PART I. ON MAGNETIC STORMS CHAP. IV. 31 1 



On the 27th, at 23*, 6, as we see, is imaginary. This shows that the perturbation-conditions at 

 the two stations at that moment do not satisfy the assumptions made. The reason of this is possibly 

 to be sought in the cross-section that the actual current must have, or perhaps in the fact that the 

 perturbation-conditions could in no way be ascribed to the effect of a more or less aggregate system. 

 We might have several simultaneously-acting systems of to some extent more local character. We 

 very frequently see at these stations in the north, that disturbances occur at one station that are not 

 noticed at another. We shall never be without these local disturbances; but the thing is that they shall 

 be slight in comparison with the total effect. 



A great local disturbance seems really to occur just about 23 b . There is a sharp deflection of 

 rather long duration, which tends to increase /V'- From the fact that there is no corresponding change 

 at Kaafjord, we may conclude that this deflection cannot be ascribed entirely to a movement of the 

 main system. 



We also, by looking at the curve for P v , obtain the impression of a new system, which would lie 

 to the north of Axeleen, as the deflection is in the opposite direction. 



At the second hour, 23 h 2o m , the great local disturbance at Axeleen is over, or at any rate fainter, 

 and we now obtain a real solution. The calculated vertical component for the time, however, is some- 

 what smaller than the observed. This is also the case on the day following. 



THE ENERGY OF THE CORPUSCULAR PRECIPITATION. 

 THE SOURCE OF THE SUN'S HEAT. 



80. We consider it to be beyond doubt that the powerful storms in the northern regions, both 

 those of long duration, and the short, well-defined storms that we have called elementary, are due to the 

 action of electric currents above the surface of the earth near the auroral zone. 



These currents, as far as the elementary storms are concerned at any rate, act, in the districts in 

 which the perturbation is most powerful, as almost linear currents, that for a considerable distance are 

 approximately horizontal. In the preceding articles, we have attempted, in some of the magnetic storms 

 described, to calculate the strength of horizontal currents such as might be the cause of the storms, 

 supposing that they acted magnetically as galvanic currents. The values found, which cannot certainly 

 lay claim to any great accuracy, will yet give an approximate idea of the strength of these currents. 



In the case of the greater storms, we found current-strengths that varied between 500,000 and 

 1,000,000 amperes, or even considerably more. 



It might be interesting to know the amount of energy per second of this current. According to 

 my hypothesis, the currents would not, in reality, be galvanic, but be formed of cathode rays, or more 

 generally of rays of electric corpuscles. We will make this hypothesis, then, the basis of our estimates. 



By energy we in the mean time understand the kinetic energy of the corpuscular current that passes 

 per second through a cross-section of the horizontal part of the current, and where the corpuscles are 

 assumed to flow in the path of the before-mentioned galvanic current. In Article 36, fig. 50 a & b, we 

 gave a diagram of the manner in which we in reality approximately imagine the corpuscles to move. 

 With the method of calculation here employed, we obtain only a small lower limit of the energy of the 

 corpuscular current. 



If we call the number of corpuscles that pass the cross section in the time-unit n, the apparent mass 

 of a particle //, and the velocity v, we obtain as the energy W. 



W = \ nf.iv' 1 . 



