CONDUCTION THROUGH GASES 129 



to a state in which all the ions are removed to 

 the electrodes as fast as they are produced by the 

 ionizing agency. 



As the sparking point is approached, the curve 

 shows that the current again rises rapidly ; the 

 applied electric force being strong enough to 

 produce ions in the gas by its own action. 

 Townsend has shown that this process is effected 

 by the collision with the gas molecules of ions 

 already present, which are driven forward by 

 the electric force with high velocity. In this 

 way are formed most of the ions which carry 

 the current in an electric spark, or in the arc 

 discharge. 



We have described already the methods of 

 calculating the velocities with which the ions of 

 liquids move under known electric forces, and of 

 determining those velocities by direct experiment. 

 For gaseous ions, the corresponding velocities 

 are much higher. They have been determined 

 in several indirect ways, with concordant results. 

 For instance, Zeleny measured the electric force 

 required to push an ion against a stream of gas, 

 moving with a known and uniform velocity in the 

 opposite direction to the natural motion of the 

 ion. Langevin, in 1902, attacked the problem in 

 another way. The gas between two parallel elec- 

 trodes was exposed momentarily to the action of 

 Rontgen rays. The ions thus produced may dis- 

 appear in two ways. Opposite ions may recombine 

 with each other, or they may pass to the electrodes 

 under the influence of an electric force. If the 

 force be great, the latter method alone is operative, 

 the number of ions recombining before reaching 

 the electrodes being very small. If, then, the 



K 



