the Motion of Electrons in Gases. 891 



The large losses of energy could only occur when the 

 electron collides with velocities exceeding those corre- 

 sponding to JiUj where n is the frequency of radiation which 

 is absorbed by the gas. In hydrogen and nitrogen there is 

 no absorption of radiant energy in the visible or infra-red 

 spectrum. In the ultra-violet spectrum nitrogen is slightly 

 opaque to rays of which the values of Tin are between 6*4 

 and 9*4 volts, but it is uncertain whether hydrogen absorbs 

 rays in this part of the spectrum. 



The velocities of the electrons are distributed about the 

 mean velocity u as in the case of molecules, but the law of 

 distribution is not the same in the two cases. If the Max- 

 wellian distribution be assumed in the case of the electrons 

 •an upper limit may be found to the number of collisions in 

 which the velocity exceeds the mean velocity by a large 

 factor. It thus appears that with a mean velocity of about 

 2 volts (& = 54) there may be a sufficient number of collisions 

 with a velocity of 6*4 volts to account for the average loss of 

 energy of electrons in colliding with molecules of nitrogen. 

 But the loss of energy is not explained on this theory when 

 the electrons are moving with smaller mean velocities corre- 

 sponding to values of k of order 10, since the number of 

 possible collisions that occur with velocities exceeding 6'4 

 volts is so small in these cases. 



When a large proportion of the energy of an electron is 

 lost in each collision the quantity k is comparatively small 

 and W is comparatively large provided no ions are formed. 

 In order that an ion may be formed the electrons must not 

 rebound from the molecule after a collision, and the tendency 

 to form ions is greatest when the molecules absorb a large 

 proportion of the energy of the electrons. Also it is found 

 experimentally that ions are only formed (even in the presence 

 of water-vapour) when the electrons are moving with com- 

 paratively small velocities, so that in order to account for 

 the formation of ions it is necessary to suppose that the mole- 

 cules can absorb small quantities of energy. Hence the 

 quantum theory indicates that large numbers of ions may be 

 formed when the gas contains impurities which absorb radiant 

 heat. The experiments are in agreement with the theory on 

 this point as it has been found that ions are not formed in 

 hydrogen and nitrogen, which are very diathermanous, but 

 are rapidly formed in gases containing water- vapour or 

 carbon dioxide which are very opaque to thermal radiation. 



In conclusion we desire to express our thanks to Mr. B. 

 Lambert for having supplied us with a quantity o( carefully 

 prepared nitrogen for use in these experiments. 



3N2 



