578 Prof. E. H. Barton on the Lateral 



electrostatic forces called into play by the displacement of 

 the electrons. 



A calculation o£ the effect to be expected on this theory is 

 given, and the displacement is found to be in the direction 

 of greater wave-length and to be proportional to the pressure 

 of the gas. Both of these results are in agreement with 

 experiment. The magnitude of the effect to be expected is 

 considerably greater than that which has been found experi- 

 mentally. This is probably due to the complexity of atomic 

 structure, as the theory is exact only for the simplest possible 

 type of atom. 



Effects due to the motion of the atoms are shown to be of 

 little importance in calculating the displacement of the lines. 

 The explanation of the discrepancy between the theoretical 

 and observed effects is probably to be sought in the shielding 

 effect of the other electrons in the atom. A similar problem 

 arises in the electron theory of dispersion, where it appears 

 to have been disregarded hitherto. 



The possibility of perturbations of magnetic origin giving 

 rise to displacements is considered at some length. The 

 magnetic effects are always found to be small compared with 

 those of electrostatic origin. 



The application of this point of view to the case of the 

 absorption spectra of dissolved substances and solids may 

 conveniently be left to a later paper. 



LVII. Tlxe Lateral Vibration of Bars treated simply. By 

 E. H. Barton, D.Sc, F.R.S.E., Professor of Expert- 

 mental Physics, University College, Nottingham *. 



[Plate XIV.] 



IN his treatment of the lateral vibration of barsf Lord 

 B;ayleigh obtains expressions for the potential energy 

 of bending and the kinetic energy of the elements of the 

 bar due both to translation and to rotation. The energy of 

 rotation, which is in most cases comparatively small, is then 

 omitted to simplify the analysis and a full solution obtained 

 giving the periods of the vibrations, the position of the nodes, 

 and the forms of the curves assumed by the bars when 

 vibrating. But, as the method here followed involves the 

 use of the calculus of variations, it seems desirable, for 

 teaching purposes, to have if possible an equally close 

 solution, but obtained by the simpler analysis familiar to 



* Communicated by the Author. 



f ' Theory of Sound/ vol. i. pp. 255-296. 



