Discharge in Rarefied Gases. 243 



At a pressure of £ inillim. the largest surface obtainable 

 by interposition of sparks had a diameter of 22 millims. 

 (measured as before, on the circumference of the tube). 



At a pressure of ^ millim. a phosphorescent surface (whose 

 diameter amounted to 12 millims.) was obtained even with- 

 out sparks ; by interposition of sparks the diameter was 

 increased up to 19 millims. 



Pressure ^g millim. Without sparks the diameter varied 

 between 9 and 11 J millims. ; with sparks it rose to 

 14 millims. 



Pressure ^ millim, Without sparks, diameter 7-8 mil- 

 lims. ; with sparks, 10 millims. 



Pressure -^ millim. Without sparks, diameter 7 millims ; 

 with sparks, not perceptibly greater. 



The size of the focal surface therefore increases in this 

 case, as it does generally, when a spark is introduced and 

 when it is lengthened. As the density of the gas decreases the 

 magnitude of the surface obtainable with metallic circuits 

 decreases; and at the same time the magnitude of the maxi- 

 mum diminishes, to which the diameter of the surface can be 

 brought by interpolation of sparks, and more rapidly than the 

 magnitude of the surfaces obtained with metallic circuits. 

 The amplitude within which the magnitude of the surfaces 

 varies becomes thus continually smaller as the evacuation is 

 continued, until at very small densities the diameter of the 

 surface remains constant and is not affected by the interpola- 

 tion of sparks. It might be supposed that the change in the 

 magnitude of the surface occupied by the ends of the rays 

 emitted by the mirror is only an apparent one ; that the sur- 

 face itself is actually of the same size, but that with different 

 degrees of exhaustion and intensities of discharge the brilliancy 

 of the phosphorescence excited is different ; if now the inten- 

 sity of the surface is not uniform, but decreases from the 

 centre, zones of the surface nearer the outside might become 

 visible when the intensity of the phosphorescence was greater 

 than when the intensity w T as less, in which case only the 

 inner portions were bright enough to be seen. The increase 

 and diminution of the focal surface thus reduces itself to an 

 increase or diminution of intensity. 



With reference to such objections, it must be noticed that 

 the brilliancy of the surface decreases only very slightly 

 from the centre to the periphery, but, with each magnitude 

 of the surface, falls away very rapidly at the outside, so that 

 the boundary of the surface is sharp and clearly marked. 

 Further experience shows that a decrease of the surface 

 takes place when the density of the gas is reduced, whereas, 



