Doppler sorting and angle of arrival have been used in an attempt to identify localized scattering features in the mesosphere. These might be caused by isolated perturbations in electron density carried along by the neutral wind or by some focusing mechanism in which a stable horizontally stratified scattering layer is perturbed by a wave motion, creating a moving ''glint.'' Discrepancies were found between interferometer and ''true'' (spaced antenna drift analysis) speeds using T=3.2-min record lengths and averaging over 1 hour. These records have good frequency resolution (1/T), but the length combined with normal wind speeds present means that single scatterers will be spread over several Doppler bins, and it is always possible that echoes from different locations combine to give a false angle of arrival. Consequently, series of shorter records (72 s) were examined in the hope of tracking echo location. Two examples have been found. Both occurred during signal strength ''bursts'' near 85 km, and their appearance suggests ''glints'' from 5-min period waves, one traveling with and the other traveling ~60¿ away from the background wind direction. The signal bursts themselves are of interest, having a regular repetition rate (30--40 min) and being very common at these heights in the summer. It is possible that these bursts are due to focusing or turbulence created by long-period waves. ¿American Geophysical Union 1987