484 ON THE ABSORPTION OF LIGHT 



shall (like AB) bisect the angle between the branches. 

 The branches, however, are of unequal length, the one 

 BCD being longer than the other, by a quantity equal 

 to half the length of the undulation or pulse of the musi- 

 cal note in question. It is evident, then, that if that 

 note be sounded at A, each pulse will subdivide itself at 

 B b, and the divided portions will run on along the two 

 branches with equal intensities till they reunite at D d. 

 They will arrive there, however, in opposite phases, and 

 will therefore counteract each other at their point of re- 

 union, and in every point of their subsequent course 

 along the pipe D E; so that on applying the ear at E 

 no sound should be heard, or at best a very feeble one, 

 arising from some slight inequality in the intensities 

 wherewith the undulations arrive by the longer and 

 shorter pipe a difference which may be made to dis- 

 appear, by giving the longer a trifle larger area for its 

 section.* 



(10.) Suppose now that the pipes instead of being 

 cylindrical were square, and that the whole surface of 

 one side of a chamber were occupied with the orifices 

 A of such pipes, leaving only such intervals as might be 

 necessary to give room for their due support, and for 



* I ought to observe, that I have not made the experiment de- 

 scribed in the text, nor am I aware that it has ever been made ; but 

 it is easy to see that it ought to succeed, and would furnish an apt 

 enough illustration of the principle of interference. Instead of a 

 pipe, inclosing air, a canal of water might be used, in which waves 

 of a certain breadth, excited by some mechanical contrivance at one 

 end, would not be propagated beyond the point of reunion, D, of 

 the twa canals into which the main channel, A B, was divided. 



