Rontg en Radiation from Gases^and Vapours. 661 



was measured, with the rays off. This gave a measure of the 

 difference between the spontaneous ionization in the two 

 ionization-ch ambers, that is o£ S x — S 2 . As the ionization- 

 chambers had been made as nearly equal as possible, this was 

 a very small amount, being usually about 5 or 6 divisions of 

 the scale in 10 minutes. It varied, however, in a curious 

 manner from day to day, and also in the course of the day, 

 being greatest in the morning, decreasing gradually until 

 late in the afternoon, when it reached its minimum value, 

 and then increasing again during the evening. This phenomenon 

 has been recently described, and studied in a paper by Wood 

 and Campbell*. It is easily eliminated by taking the readings 

 with and without the X-rays. 



The pressure in one or other of the gas -chambers was now 

 altered until the rate of leak of the electroscope was the same 

 whether the X-rays were on or off. When the apparatus 

 was working well, it was quite possible to obtain a balance 

 to within two or three millimetres of mercury in the pressure- 

 gauge. Let pi, p 2 be the pressures in A x and A 2 respectively 

 when this balance is obtained. Then, since the rate of leak 

 of the electroscope is the same whether the rays are on or 

 off, we must have the total ionization in the ionization-chamber 

 over A x less the total ionization in the ionization-chamber 

 over A 2 equal to the difference between the spontaneous 

 ionization in the two chambers. Thus we must have : — 



- (k P2 GY 2 + kC 2 + S 2 ) - S, - S 2 , 

 or on simplifying, 



Ih&oVi + Ci =p 2 G-V 2 + G 2 . 



Since the gas-chambers are subjected to the action of the 

 rays simultaneously, the intensity factor k, which depends 

 only upon the way the bulb is working, will obviously be the 

 same for both. 



The gas under experiment is then removed from A 2 and 

 the latter filled with dry, dust-free air. Keeping the pressure^ 

 in A x constant, the pressure of the air in A 2 is now altered 

 until there is again no difference in the rate of leak of the 

 electroscope whether the X-rays are on or off. Let p 2 ' be the 

 pressure in A 2 when this balance is obtained. 



Then, as before, we have 



(k' Pl G Q N 1 + //d + S,) - (Fp/GoV 2 + k'C 2 + S 2 ) = S : - S 2 , 



or p l G Y 1 -f L\ =p 2 'G V 2 + C 2 . 



* Phil. Mas Feb. 1907. 



