294 



INFRA-RED EMISSION SPECTRA. 



processes in which the radiation is more intense than that correspond- 

 ing to the sensible temperature of the tube. The luminosity of a gas is 

 due to vibrations within the molecule rather than translatory motions 

 of the molecules as a whole, which determine temperature. These 

 intermolecular vibrations result from chemical or electrical disturbances. 

 The theory of the low temperature of the vacuum-tube has been 

 worked out by Warburg.' He computes the temperature upon the as- 

 sumption that the energy, computed from the product of potential gra- 

 dient and current, is changed into heat. He also shows that for a gas 

 at a low pressure the heat conductivity is very rapid, so that in a small 

 fraction of a second the gas assumes a constant temperature. In the 

 same manner, on stopping the discharge the temperature will return 

 to its original value ; hence there will be a rise and fall in temperature 

 and a corresponding rise and fall in pressure of gas. For nitrogen he 

 com.putes the following temperature for a current of 0.0012 ampere: 



Parts of tube. 



Temperatiires. 



Temperature of inner wall=Wi.... 

 Temperature of axis of tube^Wo' 

 Mean temperature of gas, W 



o Computed for a current of 0.0032 ampere. 



This shows that while the inner wall is not above i°.5 the axis is 

 at a temperature of 21° to I32°C. 



The interior of the vacuum-tube was explored by Wood," who used 

 for the purpose a bolometer of very thin (o.ooi mm.) platinum-iridium 

 wire. The highest temperature recorded was 43° C. In all cases the 

 computed and observed temperatures were in very close agreement. 



Since the question of dissociation must be considered in this work it 

 is of interest to note that Townsend' shows that the dissociation of the 

 molecule by collision with an electron in a vacuum-tube is very small. 



All observers agree in thinking that when radiation is emitted by a 

 gas, in one case by heating it and in another case by sending a current 

 through it, the mechanism which is brought into play must differ in 

 some important respects in the two cases. 



^Warburg: Ann. der Physik (3), 54, p. 265 

 *Wood : Ann. der Physik (3), 59, 238, 1896 

 'Townsend: Phil. Mag. (6), i, 226, 1901. 



265, 1895. 



