Troughs in the auroral ionosphere have been observed in satellite and radar data since 1965. One explanation for the appearance of these troughs is the stagnation theory of trough formation. According to this model, magnetic flux tubes are convected from the dayside to the nightside across the polar cap as a result of the impressed magnetospheric electric field. These tubes return to the dayside by subsequently drifting westward or eastward. Tubes drifting westward must travel faster than earth rotation if they are to reemerge in sunlit regions, and some fail to do so, allowing the ionization in them to fall to very low levels as it recombines. The trough is the locus of flux tubes which spend a long time in darkness due to the competing effects of the convection electric field and earth rotation. Commencing in 1979, a new observing technique was implemented at Millstone Hill in which the 150-ft antenna is moved slowly under computer control as data collection proceeds. The electron density results obtained with these experiments show that when a trough does form there is a great variation in when and where it is seen and how long it is present. On some nights the trough does not form at all. Furthermore, the appearance of the trough in the data often seems to occur in regions where the electric field is very large, with the ion drift velocity often far exceeding the speed of earth rotation. None of this behavior is entirely consistent with the stagnation theory of trough formation. An alternative explanation is to suppose that troughs are formed in regions of very strong convection (>1 km/s) as a result of the increase in the rate of charge transfer between O+ and O that occurs when the ions are driven through the neutral gas at high speed. According to such a view, “troughs” would be created in localized regions during substorms. The troughs might be created in regions of strong westward flow in the dusk sector. They then could migrate south and eastward depending upon the effects of convection and corotation. In an effort to test this picture we have begun conducting observations through 360° azimuth to see if the ionosphere to the northeast of the radar is the same as that seen later to the northwest. Results are presented for December 21–22, 1980. These do exhibit significant differences, tending to confirm that the trough is not a persistent extended E-W feature.