304 ON LIGHT. 



by its pull. But in so doing the same pull will also be 

 communicated to the ball behind it, drawing it forward, 

 and so in- succession to those yet behind ; and in this 

 manner, a wave of extension will run back along the 

 series. If the tension of the string be. very violent (sup- 

 pose equal to the repulsive elasticity of the balls), this 

 wave will run back with the same velocity as the other. 

 Here we have then a case of the reflexion of a wave, 

 where, in the very instant of reflexion, its character is 

 changed ipso facto from that of a wave of compression 

 to one of extension in other words, it starts backwards 

 in the opposite phase to that of its arrival ; or, again in 

 other words, a semi-undulation is lost, or gained (for it 

 matters not which) in the act of reflexion. 



(86.) This is the extreme case of reflexion from a 

 denser medium on a rarer for here there is absolutely 

 nothing to carry on the motion beyond the terminal 

 ball. Such a case never occurs in nature as regards 

 light j since even what we call a vacuum is filled with 

 the luminiferous ether. To assimilate it to such as do 

 occur, suppose a second series of smaller balls, similarly 

 connected with each other, but not with the first set, and 

 brought end to end with it, with just room between for 

 one intermediate free ball of the smaller size to play 

 backwards and forwards as a go-between ; and let this, 

 in the first instance, be placed in contact with the last 

 ball of the first set. When the movement reaches it, it 

 will be driven off, and immediately striking the end ball 

 of the second set, will propagate along it a wave of com- 

 pression, coming itself to rest. In so doing it will carry 



