50 Dr. E. H. Cook on Experiments 



takes place at the negative than at the positive, and generally 

 the same conditions apply as in the case of potassic iodide. 

 Thus about four times as much iron is oxidized from the 

 ferrous to the ferric condition by the negative as by the 

 positive brush. 



5. Effect on the Electrodes. 



It is stated* on the authority of Wheatstone that : — 

 " Metallic dust is in every case torn away from the electrode 

 by the brush discharge/' This statement is one which would 

 be supposed to be true when it is considered that such is 

 known to be the case with the spark-discharge. But it is 

 contradicted by spectroscopic evidence, for the spectrum of 

 the glow-discharge shows no trace of metallic lines. It is the 

 same u whatever the nature of the metal, and is due solely to 

 the incandescent gas/'f In order to test the statement 

 several experiments were made. The brushes were obtained 

 from copper points, and made to play upon the surface of 

 some dilute nitric acid placed in dishes under them. The 

 points were brought as near as possible to the surface, i. e. 

 as near as possible without producing a spark-discharge. 

 They were so near that minute waves were formed during 

 the whole time of the experiment by the wind from the points. 

 No possible loss of copper could therefore occur if any were 

 torn off from the point. The machine was run at its highest 

 speed, and thus the greatest difference of potential available 

 (from 40,000 to 50,000 volts) was obtained. The experiment 

 was continued for two hours, during the whole of which time 

 very fine luminous brushes were being produced at each point. 

 At the conclusion of the experiment the acid was carefully 

 concentrated by evaporation and tested for copper. Not a 

 trace could be found in either. Thus showing that no metallic 

 particles icere torn off the points. Moreover, I have examined 

 microscopically a pair of points which I had cut and prepared 

 from a piece of guttapercba-covered wire. These points have 

 been continuously in use for some months, and it is certain 

 that at a low computation brushes must have been drawn 

 from them for at least 150 hours. The edges seem as sharp 

 as when cut, and sensible alteration has not taken place in 

 them. 



Of course these results apply to the potential and quantity 

 worked with. Higher potential and increased quantity may 

 give quite different results. 



* Silvaims Thompson's l Elementary Lessons/ p. 305. 

 t .Toubert, Foster and Atkinson, ! Electricity/ p. 103. 



