408 Prof. Thomson, On some experiments on the [May 10, 



kinetic energy, so that the radiation of a system of molecules 

 is not altered by the application of a system of forces, provided 

 the temperature be not altered ; but the temperature of the mole- 

 cules in the electric field is only altered slightly, if at all, before 

 discharge takes place, so that the radiation from the molecules 

 can only be slightly affected by the electric field, and there 

 will only be very little if any additional loss from radiation. 



The above results however are only true when the applied 

 forces are not intense enough to produce in the intervals between 

 two or at most a small number of collisions a velocity in a co- 

 ordinate of any type, which is large compared with the average 

 value of that velocity at the temperature of the gas. 



When the forces are intense enough to do this in a coordinate 

 of any type it is obvious that the energy of this type will be 

 radiated away enormously more quickly than it would have been 

 if the same amount of energy had been acquired by the molecule 

 so slowly that it had had time to get distributed by collisions 

 among other molecules. Let us apply these considerations to the 

 electric field. 



When the intensity of the field is small, the increase in tem- 

 perature produced by it must be very small, otherwise it would 

 have been observed, hence the energy lost by radiation will be 

 very small, and the gas will insulate almost perfectly. When, 

 however, the field gets so strong that in a space comparable with 

 the free path of a molecule it is able to generate in some co- 

 ordinate a velocity which is large compared with the average 

 for that temperature ; then since the molecule will have to make 

 many collisions before this coordinate is reduced to its average 

 value, and since it is losing energy by radiation all this time, 

 there will be a considerable dissipation of energy, and therefore 

 an electrical discharge. A numerical calculation will shew that 

 enough energy to discharge the electric field might easily be lost 

 by radiation. 



In air, at the atmospheric pressure, the maximum electrostatic 

 energy which can exist without there being an electric discharge, 

 is about 500 ergs per c.c, now the energy required to raise the 

 temperature of 1 c.c. of air at the atmospheric pressure 1 degree 

 is about 15000 ergs, and the temperature of the gas would not 

 have to be raised very much for it to lose its heat at the rate of 

 1 degree per minute ; so that if in the electric field the mole- 

 cules on the average radiated -^ as much as the gas at this 

 temperature the electric energy in the field would be dissipated 

 in one minute. At lower pressures the radiation would not have 

 to be nearly so much, for until we get near to the critical pres- 

 sure, the electro-motive intensity which the dielectric can stand 

 decreases almost as fast as the pressure, and as the energy in the 



