( 429 ) 
in every plane cutting the axis at right angles was nearly radial, 
the magnetic intensity therefore, on account of its solenoidal distri¬ 
bution, being inversely proportional to the distance from the axis. 
The basis and the upper surface of the two cylinders consisted in 
two horizontal metal plates. The discharge passed nearly vertically 
from a point of the upper plate to a point of the basis, and was 
carried round about the axis describing a third coaxial cylinder. 
If the space between the two cylinders C t and C\ had been filled 
up with an electrolyte, the whole mass would have rotated under 
the influence of the magnetic field; the same being true of course if 
the motion of the charged particles had taken place in a solid body. 
In the experiments of Wilson and Martyn the gas must likewise 
have been carried along, but the motion of the gas must have been 
so slow that the discharge may indeed be considered to have revolved 
about the axis in a medium at rest. The truth of this assertion 
appears if we calculate the velocity of the gas and compare it with 
the observed velocity of the discharge. 
As to the formula (1), we can show that it may be obtained in 
another way than Wilson and Martyn’s. The mechanism of the 
discharge is not necessarily such as they assume it to be. 
$ 1. Let the radius of the larger cylinder be R t that of the 
smaller one R % . Let us suppose that the discharge does not start 
from one point only but simultaneously from all points situated 
etween two cylinders, coaxial with the former, of radius Q t and 9, 
re sp. (pj > p s ). A circular discharge of this kind was indeed observed 
ui some of the experiments. It is a reasonable supposition that when 
. e ^charge does start from one point only, the velocity of revolu- 
tion of the gas is of the same order of magnitude as the velocity 
here calculated. 
Let / be the total current. Since the section of the discharge is 
q s s ), the density of the current is 
1 follows that the current between the cylinders with radius r 
nd r -\~dr resp. is 
( 3 ) 
