716 Dr. S. R. Milner on the Current-Potential 



is complete a permanent current cannot be carried unless 

 there be a supply of ions at the electrodes. However, the 

 impact of the positive ions on the surface of the cathode 

 under the influence of the high initial p.d. will raise its 

 temperature very rapidly: corpuscles will thus be emitted, 

 and if the rise of temperature of the cathode surface be 

 sufficient, the current for the rest of the oscillation will be 

 carried through the now highly ionized air with a com- 

 paratively low p.d. 



The arc p.d. is partly accounted for by a drop at the 

 cathode sufficient to keep up the temperature to the right 

 value, and partly by the field throughout the gap necessary 

 to make the ions carry the current. There is also probably a 

 potential drop at the instantaneous anode, and the conditions 

 during the main part of each oscillation are similar to those 

 which occur in the arc discharge, but there is an important 

 difference as regards the temperature of the anode. In the arc 

 the anode is hotter than the cathode, a fact which is accounted 

 for by the greater potential drop at the anode and the con- 

 sequent greater development of heat there; but in the spark 

 there is reason to believe that the anode remains comparatively 

 cool. Even if there is in the spark a greater potential drop 

 at the instantaneous anode than at the cathode, it will not 

 produce the same rate of rise of the temperature of the 

 surface layer ; for there is the difference between the two 

 electrodes that the temperature of the cathode is raised by 

 the bombardment of positive ions, that of the anode by that of 

 corpuscles. Now the positive ions have a large mass, and 

 therefore a small penetrating power, consequently impinging 

 on the surface of the cathode they will expend their whole 

 energy in raising the temperature of a layer only a few 

 molecules thick; while the corpuscles with their larger 

 penetrating power will distribute their energy over a thicker 

 layer of the anode than this. Thus although the total energy 

 which they convey to the anode is possibly in excess of that 

 given up to the cathode, the rate of rise of temperature of 

 the surface layer of the anode will be less. If sufficient 

 time were given for an approximate thermal equilibrium to 

 establish itself throughout the whole mass of each electrode, 

 the temperature of the anode might possibly finally become 

 the greater, as indeed it does in the arc; but in the spark the 

 oscillations of the discharge current are extremely rapid, and 

 long before thermal equilibrium can be established the oscil- 

 lation is over. The comparatively low temperature of the 

 instantaneous anode in the spark forms, I think, a charac- 

 teristic difference between it and the arc discharge. 



