iL't 



. A. S.MIT1IKI.I.S. II. M. DAWSON. AND II. A. WILSON: Kl.KCTlMCAI. 



differences of conductivity should be observable when salts in suitably concentrated 

 solution were sprayed in the flames of coal-gas and cyanogen. For this purpose 

 j'jj normal solutions of potassium salts were employed. In a coal-gas flame solutions 

 of this concentration conduct equally well, but in a Cyanogen flame, where no water is 

 produced by combustion, such solutions should act like the more concentrated solutions 

 in a coal-gas flame, that is to say, their individual conductivities should become 

 apparent. 



In carrying out the experiments the limited amount of cyanogen available* at one 

 time necessitated some alterations in the apparatus. The vessels for the collection of 

 coarse spray were removed, and the flame tube connected directly to the tubulated 

 cylinder, into which the sprayer projected. The flame was not separated into its two 

 cones, and just enough air used to keep the sprayer steadily in action. The amount 

 of salt entering the flame was not greatly different from what it was with the 

 apparatus in its ordinary form. 



The following results were obtained with an E.M.F. of 5 '5 5 volts : 



After each of the above experiments a measurement was carried out with a coal- 

 gas flame under exactly the same conditions of pressure. This gave the following 

 numbers : 



Subtracting the conductivity of the flame without salt, and correcting for a slight 

 variation in the condition between the two cyanogen experiments, we have the 



following values : 



One cubic foot stored over mercury in an iron gas-holder. 



