1891.] Luminous Discharge of Electricity through a Gas. 93 



so that the ratio of the kinetic energy of the atoms to the energy in 

 the electric field is 



10*.F 



Now, at atmospheric pressure, F for air is about 3Xl0 12 ; we have 

 seen that v = 1'6 x 10 10 ; hence, in this case, the kinetic energy of the 

 atoms is about 8000 times that of the electric field. If we had taken 

 the case of a gas at a lower pressure, the disproportion would have 

 been still greater. 



For this reason, the discharge along the positive column cannot be 

 carried by atoms travelling at the same rate as the discharge ; the 

 same argument would also be fatal to the view that the discharge 

 takes place by the consecutive interchange of atoms between adjacent 

 molecules. If, therefore, we are to retain the view (which seems to 

 me to be almost established by the results of recent experiments) 

 that the passage of electricity is effected by the dissociation of the 

 molecules in the path of the discharge, we are precluded from sup- 

 posing that in the positive column the discharge takes place by the 

 molecules dissociating one after another, as the discharge comes up 

 to them. In a paper in the 'Philosophical Magazine' for August, 

 1890, I suggested that we could reconcile the dissociation theory 

 with the observed velocity of propagation of the discharge (of which 

 I had, at that time, only obtained an inferior limit, and did not know 

 that it started from the anode), by supposing that the molecules of 

 the dielectric in the path of the discharge, before the discharge takes 

 place, form themselves into a series of Grotthus chains, and that for 

 the molecules which constitute any one of these chains, the dissocia- 

 tion and recombination go on simultaneously. This may, perhaps, be 

 made clearer by a somewhat crude illustration. 



A. B. C. D. 



If A, B, C, D represent consecutive polarised molecules, the 

 simplest view of the discharge would be to suppose A to split up into 

 atoms, its positive atom going up to B and combining with the 

 negative atom of that molecule ; the positive atom of B is driven off, 

 and travels to C and combines with the negative atom, and so on. On 

 this view the velocity of the atoms would be very nearly that of the 

 discharge which other preceding experiments have shown to be 

 inadmissible for the positive column. If, however, we suppose that 

 the molecules A, B, C, D, constituting a Grotthus chain, are split up 

 simultaneously, and that while the positive atom of A combines with 

 the negative of B, the positive atom of B is combining with the 

 negative of C, and so on, then, in the time which elapses between 



H 2 



