[ W5 ] 



XVII. The Absorption Spectrum of Potassium Vapour. 

 By P. V. Bevan - , M.A., Royal Holloway College *. 



[Plate II.] 



A CERTAIN amount of work f has been done on the 

 absorption of light by potassium vapour, but only in 

 the visible region of the spectrum. The absorption in the 

 ultra-violet as well as in the visible part is specially interest- 

 ing, firstly for its similarity to the corresponding absorption 

 by sodium vapour and secondly in relation to the analysis of 

 lines into series by Kayser and Plunge, and by Rydberg. 

 The recent work by Wood % has shown how the principal 

 series lines for sodium appear in the absorption spectrum in 

 much greater numbers than in any kind of emission spectrum 

 that has as yet been studied. A similar effect takes place 

 with potassium, and the author of the present paper has 

 measured the wave-lengths of lines of the principal series up 

 to the line corresponding to n = 2C) in the Kayser and Runge 

 formula. The method of experiment was the very simple one 

 of passing light through a steel tube containing potassium, the 

 •ends of the tube being closed by quartz plates, the emergent 

 light was then examined with a quartz spectrograph. The 

 tube could be heated and so a certain portion of it filled with 

 potassium vapour of density varying with the temperature. 

 Potassium under these conditions behaves in a very similar 

 manner to sodium, and there is a field for work of a similar 

 character to that of Wood on the optical properties of sodium 

 vapour. The spectrograph used was by Hilger giving a spec- 

 trum of about 8 inches long from w.l. 2100 to the red end of 

 the visible spectrum. For a source of light a Nernst lamp and 

 an arc lamp were used. The Nernst lamp has advantages in 

 giving a continuous spectrum, but when measurements of 

 wave-lengths are to be made the arc lamp is most suitable, as 

 in addition to o-ivino- more intense ultra-violet light it also 

 provides a set of standard Hues from which wave-lengths may 

 be calculated. 



First of all with regard to the visible spectrum. When 

 the tube is heated absorption is first noticed at the pair of 

 lines w.l/s 7699, 7655, and this is soon followed by channelled 



* Communicated by the Physical Society read November 12, 1909. 

 | lloscoe and Schuster, Proc. Roy. Soc. xxii. p. 302 ; Liveing aud 

 Dewar, Proc. Roy. Soc. xxvii. p. 132, etc., & xxviii. p. 352, etc. 

 + Phil. Mag., Oct. 1909. 



02 



