ness cycles places a serious difficulty in the way of a quantitative theory. Mathematical models cannot be checked and even their significance is open to question. All useful laws are, no doubt, only approximations to reality, but in the economic case the approximation is intolerably poor. Time series, trend and seasonal apart, not only fail to conform to simple sine curves, but their shapes, their amplitudes, and their periods vary. Some economists have quite correctly said that there is no conclusive evidence of the wave nature of economic life. The mere fact that time series sometimes go up and sometimes down is insufficient; nevertheless, the impression that economic evolution is cyclic cannot be so easily dismissed. The two points of view can be reconciled by adopting the hypothesis that we are dealing with an oscillatory mechanism which is disturbed.