﻿and Absorption by Resonating Gas Molecules. 697 



5 mm. effected a reduction of 1/2. This is in agreement 

 with a calculation made by Schuster in his paper on 

 " Radiation through a Foggy Atmosphere " (Astrophysical 

 Journ. vol. xxi. p. 6, 1905), though the differences in this case 

 are not much larger than the probable errors in the measure- 

 ments. The beam of light which entered the cell was made 

 accurately parallel by means of a quartz lens and passed 

 through a square aperture (measuring 5 mm. on a side) 

 perforated in a black card. In this way a beam of uniform 

 cross-section was obtained, which was of course necessary if 

 the measurements were to be of any value. Only the ultra- 

 violet 2536 light entered the cell, a quartz spectrograph 

 being used as a monochromator. 



We are now in a position to consider the amount of energy 

 diverted from the primary beam by each molecule. 



Lamb, in his theoretical treatment of the absorption of 

 light by a gas, published in the Stokes Commemoration of 

 the Camb. Phil. Soc, sums up a calculation in the following 

 words : — " Hence in the case of exact synchronism, each 

 molecule ot gas would, if it acted independently, divert per 

 unit of time nearly half as much energy as in the primary 

 waves crosses a square whose side is equal to the wave- 

 length." This means, if I am not mistaken, that if we had 

 a density such that there was one molecule in each cube the 

 sides of which were equal to the wave-length, the intensity 

 of the light would be reduced by 1/2 by traversing a single 

 layer of molecules, while a density ten or twenty times as 

 great as this ought to give selective reflexion, since the wave 

 would be practically stopped before penetrating to a depth of 

 more than a small fraction of a wave-length. 



Let us now compare this calculation with the values 

 which have been determined. At a pressure of 0*001 mm., 

 which is about the pressure used, the average molecular 

 distance is such that we shall have on the average one 

 molecule of mercury in every cube the sides of which are 

 only very little larger than the wave-length (or more exactly 

 0*0003 mm.), which quantity divided into 5 mms., the distance 

 traversed for a reduction of intensity equal to 1/2, gives us 

 16,000, that is to say 16,000 molecules must be passed 

 before one half of the energy is removed from a square 

 element on the wave-front measuring \ on each side. 



Of course this calculation is made on the assumption that 

 all of the molecules are equally effective in scattering the 

 light. It is however possible, even probable, that but a 

 small percentage are, at any given moment, in the condition 



