Abstract
The usual method of asymptotic estimation of Fourier transforms, the phase oscillation method, is examined. It is found that the usual argument of phase oscillation can be misleading due to either locations of singularities or absolute integrability of functions. As a by‐product we have found that the scaling law for the inelastic e‐p scattering should also be valid as the energy change v→∞ regardless of the momentum transferred squared κ, according to the usual approach of light cone dominance.