﻿Low-Speed Positive H-Io)is in Hydrogen. 811 



atom may not be so simple, the positive ion itself being a 

 complex system. Thus the ionization potential might depend 

 on the state of excitation of the positively-charged " strik- 

 ing " atom, 



3. Experimental evidence on Ionization by Positive Ions.. 



To test the validity of these principles, we may examine 

 existing evidence on impact ionization by positive ions, 

 taking imo account to what extent the experimental condi- 

 tions approximate to the ideal conditions for the test. 



(a) High-speed Positive Ions. — ct-rays. Millikan * has 

 shown that ionization by a-rays usually results in single 

 ionization, more rarely in double ionization. Double ioniza- 

 tion, i.e. detaching the second electron' from the already 

 ionized atom, of course requires a greater amount of energy 

 than that necessary for single ionization. Assuming that in 

 the flight of an a-ray through hydrogen there is no loss of 

 energy in non-ionizing collisions, and that each pair of ions 

 produced levies a toll on the kinetic energy of the a-particle 

 of the same amount, and knowing the initial kinetic energy 

 and the total number of ions produced, we may calculate the 

 energy required to ionize a single molecule. This energy is 

 that acquired by a single charge moving through 35 volts f- 

 Since, however, much energy may be wasted in producing 

 partial ionization or in useless kinetic energy of the ejected 

 electron, we may regard this figure as a maximum value for 

 the ionization potential for a singly-charged positive ion. 



Canal or Positive Rays. — By applying a cross-field of a 

 few volts behind a perforated cathode in a discharge-tube, 

 Stark % use d a galvanometer to measure the ionization current 

 produced by the impact of the canal rays on the gas. In 

 nitrogen at a pressure of 0*134 mm. he obtained signs of 

 ionization when the cathode P.D. had a value of 500 volts. 

 These conditions are very different from the ideal conditions 

 necessary. At such high pressures the mean free path of 

 the positive ion would be very small, and therefore we do not 

 know to what extent when they produce ionization their 

 kinetic energy is comparable with that derived from the 

 total P.D. of 500 volts. 



Moreover, in both these cases of ionization by high-speed 

 positive ions it is probable that the mechanism is quite 



* Millikan, Pkys. Rev. Dec. 1921, p. 456. 



t Rutherford, ' Radioactive Substances and their Radiations; p. 159. 

 t J. Stark, Annalen der Phi/sik, 1906, p. 427. See also K. Grlimme 

 and J. Koenigsberger, Zeits.f'dr Phi/sik, 6, iv. pp. 276-297 (1921). 



3G2 



