Displacement of Spectral Lines produced hy Pressure. 511 



kinetic energy shows that unidirectional shifts from this 

 cause must also be negligible. 



Even if we suppose the whole atomic mass to be due to 

 negative electrons, the magnetic effects are too small to account 

 for those observed. With the electrostatic view the difficulty 

 lies the other way, so there appears to be little doubt that 

 the pressure displacement arises by electrostatic resonance, 

 although the precise nature of the vibrators leaves plenty of 

 room for further investigation. 



§ 8. Broadening of the Lines. 



It will probably have been noticed that the present theory 

 furnishes an explanation of the broadening of spectral lines 

 at high pressure. Referring to equation (8) we see that the 

 first term depends on the sign of u. The displacement due 

 to this term will therefore be sometimes positive and at others 

 negative. If the gas were symmetrically distributed about 

 the atom A, the effect of this term would be equal to its 

 average effect, that is, zero. Owing to the motion of the 

 atoms of gas, however, sometimes one side sometimes the 

 other will have the advantage, but the mean position will 

 always be the most probable. We should therefore expect 

 the light to be distributed about the mean wave-length ac- 

 cording to a law similar to the law of errors. Allowing for 

 peculiarities in the behaviour of individual lines, arising from 

 causes not taken account of in this theory, this behaviour is 

 in substantial agreement with the observed phenomena. 



A reference to § 5 will show that the method used for 

 treating the problem there considered fails when /3— y is 

 small. This may be one- cause of the well-known fact, 

 usually attributed to selective absorption, that an emitting 

 atom when surrounded by a large number of similar atoms, 

 as when the density of the metal vapour in the arc is great, 

 exhibits enormous broadening but no definite unidirectional 

 displacement. 



The importance of collisions in the immediate neighbourhood 

 of the emitting atom in giving rise to broadening as opposed 

 to displacement has already been pointed out (p. 570). 



§ 9. Conclusion. 



The object of this investigation has been to try to establish 

 the view that the displacement of the lines observed in the 

 arc spectra at high pressures is an effect of the reaction on 

 the emitting atom of forced vibrations set up in the neigh- 

 bouring atoms. The forces causing the vibrations are the 



