KINETIC OK MhXHANICAL VIEW OF NATLllK. 45 



the one side we have a great volume of purely analytical 

 reasoning begun l>y Cauchy in 1- ranee, and purHued 

 under varying assumptions by Green and MacCJuUagh in 

 England, by F. Neumann and others in (Jermany. On 

 the other side we have the purely experimental work 

 beginning with WoUaston and Brewster in England, the 

 refined methods for measuring the speed of light invented 

 ]jy Fizeau and Foucault, the beautiful contrivances for 

 experimental research and verification of Jamin and 

 nianv others. Out of so many fruitful conceptions 

 which have resulted in an enormous accumulati(jn of new 

 knowledge of actual phenomena of light and wave-motion 

 — the real and sole end and aim of all theory — I will for 

 the purpose of illustration single out one which in llie 

 middle of the century opened out an entirely new field 

 of iniiuirv, formino- almost a new science by itself. I 8i. 



^ ■' ■" Spectrum 



refer to spectrum analysis. analysis. 



The phenomena of dispersion (rainbow scattering) and 

 absorption (partial or complete extinction) of light were 

 among the earliest known, and had been among the 

 longest studied, properties of bodies. I>eing, besides, 

 connected with the physiological, subjective, and artistic 

 effects of light, they have always commanded special 

 interest. And yet, so far as either the emission or 

 the undulatory theory is concerned, they have always 

 presented special ditticulties. When the wave theory 

 was first propounded, it was generally understood on 

 the analogy of the phenomenon of sound that diflerence 

 of colour depends upon difierence of frequency, or where 

 the velocity of propagation (as in vacuo or in atmospheric 

 air) is the same, on the length of the waves. Thr ditli- 



