﻿and Absorption by Resonating Gas Molecules. 699 



nearly ^ of: the intensity of the adjacent primary radiation. 

 Its intensity fell off with increasing distance as follows : — 



Distance. 



Ratio 



0*5 mm. 



1/3 



1-5 „ 



1/6 



2-5 „ 



1/10 



3-5 ,, 



1/30 



Four photographs of the phenomenon are reproduced on 

 PI. XI. fig. 4. A vacuum-tube was put in circuit with the 

 tube leading to the pump, to serve as an auxiliary mano- 

 meter, and it was found that the secondary radiation did not 

 appear in its maximum intensity until the green fluorescence 

 due to cathode rays appeared in the vacuum-tube. 



The arrangement of the apparatus is shown in PI. X. 

 fig. 2, which requires no explanation. It is clear from 

 the photographs that when the mercury vapour is in the 

 highest possible vacuum, the light which it emits is capable 

 of exciting a secondary radiation in the surrounding vapour 

 which lies wholly outside of the path of the primary ex- 

 citing beam. The presence of air at I or 5 mm. pressure, 

 while it materially decreases the intensity of the primary 

 resonance radiation, causes the secondary radiation to dis- 

 appear entirely. In the four photographs which are re- 

 produced I have recorded the pressure of the residual air 

 in each case. To make sure that the disappearance of 

 the secondary radiation was not due to a weakening of the 

 primary radiation, I gave an exposure of four times the 

 normal one, with air at 4 mm., and though the cone was 

 very much blacker on the plate than on any of the others, 

 there was no trace of any luminosity in the surrounding 

 vapour. 



This action of a small trace of air is most remarkable, and 

 it is of the utmost importance to determine the explanation. 

 Although the vapour which is in the path of the primary 

 beam glows with diminished brilliancy, the light which it 

 gives out seems powerless to excite the surrounding vapour 

 to luminosity. It seemed possible that the damping due to 

 molecular collisions rendered the emitted light less homo- 

 geneous and therefore incapable of exciting the vapour, in 

 spite of the fact that the absorption band must be regarded 

 as broadened to a similar amount. This point is one which 

 requires very careful investigation, and if the cause can be 

 found much light may be thrown upon the nature of the 

 effects of molecular collisions upon radiation. It occurred 



