Theory of Magnetic Storms. 675 



the current per cm. 2 is ~/-e. Since l//x is about 100 times 

 larger than X for most substances the current may be 

 assumed to be - - 1 -\ 



The electromotive force necessary to balance this is negli- 

 gible. Since plenty of electrons are present X = — 



i being the current, n\ the number of electrons per cm. 3 , 

 m the mass and e the charge, r the velocity due to thermal 

 agitation, I the free path, and X the potential gradient. If 

 N x is ihe number of atoms per cm. 3 and a their diameter, 

 therefore / = 1/N ^a 2 , so that if v x is the fraction of atoms 

 ionized, pi the fraction of the total mass consisting of radio- 

 active matter of average life r, and a the atomic weight of 



the electron 1/1830, one finds X=l"6— -. — . — . — , 



where A is the atomic weight of the radioactive substance. 

 On the earth, and presumably on the sun, since a tempera- 

 ture of 6000° is far too low to influence radioactivity, A is 

 of the order 200 and /jl about 5. p : must be less than 1 and 

 r greater than 10 -9 , since the sun does not change funda- 



i 11 • qa xw v- 1*2. 10" 11 volt _, . 



mentally in oO years, so that X< . Inis 



J J v { cni. 



figure is almost certainly still far too large, but even were it 



accepted and v x put as low as 10 -6 , the potential gradient in 



the chromosphere could not exceed 1'2 volts per kilometre. 



It seems certain, therefore, that no intense electrostatic 

 force can obtain in the solar atmosphere, and the high 

 velocity of the protuberances cannot be attributed to electro- 

 static repulsion. There is therefore no reason to assume 

 that they are charged electrically. Prof. Strutt has shown 

 that such velocities cannot be due to ordinary gaseous 

 expansion, and one is therefore reduced to considering the 

 possibility of their being caused by radiation pressure. 



Without going into details it is clear that an atom which 



scatters the energy hv subtracts the momentum from the 



incident radiation. If the incident radiation is due to an 

 incandescent sphere it therefore acquires a radial velocity 



— , m being its mass. If \ r is its ionization potential the 



atom will be ionized, /. e. the vibrating electron removed 

 from it altogether if it absorbs the radiation 7tv =eV, <' being 

 the charge on the electron. It will then of course no longer 



