of the confinement of the electric radiation causing magnetic polar storms, to certain 

 sources, which send out narrow pencils of rays, was deduced as a natural consequence of the character 

 of the variation in storminess, and is the one that I have adopted in my previous works on these 

 matters, as, for instance, in "Recherches sur les Taches du Soleil", read before the Christiania Videnskabs- 

 selskab on Feb. 24, 1899, where, on page 2, the view is clearly expressed in the following terms: 



"Dans un me'moire inse're' aux Archives des sciences phys. et. nat. Geneve, juin 1896, j'ai chcrche 

 a expliquer la relation existant entre les taches du Soleil d'une part, et les aurores bor6ales et les 

 perturbations magnetiques de 1'autre. Dans mon hypothese le Soleil 6met de longs faisceaux de rayons 

 cathodiques, qui sont en partie 1'objet d'un succion dans 1'atmosphere terrestre de la part des poles 

 magnetiques, chaque fois qu'un des faisceaux cathodiques en question frole notre planete d'assez pres.' 



It is a matter of great interest that subsequently Mr. MAUNDER, from a study of perturbations 

 observed at Greenwich, was led to the very same conclusions. 



Nor does the physical side of the question present any serious difficulties. In the corona and the 

 comets' tails, we are actually examining radiations having definite directions. The difficulty in this 

 respect is not that we are in want of a possible explanation, but rather that we have too many 

 of them. 



In order to explain the properties of the corona and the tails of comets, it has long been supposed 

 that the sun should possess an electric field, in which case the cathode rays might leave the sun in a 

 direction perpendicular to its surface, just as, in a vacuum-tube, the cathode rays start perpendicularly tc 

 the cathode surface. 



(') K. UIRKEI.AND: C. R. 1910. 



