the Electric Discharge in Gases. 433 



supposes that the rays of the sun are able to dissociate the 

 compound gases, chiefly hydrocarbons, which in a very rare 

 state he supposes distributed throughout the universe; while, 

 when these gases exist at pressures comparable with that of 

 the atmosphere, they are able to transmit the sun's rays with- 

 out suffering dissociation. These considerations would seem 

 at first sight to indicate that the electric strength of gases 

 w r ould continually decrease with the density ; whereas we 

 know that it only does so to a certain point, and that after- 

 wards the electric strength increases as the density decreases. 



We have in the above reasoning, however, supposed that 

 whenever we got chemical decomposition at all we had always 

 sufficient energy absorbed to exhaust the electric field. In 

 consequence of the great absorption of energy in chemical 

 decomposition, this is legitimate, unless the gas be very rare ; 

 but for a very rare gas it will be necessary to decompose a 

 larger proportion of the molecules of the gas, and it will 

 require a more intense electric field to do this. If the gas 

 were very rare, it might be that the energy required to decom- 

 pose all the gas was not sufficient to exhaust the energy of the 

 electric field. In this case all the electricity could not be dis- 

 charged at once; while in an absolute vacuum there would be 

 no chemical decomposition to lessen the energy of the electric 

 field, and there would be no electrical discharge at all. Thus 

 there are two causes at work which produce opposite effects 

 on the electric strength as we rarefy a gas. The first is that 

 the gas is more easily dissociated as we rarefy it ; this dimi- 

 nishes the electric strength of the gas. The second is that, as 

 there are fewer molecules, a larger proportion of them must 

 be decomposed in order to exhaust the same amount of energy, 

 and it will require a more intense electric field to separate the 

 larger proportion ; this will tend to increase the electric 

 strength of a gas as we rarefy it. The second of these consi- 

 derations is not important at pressures comparable with that 

 of the atmosphere, as in this case the percentage of the gas 

 which has to be dissociated in order to exhaust the energy of 

 the electric field is extremely small; so that, starting from the 

 atmospheric pressure, we should expect the gas to get electri- 

 cally weaker as it gets rarer. With very rare gases, on the 

 other hand, the second consideration, as the extreme case of a 

 perfect vacuum shows, is the more important; and thus with 

 very rare gases the electric strength should increase as the 

 gas gets rarer. Both of these results agree with the results 

 of experience. 



It may be worth while to point out that, according to the 

 view taken in this paper, a perfect vacuum possesses infinite 

 electric strength ; and thus it is in opposition to the theories 



