532 



Mr. A. Schuster. 



heated only during as short a time as possihle, while the other was 

 kept glowing for several hours. The former, while hot, became electro- 

 negative towards the gas in -contact, while the latter became electro- 

 positive. It seems probable that the difference is due to gases dissolved 

 in the platinum and given up by long-continued heating, and we are 

 led to the opinion that in the normal state of platinum, when it is 

 freed from its dissolved gases, it becomes electro-negative on heating. 

 The authors do not state whether the unilateral property of hot 

 platinum altered together with the sign of its electrification, though 

 their own experiments would lead one to think that this was the case. 



It is unnecessary to enter into the question how the dissolved gases 

 -can affect the result. We observe very similar phenomena with 

 platinum electrodes in liquids. The great differences observed, 

 owing to comparatively small changes in the state of the platinum, 

 incline me to a chemical, rather than to a physical, explanation. An 

 important experiment bearing on the question is worth quoting. 

 Berliner* showed that the disintegration of platinum electrodes, so 

 commonly observed in vacuum tubes, altogether ceases when the metal 

 has been deprived of its dissolved gases. While disintegration is 

 taking place the surface is constantly renewed, and any surface 

 action between the platinum and the surrounding gases could be 

 kept up continuously, but would stop as soon as the disintegration 

 itself has stopped. 



Similar effects to those described by Elster and Geitel had been 

 previously observed by Goldstein and Warburg. After Hittorff 

 had discovered that the fall of potential at the negative electrode, 

 which is necessary to produce a continuous discharge, disappears if a 

 glowing platinum wire is taken as kathode, Goldstein found that the 

 platinum wire permanently loses the power of acting if it is kept red- 

 hot for a sufficient time. Warburg J states that, on hardening the 

 wire by drawing it may be brought back to its original condition, so 

 that on heating it will again discharge negative electricity freely. 

 But I learn from private information which Professor Warburg has 

 been kind enough to supply me that the surface of his wire might 

 have become coated with greasy matter in the act of hardening, and, 

 if that is so, we are at liberty to connect the free discharge of negative 

 electricity with a chemical action at the electrode. 



Behaviour of Carbon Electrodes. — It is clear that the behaviour of 

 carbon electrodes cannot give us any certain information on the 

 question as to how far unipolar effects depend on chemical Action ; 

 but, as some phenomena shown by glow lamps have received the 

 attention of electricians, it may be useful to point out that, as far as 



* ' Wiedemann, Annalen,' vol. 33, p. 289 (1888). 



f 1 Wiedemann, Annalen,' vol. 21, 1884. 



t ' Wiedemann, Annalen,' vol. 81, p. 592 (1887). 



