PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. VI. 703 



36JU/ -f 4 ( 2,1/r 2 33alM) > 







id, as r=, 



But we found above that 



Hence it followed that 11 > 0, which however is not the case. It is hereby proved that the particle 

 iiinot describe a path that asymptotically approaches the boundary-circle. 



The question might now arise as to whether the particle could move in such a manner that it did 

 st reach the boundary-surface, either after a finite or an infinite length of time. This would only be 

 )ssible if the integral curve we obtain from (6) -- which may be said to be the curve of projection 

 f circles r = constant in a meridian plane of the path of the particle -- has an infinite length within 

 e boundary-surface. It would then be an important point to decide whether the path of the particle 

 mid approach asymptotically a closed curve. 



We have however not yet succeeded in solving this problem quite generally. 



Let us now at last try to find out, whether trajectories could exist in the plane 



z = kx. 

 The equations of motion are 



dy dx 

 ~~~ 



By substitution of z = kx we obtain 



3iMx t 

 ~ r> \* 



dy_ 

 ttt 



Here we must assume, that k ^ 0. 

 From the ist and 3rd equation we obtain 



(a) 



