G82 Prof. F. A. Lindemann on the 



average speed v to be 8.10 7 cm. as above, one finds the total 



2U 

 mass M= =- =3'1.10 7 grammes. In realitv, of course, the 



mass will be greater than this, but how much greater 

 depends upon the efficiency h x of the stream, i. e. how much 

 of the slowing up of the ions is due to electrostatic attrac- 

 tion by the electrons which have been stopped in their early 

 encounters and how much is due to loss of energy by 

 collisions. If more were known about the outer layers this 

 could probably be estimated ; a rough attempt to evaluate 

 the order of magnitude is given below. 



This mass ■ : grammes will be diluted by at least 



2"5 . 10 I2 grammes of air, so that there is no reason to expect 



the spectral lines to show. It does not seem impossible, 



however, that the green line might be due to the unknown 



element projected from the sun. Since the gas recombines 



comparatively quickly however, the dilution is in reality far 



greater than that given above, so that the appearance of the 



spectrum of the injected gas does not seem very probable. 



If the substance leaving the sun is a gas of atomic weight 



A and carries n charges, the total charge E of either sign 



1 M 

 which is projected onto the earth is j--r-nF, F being 



/tj A 



Faraday's constant 9-654. 10 3 coulombs or 2-896. 10 u E.S.U. 



Therefore 



F _ 2 UnF _ 9-1 . 10 21 



\Av 2 ~ Aj 



if the gas is assumed to be hydrogen. tt 



The potential difference P cannot exceed I,- , i. e. 



Av 2 

 P = /c 2 — -== = 1T1 k 2 in electrostatic units. At first sight 



this may appear unexpectedly small. It must be borne in 

 mind, however, that the conductivity of a gas increases 

 enormously when the potential drop on a free path becomes 

 large enough for the ion to form new ions by collision. It 

 is clear that such a potential difference could not be 

 materially exceeded. As has been pointed out above, a 

 hydrogen ion moving at 8 . 10' cm. /sec. could, at its best, only 

 penetrate a column of air containing some 10 1(3 molecules 

 per cm. 2 . The free path of course varies with the density of 

 the gas. Assuming, isothermal equilibrium it may be shown 

 that the distance A/t between the first collision, at which 

 the electrons are presumably stopped, and the *>th collision 



PT P 



is ^rlogz>. If v=^-« where V is the ionization potential, 



gA tt V l 



