On July 14, 1982 the Sweden and Britain Rarar-Aurora Experiment (SABRE) observed the ionospheric flow reversal boundary at ~0400 MLT to move equatorward across the radar field of view and then later to return poleward. The polar cap appeared to be considerably inflatted at this time. Concurrent observations by ISEE-3 at the L1 libration point of the solar wind speed and density, and of the interplanetary magnetic field (IMF) indicated that the solar wind conditions were unusual throughout the interval under consideration. A mapping of the solar wind parameters from the L1 point to the subsolar magnetopause and thence to the SABRE local time sector indicates that the equatorward motion of the polar cap boundary was controlled by a southward turning of the IMF. The inference of a concomitant increase in open magnetic flux is supported by a comparison of the magnetopause location observed by ISEE-1 on an inbound pass in the 2100 MLT sector with a magnetopause model based upon the solar wind measurements made by ISEE-3. Some 20 minutes after the expansion of the polar cap boundary was first seen by SABRE, there was a rapid contraction of the boundary, the cause of which was independent of the IMF and solar wind parameters, and which had a poleward velocity component in excess of 1900 m s-1. The boundary as it moved across the radar field of view was highly structured and oriented at a large angle to be ionospheric footprints of the magnetic L shells. Observations in the premidnight sector by the Air Force Geophysics Laboratory (AFGL) magnetometer array indicate that the polar cap contraction is caused by substorm draining of the polar cap flux and occurs without a clearly associated trigger in the interplanetary medium. The response time in the early morning local time sector to the substorm onset switch is approximately 20 minutes, equivalent to an ionospheric azimuthal phase velocity of some 5 km s-1. ¿ American Geophysical Union 1990