The tunnelling probability through a pair of high barriers separated by a distance a=1, as a function of wavenumber k. Note the regularly-spaced, narrow resonances where the probability is unity. (The first few peaks are so narrow that they can't be plotted accurately. A higher resolution would show that all of them peak at 1.) The resonances sit close to multiples of π, the allowed values of k for an infinite square well of width 1. As the energy increases, it gets easier for the particle to tunnel through the barriers and so the resonances get broader.