r =r>>--§r 



1034 Superficial Resonance of Sodium Vapour. 



intensity removed from the line by the resonating molecules, 



2C f V*x! _ ., 



VhS e%du > 



V »- %, 



or for the ratio 



r_ 2 rvL, _ w2 



7" — — >■= <? " dw. 



1 VttJo 

 I' 

 Now the value of y has been determined by the photo- 



1 

 metric experiments. For the concentration ^tttq? or the 



j/ -U-i" 

 flame containing the least amount of sodium y =0'25. 



From this value and the tables of integrals (Calculus of 

 Probabilities of M. Bachlier) we can calculate the value of 

 the upper limit 



y/lx 1 = 0-0225; ...... (4) 



and by division (equations 2 and 4) 



f=0"27 (5) 



"We may obtain an approximate value of X\ if we take for 

 € the value given for a feeble sodium flame by Fabry and 

 Buisson, 



e = 0-04 A. 



Inserting this value in equation (5) gives us 



^ = 0-01081, 



or, since .i\ is the half width of the region required, for the 

 width of the region effective in exciting resonance, 



2x 1 = '021 A, 

 the probable width of the resonance lines in contrast to 



2e = '08 A, 



the width of the flame lines. 



We thus see that by means of sodium vapour at low 

 temperature we can manufacture, so to speak, light much 

 more homogeneous than the incident light, the method being 

 somewhat analogous to that of the Residual Rays of Rubens 

 and Nichols. 



It is highly probable that the width of the region removed 

 from the exciting line is identical with the width of the 

 re-emitted resonance radiation. The lines obtained in this 



