ON LIGHT. 305 



off some, but not all of the motion of the terminal ball of 

 the first set. This will still continue to advance after 

 the blow, but to a less extent, and with less momentum 

 than in the former case, and, just as in that case, will 

 propagate backward a wave, though a feebler one, of 

 extension. Starting, then, from the same place at the 

 same moment, the two waves the reflected portion (or 

 echo) and that which runs forward in the second set of 

 balls, set out each in its own direction in opposite 

 phases. 



(87.) The intensity of the reflected wave or echo will 

 be feebler the nearer the balls of the two sets approach 

 to equality (or the less the difference of density in the two 

 media). If they are exactly equal, the go-between ball 

 will carry off all the motion of the ball which strikes it 

 or there will be no reflected wave, no echo. And this 

 agrees with fact. At the common surface of two trans- 

 parent media of equal refractive power, however they 

 may differ in other respects, there is no reflexion. But 

 suppose the second set of balls, as also the single inter- 

 mediate one, larger than the first. In that case (still 

 according to the laws of elastic collision) the last ball 

 of the first set not only will not advance after the shock, 

 but will be driven back, and the wave which it will pro- 

 pagate backwards will no longer be one of extension, 

 but of compression. This being also the case with that 

 propagated onwards in the second series, in this case 

 both will start on their respective courses from the point 

 of reflexion in the same phase. 



(88.) In the undulatory theory of light the "denser* 1 



