required to produce Discharge through various Gases. 447 



of potential in air, as has been pointed out by Professor J. J. Thom- 

 son.* The following explanation may be offered of the fact that there 

 is this minimum striking potential, and that it is approximately con- 

 stant. 



The negative glow in any gas, as has been shown by Warburg,! re- 

 quires for its production a definite difference of potential (about 340 

 volts in the case of air) independent of the pressure and constant, so 

 long as the glow is not crushed into a smaller space than that which 

 it would naturally occupy. If the glow is crushed, the necessary poten- 

 tial is greater. 



Let us now suppose that the discharge takes place between two 

 parallel plates. A- part of the space between these plates is occupied 

 by the negative glow, a part by the positive column. So long as any 

 of the positive column remains, it is clear that the negative glow is not 

 constricted, and consequently it only requires 340 volts to produce it. 

 The greater the length of the positive column, the greater the corre- 

 sponding potential difference, so that the striking potential will be the 

 least possible when the pressure is low enough to make the negative 

 glow occupy the whole space between the plates, but not low enough 

 to make it require more. 



My experiments have been undertaken with a view to obtaining 

 further experimental evidence on these ideas. The sparks were taken 

 between large metal plates, f mm. apart. For details of the apparatus 

 and method of experimenting, the original paper must be consulted. 

 There also will be found curves showing the relation between spark 

 potential and pressure for the following gases : atmospheric air, hydro- 

 gen, nitrogen, helium. 



I give here only the minimum value of the spark potential found 

 for each, together with the cathode fall of potential given by 

 Warburg : — 



Nature of gas. 



Cathode fall. 



Minimum spark potential. 



Nitrogen specially freed 

 from all traces of oxygen 



TOlts. 



340—350 

 300 



Variable, 315—340 

 230 



226 



volts. 

 341 

 302, 308 

 347. 351, 369, 388 

 251 



Variable, 326—261 



It will be seen that on the whole, the evidence is in favour of the 

 views explained above. 



* ' Recent Researches in Electricity and Magnetism,' p. 158. 

 f ' Wied. Ann.,' vol. 31, p. 579. 

 VOL. LXV. L 



