

of Electricity through Gases. 327 



the discharge through gases we have seen that when a foreign 

 substance (water-vapour) is present the potential-difference 

 which the gas can support without discharge taking place is 

 approximately steady, but that when the gas is carefully 

 dried it can support an abnormally large potential-difference, 

 though when once the discharge has started the potential- 

 difference falls at once to its normal value. The passage of 

 the spark produces a supply of modified gas which persists for 

 some time after the discharge has stopped; during the ex- 

 istence of this gas the potential-difference required for spark- 

 ing has only its normal value, whether water-vapour be 

 present or not. If, however, the gas is allowed to rest for a 

 sufficient time for this modified gas to return to its original 

 condition, the gas can again sustain an abnormally great 

 potential-difference before a luminous discharge passes. This 

 result shows, I think, that the decrease in the potential-dif- 

 ference required to produce a luminous discharge through 

 dry hydrogen immediately after a spark has passed through 

 the gas is not due to the formation of water- vapour by the 

 combination of hydrogen with a trace of oxygen that may 

 have been present, for if it were the diminution in electric 

 strength would be permanent instead of temporary. 



The most obvious explanation of the great difficulty which 

 the fresh spark experiences in passing through a dry gas is 

 that the passage of a luminous discharge through a gas is 

 preceded by the condensation of some of its molecules into a 

 more complex state of aggregation, and that when these 

 aggregates are formed the potential-difference the gas can 

 support has its normal value; further, that the formation of 

 these aggregates requires or is very much facilitated by the 

 presence of nuclei of a foreign substance, and that when 

 these nuclei are removed or very much diminished in number 

 the gas can sustain an abnormally large potential-difference 

 without its molecules condensing into these aggregates, the 

 gas, however, being in an unstable condition, for as soon as 

 condensation takes place the potential-difference sinks to its 

 normal value. On this view the discharge through a gas 

 does not consist in tearing the atoms of a single molecule apart, 

 but rather in tearing atoms from off a complex aggregate of 

 molecules. This view explains the difficulty which I have 

 alluded to elsewhere (' Notes on Electricity and Magnetism,' 

 p. 193), that on the electrical theory of chemical combination 

 the force holding the atoms together in a single molecule is 

 enormously greater than the force tending to separate them 

 in an electric field strong enough to produce discharge. 



