Stoney — Cause of Double Lines in Spectra. 569 



CHAPTER II. 



THE PROBLEM TREATED DYNAMICALLY. 



The alternations of electro-magnetic stress in the aether which constitute light 

 form an undulation which is propagated under the same laws as the transverse 

 vibrations of a suitable medium. We shall in the present chapter treat the subject 

 under this purely dynamical hypothesis, and will in the following chapter make 

 those corrections which are required to convert the investigation into one under the 

 electro-magnetic theory of heat and light. 



We shall accordingly, for the present, regard certain points in the molecules 

 of the gas as acting dynamically on an aether capable of receiving and transmitting 

 only transverse vibrations, and we have to inquire what motions of these points 

 within the molecules would impart to the medium the oscillations which correspond 

 to the observed lines in the spectrum. 



Let us then fix our attention on a particular molecule 3f, and suppose that a 

 point P in it which acts on the sether has been set moving along some orbit within 

 the molecule by the last of the inter-molecular encounters to which Mhas been 

 subjected. We are in ignorance as to what the forces are, under the influence of 

 which the point P will continue its motion during the flight of the molecule ; but, 

 nevertheless, there is one case which admits of treatment up to a certain point ; 

 and on comparing the conclusions of this treatment with the simplest spectra — ■ 

 those of the light monad elements — we find that the conditions which lead to it 

 occur in them. We shall confine our attention in the present Memoir to this case. 

 Tt presents itself whenever one or some forces acting on P are predominant over all 

 the others, and the treatment to be employed is the same as that with which we 

 are familiar in the lunar and planetary theories. In applying this method the real 

 course of the point P is to be arrived at by first la3ang down its " dominant 

 orbit," that is the path which P would pursue if the dominant forces were the only 

 ones acting on it, and by then subjecting this orbit to perturbations while P is 

 traversing it. These perturbations are of two kinds : — (1°) such a gradual shifting 

 of the position of the dominant orbit while P is revolving round it, as will bring P 

 at each instant to the real position which it actually does then occupy under the 

 influence of all the forces ; accompanied by (2°) such a gradual change of the form 

 of the dominant orbit as may be necessary to render it at each instant the orbit 

 which P would describe if the perturbating forces were then suddenly to cease 

 acting. If the pertm'bating forces be feeble these changes will be slow as well as 



