The Light from Mercury Vapour. 61 



of a very obliquely held screen, the fringe systems observed 

 at and near the surface of the screen on either side show features 

 which require for their explanation the complete analytical 

 investigation of diffraction given by Sommerfeld. On 

 one side of the screen we have the region of shadow, and 

 adjoining it diffraction fringes, the maxima and minima of 

 illumination in which show contrasts more marked than 

 those in the diffraction fringes of the Fresnel type, their 

 positions, however, being the same. This has been verified 

 by photometric observation. On the other side of the screen, 

 the fringes due to the interference of the direct and reflected 

 wave-trains are observed, and these are modified by dif- 

 fraction in a manner which can be fully explained only in 

 terms of the complete analytical solution of the diffraction 

 problem. 



The solution obtained by Sommerfeld with the boundary 

 condition 5 = at the surface of the mirror, agrees with the 

 results observed at oblique incidences in the part of the field 

 under discussion, irrespective of the plane of polarization of 

 the incident light. 



The investigation described in the paper was carried out in 

 the Palit Laboratory of Physics. 



Calcutta, 

 16th March, 1918. 



Y. The Light from Mercury Vapour. 

 ByC.D. Child * 



IN a recent article by Strutt f a description is given of 

 experiments on the luminous vapour coming from the 

 discharge through gases at low pressure. The conclusion 

 reached in that paper is that the light given off by the vapour 

 very largely comes from the negative ions. The conclusion 

 reached during an investigation of the luminous vapour 

 distilled from the mercury arc performed by myself was 

 that the light did not come from either the positive or the 

 negative ions alone, but from the two at the moment of 

 recombining J. 



Two questions present themselves in connexion with these 

 papers. First, is the same fundamental action being studied 

 in the two sets of experiments ? Secondly, do the experiments 

 of Strutt prove his conclusions ? Apparently neither of these 



* Communicated by the Author. 



t Proc. Roy. Soc. A, xciv. p. 88 (1917). 



+ Phil. Mag-. [6] xxvi. p. 906 (1913). 



