An attempt is made to understand the negative results of Panofsky and Saxena and the positive results of Frisch and Olson in terms of the theory of the interference effects in high-energy bremsstrahlung from crystals worked out by Überall. Several of the theoretical approximations are examined in detail: the validity of the Born approximation, the calculation of temperature effects, the validity of the closure approximation for the crystal lattice, and the use of the Debye form for the lattice vibration spectrum. It is concluded that all of these are justified, except that a partial failure of the Born approximation may be responsible for the nonappearance of the central minimum in the Frisch-Olson experiment. Crystal imperfections and multiple and inelastic scattering of the primary electrons are also considered briefly, but are found to be unimportant. No explanation is found for the Panofsky-Saxena results. The interference should be enhanced by making the primary electron energy as large as possible and the ratio of photon to electron energy as small as possible. In the case of a diamond-type crystal the  direction for the electron beam is to be preferred to either the  or the  direction. Little advantage is to be gained from cooling the crystal.