Theory of Magnetic Storms. (373 



Ro = 2'2.10 12 c] 



T ° /-- RqV 2U 



R = 2 , 2.1Q 12 cm. and the total number emitted therefore 



of the order ( - 1 . -I — -«, Ti being the duration of the 



storm, R the distance of the sun, and t the time of one 

 revolution of the sun. Since the only way to account for a 

 large value of v is to assume the presence of a radioactive 

 substance, and since even in this case it cannot be much 

 greater than 2.10 9 without leading to an inordinately short 

 life, this number must be large and involve an impossible 

 concentration of radioactive substance in the part of the sun 

 involved. Therefore it seems unlikely that radioactive pro- 

 cesses are involved, and v must be so small that the time 

 of the particles in transit is not short compared with the 

 duration of the storm. In this case of course the number 

 is large, but since there is no need to postulate radioactive 

 origin for the particles this does not matter. Further, as 

 was stated above, there is no undue tendency for the particles 

 to spread since their resultant electric charge is small. 



The hypothesis to be examined therefore is that an 

 approximately equal number of positive and negative ions 

 are projected from the sun in something of the form of a 

 cloud and that these are the cause of magnetic storms and 

 aurorse. It is known that such clouds are emitted in the 

 neighbourhood of sun-spots, as is observed in the protuber- 

 ances, and it will be shown that the observed velocities 

 might well be acquired by a neutral gas under the influence 

 oE light pressure. It will further be shown that the gas 

 would become almost completely ionized as the pressure 

 diminished, that it would spread out by approximately the 

 right amount without recombining on its passage from the 

 sun to the earth, that its density on reaching the earth would 

 be of the right order of magnitude, and that the observed 

 magnetic and auroral effects would be accounted for bv the 

 difference in range of the positive and negative particles. 

 The connexion of magnetic storms with the sun-spot period 

 and the tendency of storms to recur after periods corre- 

 sponding to one solar revolution, would on this theory be 

 obvious consequences of the tendency of protuberances to 

 occur above sun-spots. 



It is beyond the scope of this paper to attempt to explain 

 why masses of gas are ejected by the sun at various spots, 

 -and it is sufficient to accept this phenomenon as an observed 

 fact. Strictly speaking it might be considered enough also 

 to accept the fact that the radial velocity ot % these masses 

 of gas is comparatively great, often exceeding 8.10" cm. sec. 



