ON THE NATURE OF LIGHT AND COLOURS. 46l 



by undulations, propagated in all cases through an ethereal medium. Tlie 

 mechanism of thesf supposed undulations is so complicated, and attended by 

 so many difficulties, that the few who have examined them have been in 

 general entirely dissatisfied with them: and the internal vibrations of the 

 particles of light themselves, which Boscovich has imagined, appear scarcely 

 to require a serious discussion. It may, therefore, safely be asserted, that 

 in the projectile hypothesis this separation of tlie rays of light of the same 

 kind by a partial reflection at every refracting surface, remains wholly unex- 

 plained. In the undulatory system, on the contrary, this separation follows 

 as a necessary consecjuence. It is simplest to consider the ethereal medium 

 Avhich pervades any transparent substance, together with the material atoms 

 of the substance, as constitutmg together a compound medium denser than 

 the pure ether, but not more elastic; and by comparing the contiguous 

 particles of the rarer and the denser medium with conunon elastic bodies of 

 different dimensions, we may easily determine not only in what manner, 

 but almost in what degree, this reflection must take place in different circum- 

 stances. Thus, if one of two equal bodies strikes the other, it communi- 

 cates to it its whole motion without any reflection; but a smaller body 

 striking a larger one is reflected, Avitlr the more force as the difference of 

 their magnitude is greater; and a larger body, striking a smaller one, still 

 proceeds with a diminished velocity ; the remaining motion constituting, in 

 the case of an undulation falling on a rarer medium, a part of a new scries 

 of motions which necessarily returns backwards with the appropriate velocity: 

 and we may observe a circumstance nearly similar to this last in a portion 

 of mercury spread out on a horizontal table; if a wave be excited at any 

 part, it will be reflected from the termination of the mercury almost in tlie 

 same manner as from a solid obstacle. 



The total reflection of light, falling, with a certain obliquity, on the 

 surface of a rarer medium, becomes, on both suppositions, a particular case 

 of refraction. In the undulatory system, it is convenient to suppose the 

 two mediums to be separated by a short space in which their densities ap- 

 proach by degrees to each other, in order that the undulation may be turned 

 gradually round, so as to be reflected in an equal angle: but this supposition 

 is not absolutely necessary, and tlie same effects may be expected at the 

 surface of two mediums separated by an abrupt termination. 



