﻿1034 Prof. Townsend and Mr. Bailey on the 



tapered and ground to fit into metal sockets in the outer case 

 of the instrument, the wax used for sealing being applied 

 only on the outside of the joint. A great improvement was 

 thus obtained, and two instruments of different dimensions 

 w T ere constructed, one with a slit 2 centimetres from the 

 receiving electrodes suitable for measuring velocities in 

 gases like argon where the lateral diffusion of a stream of 

 electrons is very wide, and the other similar to that which 

 had been previously used with the slit 4 centimetres from 

 the receiving electrodes. When tested with hydrogen no 

 change was observed in the velocities after the gas had been 

 in the apparatus for several days, and with pure argon the 

 changes in two or three days were extremely small. 



In the instrument with the slit 2 centimetres from the 

 receiving electrodes the guard-rings and the electrodes were 

 fixed in the positions shown in fig. 1. The electrons are set 

 free from the copper plate P by ultraviolet light admitted 

 through a quartz plate sealed in the cover of the instrument, 

 and the stream of electrons that passes through the gauze Gr 

 and the slit S is received by the electrodes E l5 E 2 , and E 3 . 

 These electrodes were mounted on two strips of plate glass 

 fixed to the guard-ring R 1? so that the upper surfaces of the 

 electrodes were in the same plane with the upper surface of 

 the ring. The ring R^ was 7*8 centimetres internal diameter 

 and 11 '6 centimetres external diameter, and was at zero 

 potential. The ring R 2 , of the same size as R l? was insulated 

 and fixed at a distance of one centimetre from R^ The 

 slit S was 2 millimetres wide and 1*5 centimetres long in a 

 sheet of silver foil stretched inside the brass ring A, and 

 fixed at a distance of 2 centimetres from the receiving- 

 electrodes. The gauze of silver wire G was at a distance of 

 3 centimetres and the plate P at a distance of 6 centimetres 

 from the receiving electrodes. A uniform electric field was 

 obtained by maintaining the ringR 2 , the plate A, and the gauze 

 Gr at potentials V, 2V, and 3V proportional to their distances 

 from the receiving electrodes E. In most of the experiments 

 the plate P was maintained at the potential 6V, and the 

 electric force from this plate to tha gauze was the same as 

 the force in the lower part of the field. The object of the 

 gauze was to ensure that the electrons should have attained 

 the steady state of motion corresponding to the force Z in the 

 lower part of the field before passing through the slit. This 

 condition may be obtained without the gauze by fixing the 

 plate P at the potential 6V, and for experiments with gases 

 at low pressures this gauze is unnecessary. But with large 

 pressures above 20 or 30 millimetres the currents become 



