THE DISPLACEMENT OF SPECTRAL 

 LINES BY PRESSURE 



By H. SPENCER JONES, B.A., B.Sc. 



Late Isaac Newton Student in the University of Cambridge ; Chief Assistant, Royal 



Observatory, Greenwich 



Up to the year 1896, the Fraunhofer lines in the solar spectrum 

 had been regarded as fixed marks of reference, subject to no 

 possible change in position. In that year, L. E. Jewell, 1 when 

 engaged in carrying out some measurements of their positions 

 for Rowland's " New Table of Standard Wave Lengths," dis- 

 covered certain systematic differences between the wave lengths 

 of the metallic lines in the solar spectrum and of the correspond- 

 ing lines in the arc and spark spectra obtained experimentally : 

 the differences of wave length were found to vary from line to 

 line, proving that the displacements were not due to the Doppler 

 effect, arising from a motion in the line of sight. He suggested, 

 as a possible explanation, that the wave length of a line might 

 depend upon the physical conditions under which it was pro- 

 duced, or, in other words, that the vibration period of an atom 

 might depend to some extent upon its environment, and that 

 presumably an increase of density or of pressure would produce 

 a damping effect upon the vibrating and radiating systems. 



Following upon this suggestion, Humphreys and Mohler 2 

 investigated experimentally the effect of pressure upon the 

 positions of the lines in metallic spectra, by placing the arc in 

 a vessel containing air, with an arrangement by which the 

 pressure could be varied by known amounts. This proved but 

 the commencing point of a long series of experiments by 

 Humphreys, 3 Hale and Kent, 4 Anderson, 5 Duffield, 6 Rossi, 7 Gale 



1 Astroph. Journ. 3, p. 92, 1896. 

 * Ibid. 3, p. 114, 1896. 



3 Ibid. 4, p. 249, 1896; 6, p. 169, 1897 ; 22, p. 217, 1905 ; 26, p. 18, 1907. 



4 Ibid. 17, p. 154, 1903. 

 4 Ibid.. 24, p. 221, 1906. 



6 Ibid. 26, p. 375, 1907 ; Phil. Trans. A. 208, p. in, 1908. 



7 Proc. P.S., A. 83, p. 414, 1910; Phil. Mag.^ 21, p. 499, 191 1. 



43S 



