Residual Ionization in Gases. 419 



leather washers. The lower end of the covering-tube A 

 had a ground conical surface, and was held hermetically 

 connected to the body of the instrument by means of a 

 threaded shoulder-piece. All the joints were carefully 

 examined before each experiment, and found to be air-tight 

 before measurements were undertaken. All the gases used 

 were carefully dried before being admitted into the instru- 

 ment. The latter, however, was provided with a small 

 drying-chamber opening directly into its interior, and a 

 small glass vessel in this chamber was always kept filled 

 with metallic sodium. 



In all experiments made with air it was found that the 

 ionization was the same whether the sodium was present or 

 not. 



The theory of the Wolff electrometer has been given 

 already in the paper by McLennan & McLeod referred to 

 above, and it will suffice here to give only the final formula, 



N = 1'03(AV-AV 1 ), . . . . (1) 



for determining the number of ions generated within the air 

 in the electrometer per cubic centimetre per second. In this 

 formula AY is the loss in the voltage per hour of the 

 ■electrical system with the tube SS up and the ions coming 

 to the system from the gas in the whole of the chamber, and 

 AVi the loss in voltage per hour with the tube S S down 

 and the ions coming to the system from the volume of gas 

 -enclosed by the tube, i. e., 31'5 c.c. 



III. First Investigation, 



In the first set of experiments the gases used were air, 

 hydrogen, and acetylene. The observations were made in 

 the open at Bowland, Scotland, in a hotel in London, 

 England, on the Atlantic Ocean in s.s. ' Megantic,' and in 

 the Convocation Hall of the University of Toronto, the last- 

 mentioned building being a new structure and one into 

 which no radium or other radio-active substance had been 

 brought hitherto. 



The readings taken at the:ie various stations in the 

 different gases are recorded in Table I. Those at Bowland 

 with hydrogen, acetylene, and the mixture of methyl iodide 

 and air were taken in a hurry as the time available was 

 short. The acetylene used was taken from a private house 

 supply, and was carefully dried before being used. The 

 hydrogen was generated in an improvised piece of apparatus 



2 E 2 



