From July 1976 to November 1977 regular incoherent scatter observations of E and F region ion-drifts were conducted at Millstone Hill (42.6¿N) with the steerable (L band) and fixed (UHF) radars. Data were collected on a total of 46 days and analyzed for ionospheric electric fields and daytime neutral winds in the lower thermosphere at 105, 115, 125, and 135 km. The wind measurements characteristically exhibited the dominance of the semidiurnal oscillation with both wind components, showing a 12-hour variation and an eastward wind maximum 3 hours ahead of a southward wind maximum. A model fit to the full set of wind measurements alos allowed the smaller annual mean diurnal winds to be determined. The winds measured on 38 geomagnetically quiet days were grouped by day-number to determine the seasonal dependence of the semidiurnal tide as well as the prevailing wind. The semidiurnal winds exhibited the following general characteristics. They typically grew in amplitude reaching a maximum in the vicinity of 125 km with larger values at the equinoxes than the solstices. The wind field at 105 km showed a significant polarization, with southward wind amplitudes larger than eastward amplitudes. Seasonal variations in tidal phase were generally less than one hour. The amplitude and phase of the measured diurnal wind were in reasonable agreement with theoretical expectations for the in situ tidal component, and there was no indication of the presence of an upward propagating diurnal component. On the other hand, the measured semidiurnal winds were consistent with being the result of the upward propagation of tidal energy from below 100 km rather than in situ tidal component, and there was no indication of the presence of an upward propagation of tidal energy from below 100 km rather than in situ excitation. Attempts to match the measured vertical wind structure with theat expected for the different semidiurnal Hough mode extensions in the lower thermosphere lead to no definite mode identification. The measured vertical wavelengths (70--100 km) and the altitude of maximum amplitude were in best agreement with the S2.2 mode, while the temperature/wind ratio suggested the S2.4 mode. There isoverall agreement of the Millstone Hill southward wind measurements with results from the French incoherent scatter radar at Saint Santin, but diffrences are noted as to the seasonal variation, the altitude of maximum wind, and the vertical wavelength. Some of these differences might be explained by a solar cycle dependence of tidaldissipation in the lower thermosphere.