SPECTRAL LINES 455 



conclusions would be very rash. It should be remembered also 

 that too good an agreement cannot be expected when one 

 recollects that some of the assumptions which underlie the 

 theory are somewhat ideal, and are certainly departed from in 

 nature. The atom has been pictured above as a collection of 

 electrons each vibrating about a position of equilibrium, and 

 each, by its vibrations, emitting radiation of a definite frequency 

 and so giving rise to a single spectral line. The limitations of 

 mathematical analysis and our lack of knowledge of the definite 

 arrangement of the electrons inside an atom compel some such 

 simple assumption, which indeed is in a sense justified by the 

 success with which it has explained many of the phenomena of 

 absorption and dispersion. Yet it is much more probable that 

 one has really to deal with the vibrations of groups of electrons, 

 which are jointly responsible by their radiation for the produc- 

 tion of a number of spectral lines ; the existence of spectral 

 series supports this view. The limitations of the above theory 

 are shown in a marked manner by one of the results obtained ; 

 it was proved that with increase of density there is, apart from 

 the displacement, a symmetrical broadening of the absorption 

 lines. In many cases such actually occurs, but in many others 

 there is a marked dissymmetry in the broadening, generally to- 

 wards the direction of longer wave length. The case of mercury 

 vapour, investigated by R. W. Wood, 1 is a very striking example. 

 Of such abnormal effects, as they may be called, the theory in 

 its present form can give no explanation. Neither can it account 

 for the anomalous behaviour of. the enhanced lines. The chief 

 experimental results have, however, been — qualitatively, at least 

 — explained by it in a remarkable manner, and therein lies the 

 justification for the belief that its fundamental assumptions con- 

 tain the germ of truth. For the present this must suffice; and 

 just as in the development of other branches of physics such 

 ideal conceptions as, for instance, those of a perfect fluid or of a 

 perfectly rigid body have been found most fruitful, so also the 

 present theory may be regarded as throwing some light upon a 

 complicated series of phenomena. A more definite discussion 

 must wait until experimental physicists have obtained a com- 

 pleter knowledge of the structure of the atom. 



1 Phil. Mag. August 1909. 



30 



