PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. V. 617 



If condition (3) is fulfilled, the range of the radiation will be R-, as given by equation 2 b. 

 We suppose a to be constant and let c vary. 



When c decreases towards -=- , R z will increase and approach the value R-, = c. 



y2 1 



The greatest range which the rays can have without going towards infinity will be 



R-i = -, a - = 2.414 a. 



\2 I 



We then get the very simple result: 



If electric radiation starting from the surface of a sphere in the plane of the magnetic equator, and 

 only subject to the influence of the magnetic field of the sphere, reaches a distance from the centre 

 greater than 2.414 times the radius to the sphere, the radiation will not be able to return to the sphere, 

 but will pass on towards infinity. This result will hold independent of the magnetic moment of the 

 sphere and the stiffness of the electric rays. 



This result supposes that relation (i) holds good close up to the surface of the sphere. This relation 

 actually holds good provided the sphere is uniformly magnetised or it will be more or less true for 

 any magnetisation which makes the magnetic force in the magnetic equator a function of the distance 

 from the centre. 



If the radiation shall return to the sphere, the following condition must hold: 



c> 2.414. a or 



iw- 



M > 2.414. a. 



This result corresponds to the rays starting in the direction 6 = - 



2 



If we consider the radiation starting normally we get 



^2, max .,= 2rt, 



or if radiation starting normally reaches a distance greater than 2 a from the centre, it will pass on to 

 infinity. 



If the radiation starting normally shall return to the sphere, we must have 



c > 2 a or 



M 



^ M 



Application to the sun. 

 In order that radiation shall emanate from the sun 



.86X 10** Ho, 



when starting in the direction B = - and 



m 



when starting normally. 



For the stiffest ft rays starting normally we get 



.9 X 

 and for a rays 



X IO 27 . 



When M is of the order io 28 as estimated by me in C.R. Jan. 24, 1910, it supposes that 

 o > 5 X io 5 for normally starting rays if the rays shall be able to emerge into infinity. 



