During a 40-day period in 1983, ISEE 3 was located about 225R(E) behind the Earth and remained within 12 R(E) of the nominal tail axis. During this time the spacecraft spent at least 70% of its time in the magnetotail with occasional excursions into the magnetosheath. However, during five geomagnetically disturbed intervals of 1-3 days duration during this period, ISEE 3 remained within the magnetosheath for extended intervals, even when it was very near the center of an average tail. Simultaneous observations of the solar wind direction and thermal pressure suggest that nonradial solar wind flow associated with interacting solar wind streams moves a compressed tail away from the nominal position at these Limes and explains most of these observations. However, during several few-hour intervals of strongly northward interplanetary magnetic field (IMF) within these periods, the solar wind is more radial and cannot explain the residence of the spacecraft in the magnetosheath. At these times ISEE 3 seems to be moving back and forth between two regions, one a higher-density, lower-temperature magnetosheathlike region but with density somewhat lower than the normal magnetosheath, the other a lower-density, higher-temperature taillike region but with density higher than the normal tail. Both regions have large B(z) components and B(x) components that tend to vary as if the spacecraft were moving from one hemisphere of the tail to the other. It is suggested that the magnetotail at these times of northward IMF consists mostly of field lines that close Earthward of the spacecraft with a narrow remaining tail at 225 R(E) waving back and forth across the spacecraft. This suggestion is supported by recent MHD simulations of J. A. Fedder and J. G. Lyon (submitted manuscript) for northward IMF conditions that show a completely closed tail that extends only to 155 R(E). If relatively rare intervals of long-duration, very northward IMF can eliminate the extended tail, it seems likely that more common, less northward IMF might well have very important, though less drastic, effects on the tail configuration.
