﻿and Absorption by Resonating Gas Molecules. 709 



when both compartments were completely exhausted. This 

 disposes of the second hypothesis, and is fully in accord with 

 the absorption explanation. 



Transition from Irregular Scattering to Regular Reflexion 

 by the Molecular Reso?iators. 



I have already alluded to the selective reflexion of mono- 

 chromatic light by very dense mercury vapour, which I made 

 a study of some years ago, and the transition from diffuse 

 scattering to this regular reflexion has been made the subject 

 of experimental investigation. Regular reflexion can occur 

 only when the resonators are so closely packed together that 

 the primary wave is stopped completely by a layer the thick- 

 ness of which is of the order of magnitude of the wave- 

 length. If the resonators are so sparingly distributed in 

 space that the incident wave penetrates to any appreciable 

 depth, their radiations do not combine to form a reflected 

 wave in the usual sense of the term. Planck, I think, has 

 somewhere mentioned the circumstance that the radiations 

 from the resonators destroy the primary wave on account of 

 their lag in phase, but that those in the opposite direction, there 

 being no primary wave travelling in this direction with which 

 they interfere, give rise to a wave which constitutes selective 1 

 reflexion. If this occurred, except in the special case in which 

 the wave is stopped at the boundary, it would constitute what 

 we might term " volume reflexion " as contrasted to surface 

 reflexion, but the phenomenon has no existence, I feel sure. 

 It seemed so easy, however, to get evidence of volume re- 

 flexion if it existed, that I made one experiment, rather with 

 the idea of convincing myself of its non-existence than any- 

 thing else. A cell was made of the form shown in PI. X. 

 fig. 7 and a divergent beam passed into it through the 

 oblique quartz plate. If even a very small percentage of the 

 energy emitted by the resonators was returned towards the 

 source, ?'. <?., in the direction of the primary rays reversed, we 

 should have a converging system of rays which would be in 

 part reflected from the oblique quartz plate and come to a 

 focus on the photographic plate P. The plate, however, 

 showed only a uniform darkening due to the light irregularly 

 scattered by the resonators. 



We do, however, obtain regular reflexion when the; 

 density of the vapour is so great that the wave can penetrate 

 only to a very small depth, and I next investigated the 

 subject of the density at which the change from scattering 

 reflexion takes place. A quartz bulb ^containing 



