Plate 
to random modulation; you do not get a breaking of waves owing to 
the fact that they are real waves. Also, you do not find waves which 
are skewed like the ones I showed in my paper. And I think that you 
arrive at wrong conclusions, or no better conclusions, from this 
kind of study than those which you would get by just using a sequence 
of experiments with sinusoidal waves of increasing frequency and 
drew your response diagram in the usual manner, because you are 
basically trying to solve a non-linear wave surface problem by a 
linear superposition method. If you allow yourself to make this appro- 
ximation, why not go all the way and just use sinusoidal generators, 
which would cost you a lot less money? 
1396 
