On the assumption that light was wave-motion, all his experiments on interference were accounted for; on the assumption that light was flying particles, nothing was explained.  In the time of Huyghens and Euler a medium had been assumed for the transmission of the waves of light; but Newton raised the objection that, if light consisted of the waves of such a medium, shadows could not exist.  The waves, he contended, would bend round opaque bodies and produce the motion of light behind them, as sound turns a corner, or as waves of water wash round a rock.  It was proved that the bending round referred to by Newton actually occurs, but that the inflected waves abolish each other by their mutual interference.  Young also discerned a fundamental difference between the waves of light and those of sound.  Could you see the air through which sound-waves are passing, you would observe every individual particle of air oscillating to and fro, in the direction of propagation.  Could you see the luminiferous ether, you would also find every individual particle making a small excursion to and fro; but here the motion, like that assigned to the water-particles above referred to, would be across the line of propagation.  The vibrations of the air are longitudinal, those of the ether transversal.

The most familiar illustration of the interference of sound-waves is furnished by the beats produced by two musical sounds slightly out of unison.  When two tuning-forks in perfect unison are agitated together the two sounds flow without roughness, as if they were but one.  But, by attaching with wax to one of the forks a little weight, we cause it to vibrate more slowly than its neighbour.  Suppose that one of them performs 101 vibrations in the time required by the other to perform 100, and suppose that at starting the condensations and rarefactions of both forks coincide.  At the 101st vibration of the quicker fork they will again coincide, that fork at this point having gained one whole vibration, or one whole wavelength, upon the other.  But a little reflection will make it clear that, at the 50th vibration, the two forks condensation where the other tends to produce a rarefaction; by the united action of the two forks, therefore, the sound is quenched, and we have a pause of silence.  This occurs where one fork has gained half a wavelength upon the other.  At the 101st vibration, as already stated, we have coincidence, and, therefore, augmented sound; at the 150th vibration we have again a quenching of the sound.  Here the one fork is three half-waves in advance of the other.  In general terms, the waves conspire when the one series is an even number of half-wave lengths, and they destroy each other when the one series is an odd number of half-wave lengths in advance of the other.  With two forks so circumstanced, we obtain those intermittent shocks of sound separated by pauses of silence, to which we give the name of beats.  By a suitable arrangement, moreover, it is possible to make one sound wholly extinguish another.  Along four distinct lines, for example, the vibrations of the two prongs of a tuning-fork completely blot each other out.[12]