﻿12 G Mr. Clinton J. Davisson on the Positive 



presents difficulties which are great compared with ^ the 

 usefulness of an exact solution. Richardson (I. c.) has given 

 an approximate solution for the displacement of the maximum 

 of i 2 j (ii + i 2 ) on the assumption that the initial velocities of the 

 ions are small compared with the velocities which they 

 acquire in the electric field, and on the further assumption 

 that ions which leave the strip normally and reach the point 

 of maximum current density when H/ = 0, also reach the point 

 of maximum current density when IL^O. These assumptions 

 have been tested by calculation and experiment and have 

 been found sufficiently valid to insure values of elm accurate 

 to within a few per cent., which is greater than the accuracy 

 of the measurements. 



The approximate solution referred to may be written 



Hd 2 / 2 



3 V v 



7 



m 



where x is the displacement of the maximum of 4, 0'i -l - ^)* an( l 

 H is the value of H^. 

 From this we have 



e 9 Yx* 



m 



2 rW 



In this expression, with the exception of .r, the quantities 

 upon which e\m depends admit of ready and fairly accurate 

 measurement. The quantity x, which is the displacement 

 in cms. of the maximum value of i 2 /(ii + 2*2) by the magnetic 

 field H, is measured in the following manner : — The current 

 which heats the strip (fig. 1) is gradually increased until 

 the thermionic current from the salt is sufficiently large and 

 steady for measurement. The magnetic field and electrical 

 potential are then adjusted to appropriate values and deter- 

 minations of 6. 2/6 are made, as already described, with the 

 slit in a series of different positions. The screw used in 

 changing the position of the slit has a pitch of 0*0635 cm., 

 and the setting can be accurately made to within less than 

 1/20 of a turn. 



The values of 0J6 which represent approximately 0'041 

 x i 2 /(i 1 + 1-2) * ire plotted against the positions of the slit and 

 readings taken until the data indicate that the slit has parsed 

 through the maximum in ij{h + i$)* The magnetic field is 

 then reversed and the measurements continued until the 

 maximum has been located in its new position. Curves are 

 drawn through the two sets of points. The distance between 

 the two peaks indicates the value of 2x. 



