Abstract
In the case of electron diffraction by crystals containing heavy atoms, it is possible to improve the usual kinematical theory by formulating a pseudokinematical theory, which is expressed in terms of rigorous atomic scattering amplitudes rather than their Born-approximation values. A corresponding improvement can be made in the treatment of the problem by the dynamical theory, and consists in taking into account the coupling of a given diffracted beam with not only the incident beam (as is usually done in first approximation) but also with other diffracted beams. The pseudokinematical and dynamical theories are compared for very thin crystals and the conditions under which both give identical results are discussed. In particular, both predict a failure of Friedel's law in the case of acentric crystals.