506 Prof. J. S. Town send on the Collisions of 



of a in the exponential term. Admitting that the ionization 

 due to positive ions produces appreciable effects at the larger 

 distances, the currents at the distances '8, I/O, and 1*1 centi- 

 metre between the plates cannot be represented by a formula 

 of this type. 



Thus in the notation adopted by Pedersen the only 

 quantity which can be determined directly from these expe- 

 riments is the product 7a, and by means of a theoretical 

 investigation the ionizing potential in volts is found to be 16. 

 But it appears from other experiments that no reliable 

 estimate of the velocity of an electron required to ionize a 

 molecule can be obtained from the value of ya by a calcu- 

 lation founded on the supposition that a large proportion of 

 the kinetic energy of an electron is lost at each collision. 

 I have already pointed this out in various publications, but 

 it maybe of interest to refer to a series of experiments which 

 show how the velocities of electrons nre distributed when 

 moving under an electric force in a gas, and afford a more 

 reliable means of estimating the velocities required to generate 

 ions by collisions. 



2. In the first* of this series of investigations of the 

 motion of electrons in gases, experiments were made on 

 the diffusion of electrons moving in air under forces much 

 smaller than those required to generate ions by collisions. 

 It was found that in an electric field the electrons moved 

 freely and acquired a mean velocity of agitation which is 

 much larger than the normal value corresponding to thermal 

 equilibrium with the molecules of the gas. This shows that 

 when an electron collides with a molecule of air it retains to 

 a remarkable extent a large proportion of the kinetic energy 

 acquired under the electric force. A similar result was 

 obtained in oxygen, hydrogen, and carbonic acid. For a 

 given force Z, the maximum pressure of the gas at which 

 the electrons move freely, varies in different gases; this 

 pressure is greater in hydrogen than in air, and in carbonic 

 acid it is less than in air. Subsequently Franck and Hertz f 

 found by another method that electrons after colliding with 

 molecules of helium or hydrogen retained a large proportion 

 of the energy acquired under an electric force, and the 

 collisions were described as being of an elastic type. Re- 

 ference has been made so frequently to this result that the 

 phenomenon is generally supposed to be peculiar to mon- 

 atomic gases or l^drogen. 



* J. S. Townsend, Proc. Rov. Soc. lxxxi. p. 464 (]908). , 

 t J. Franck & G. Hertz, Verh. d. I). Phys. Ges. xv. 9. p. 373, May 

 1913. 



