PREFACE TO THE FIFTH EDITION, xiii 



It appears to me that heat and light may be considered 

 as affections ; or, according to the Undulatory theory, vibra- 

 tions of matter itself, and not of a distinct ethereal fluid 

 permeating it. These vibrations would be propagated, just 

 as sound is propagated by vibrations of wood or as waves by 

 water. To my mind, all the consequences of the Undulatory 

 theory flow as easily from this as from the hypothesis of a. 

 specific ether ; to suppose which namely, to suppose a fluid 

 sui generis, and of extreme tenuity, penetrating solid bodies 

 we must assume, first, the existence of the fluid itself; 

 secondly, that bodies are, without exception, porous ; thirdly, 

 that these pores communicate ; fourthly, that matter is 

 limited in expansibility. None of these difficulties apply to 

 the modification of this theory which I venture to propose ; 

 and no other difficulty applies to it which does not equally 

 apply to the received hypothesis. With regard to the plane- 

 tary spaces, the diminishing periods of comets is a strong 

 argument for the existence of an universally-diffused matter: 

 this has the function of resistance, and there appears to be no 

 reason to divest it of the functions common to all matter, or 

 specifically to appropriate it to certain affections. Again, the 

 phenomena of transparency and opacity are, to my mind, 

 more easily explicable by the former than by the latter 

 theory, as resulting from a difference in the molecular 

 arrangement of the matter affected. In regard to the effects 

 of double-refraction and polarisation, the molecular structure 

 gives at once a reason for the effects upon the one theory, 

 while upon the other we must, in addition to previous 

 assumptions, farther assume a different elasticity of the ether 

 in different directions within the doubly-refracting medium. 

 The same theory is applicable to Electricity and Magnetism ; 

 my own experiments on the influence of the elastic inter- 

 medium on the voltaic-arc, and those of Faraday on electrical 

 induction, furnish strong arguments in support of it. My 



