“…We suppose that (12a) is generally valid, since its simple form suggests that the only essential requirements are the periodicity of the average crystal and, classically, the occurrence of the Doppler effect. Calculations of TDS with anharmonic motions, but retaining the convolution approxi-* This part of the paper was initiated by a correspondence with Professor R. F. Stewart, Groningen. marion, were performed by Hahn & Ludwig (1961), Hahn (1961) and Ludwig (1967. In these papers, (12a) was not explicitly established, but it appears that anharmonieity changes the intensity distribution of TDS, but neither its loearion between the reciprocal lattice points nor its inelasrie nature.…”