994 Sir J. J. Thomson on a 



the luminous parts of the striations. This was very con- 

 spicuous in some experiments I made on the distribution of 

 the electric force in the striated discharge (Phil. Mag. xviii. 

 p. 441, 1910). The bright parts of the striations are on the 

 cathode side of the maxima of the energy of the electrons. 

 The distribution of the energy of the electrons can be deter- 

 mined by measuring the temperature attained by a thin 

 platinum wire at various places along the positive column. 

 This has been done by Professor R. W. Wood ( Wied. Ann. 

 xxxix. p. 238, 1896J, and the curves he gives show that the 

 places of maximum temperature are quite distinctly on the 

 anode side of those of maximum luminosity. 



The thickness of the bright part of the striations need not 

 be the same as the thickness of the spaces over which there 

 is ionization, though the space between the bright parts of 

 two striations will be equal to the distance between two 

 maxima on the curve representing the distribution of the 

 energy of the electrons. The breadth of the luminous por- 

 tions will depend upon the way the positive ions combine 

 with the electrons. Thus, to take an extreme case, if a 

 positive ion did not combine at all with an electron whose 

 energy was greater than E\ but combined at every collision 

 with electrons whose energy was less than this, the light 

 given out by the positive ions would be condensed into a 

 thin layer on the cathode side of the places where the energy 

 of the electrons was equal to E ; . 



In a mixture of gases the places where the ionization is 

 greatest will be the same for each component of the mix- 

 ture ; and if the positive ions were immovable, the places 

 where the spectra of the different constituents are brightest 

 would coincide. In consequence, however, of the drift of 

 the positive ions under the electric field, the places where 

 the light is emitted will be nearer the cathode than those 

 where the ions are produced, and the displacement will vary 

 with the character of the ions. If the bright parts of the 

 striations are thin, the difference in the displacement may be 

 sufficient to make the bright part due to one gas clear of 

 those due to the other gases, making the bright parts of the 

 striated column consist of bright layers corresponding to 

 different gases in the mixture. 



If the gas is one whose atoms or molecules can be excited 

 to luminosity without the detachment of electrons, then some 

 of these will be luminous though they have not previously 

 been positively charged ; they will not move under the 

 electric force, and the light they emit will come from all 

 those parts of the tube which correspond to points on the 



