Power with the Velocity of Cathode Rays. 651 



the solenoid was adjusted by resistances to a definite value 

 indicated by a Weston ammeter and the values of Q 1 and C 2 

 observed on the galvanometer scale. The solenoid current 

 was then altered so as to cause a different part of the 

 spectrum to enter the chamber and the galvanometer 

 deflexions again read. Readings were taken in this way 

 throughout the range of the spectrum. The pressure was 

 then altered and a fresh set of readings taken over the range 

 of velocities obtained. 



In practice, the range of velocities which can be used is 

 limited both at its lower and upper ends. The velocity of 

 the rays produced by the discharge depends on the pressure 

 in the tube. The highest pressure at which cathode rays 

 appear at s was found to be '21 mm. of Hg. It was found 

 that the slowest rays produced at this pressure had a velocity 

 corresponding to 6 amperes, i.e. it required a current of 

 6 amperes in the solenoid to deflect them into the ionization 

 chamber. The upper limit to the velocities employable is 

 governed by the following considerations. To produce 

 fast rays the pressure in the apparatus must be reduced and 

 consequently the number of ions produced by a ray in 

 crossing the chamber is also reduced. Now theoretically the 



most accurate values of -^ — ~ 2 ( = N) are obtained when 



vi + Oj 



the difference between Ci and C 2 is as great as possible. 

 For small errors in reading both C\ and C 2 produce a much 

 greater proportionate error in their difference. So that 

 when Cx — C 2 is small the liability to error in the value of N 

 is very great. Although a few readings were taken with a 

 low pressure and velocities corresponding to 10 or 11 

 amperes they were discarded for the above reasons and no 

 readings are given for velocities greater than 9 amperes. 



The results of the experiments in air are shown in Table I, 

 Column I. gives the pressure in the apparatus in readings of 

 the McLeod gauge ; column II. gives the current flowing- 

 through the solenoid in amperes ; column III. the mean 



Q Q 



values of the ratio n l n 2 , i. e. the actual number of ions 



^1 + ^2 



made by a ray in traversing the chamber. 



In fig. 3 (p. 653) the values of N have been plotted against 

 the pressure in the tube for different velocities of rays. It is 

 evident that the number of ions made should be proportional 

 to the pressure in the tube and the curves connecting the 

 two should be straight lines passing through the origin of 

 coordinates. The curves in fig. 3 have been obtained by 



