CH. v.] DIRECT PROCESS 57 



with the density of the residual gas and with the 

 intensity of the electric field. 



In order to prevent the diffusion and spreading out 

 of the rays, on their passage down the tube, a longi- 

 tudinal magnetising helix was applied to it, so as to 

 concentrate the rays along the axis. 



If u is the velocity to be measured, and I the 

 distance they travel to give the first zero, then 

 u knl, where n is the frequency of the electric 

 oscillation in the circuit MN. 



In one experiment I was 39 centimetres, and n 

 was 32 million per second; wherefore u comes out 

 4*5 x 10 9 cm. per sec. 



As soon as the velocity is known, the ratio e/m can 

 be determined by measuring the magnetic deflexion 

 of the rays. The range of uncertainty for this 

 determination, as made by Wiechert, lies between 

 1*55 and 1*01 times 10 r . It can hardly be con- 

 sidered the most accurate method of measurement, 

 but its directness and ingenuity entitle it to atten- 

 tion. 



