734 Mr. A. E. Garrett on Electrical 



The following were the results obtained: — 



Temperature in degrees Leak with voltage sufficient to 



Centigrade. produce saturation current. 



230 22 



242 35 



248 52 



252 67 



254 80 



256 110 



258 150 



259 180 



When these are plotted with ^ for abscissas and 



^ log e — log e I for ordinates, it is seen that a decided break 

 occurs in the straight line in the neighbourhood of temperature 

 250° C. 



The probable reason for this is that some fresh source of 

 ionization is tapped at this temperature. It was subsequently- 

 found that at reduced pressures, negative ions as well as 

 positive could be detected at that temperature. 



III. Experiments on Halogen Salts of Zinc at reduced 

 pressures. 



The apparatus now used is shown in fig. 3. A is a glass 

 tube ; B, C, w r ires for heating ; D, ebonite stopper made 

 airtight ; F is the electrode ; G, iron tube earthed by means 

 of H ; J is a gauge ; K, thermometer ; L, iron dish con- 

 taining a small platinum dish in which materials to be 

 tested were placed. Experiments with zinc bromide were 

 tried with this apparatus. The pressure was reduced to 

 200 mm. of mercury, and it was then found that the leak 

 with positive ions could be detected at somewhat lower 

 temperatures than when the substance was heated at atmo- 

 spheric pressures. The following is a typical set of readings 

 obtained when the electrode F was charged negatively to a 

 potential of 80 volts : — 



Temperature, 



Leak 



in degrees Centigrade. 



(scale-divisions). 



195 



7 



207 



17 



216 



33 



From this it appears that extra leak can already be detected 

 before 200° C. is reached. 



