Manchester Memoirs, Vol. Hi. (1908), iV^. 10. 19 



perhaps to involve the corollary that sensible internal 

 vibratory disturbance in the former is far more difficult to 

 excite than in the latter. In the case of molecules of 

 easily condensable gases, between which, therefore, strong 

 mutual forces come into play, there seems to be evidence 

 that the natural thermal collisions alone can excite selective 

 radiation, as indicated by the appearance of band spectra 

 under conditions of high temperature without chemical or 

 electric action : in the case of atomic line spectra the 

 negative results of Fringsheim and other observers seem 

 consonant with what was to be expected. Line spectra 

 are of very great luminous intensity compared with 

 any natural continuous spectrum on which they can 

 be superposed, unless the temperature of the latter is 

 extremely high, and therefore the molecular collisions 

 very violent. This seems to afford sufficient reason why 

 they cannot be considered as in any kind of energy equili- 

 brium with the surrounding continuous radiations. 



An adequate interpretation of the master clue to 

 dynamical molecular structure afforded by the spectrum 

 is still lacking. The researches of Liveing and Dewar, 

 Balmer, Rydberg, Kayser and Runge, Rayleigh, Schuster, 

 and others, have led to the division of the simpler 

 line spectra into correlated series of lines, with the 

 successive vibration frequencies in each series, after 

 the first one or two, determined by simple approxi- 

 mate formulas, obviously the asymptotic forms of more 

 complex exact relations which remain to be dis- 

 covered ; but very little progress has yet been made 

 towards the dynamical interpretation of this ordered 

 system. The radiations from electrons involve their 

 accelerations, while those from ordinary material vibra- 

 tors, as, for example, in the case of sound waves, depend 

 only on velocities ; thus, as Lord Rayleigh has remarked. 



