Measurements of the width of the OI λ5577 line made at night with a Fabry-Perot interferometer are presented and the observed variations discussed. It is shown that one type of variation was associated with relatively small magnetic disturbances. Another type was apparently due to the rapid recombination in the F-region in the evening, but the observations pointing to this were not sufficiently numerous to make it a certainty. A third and rarer type is not properly explained. All these types are believed to have been caused by λ5577 emission from above 200 km. A fourth type was almost certainly due to a temperature variation in the region emitting the green line by the Chapman process, and this variation appeared to be seasonal, coinciding in phase with the seasonal variation of the green line intensity. Some preliminary attempts to detect a Doppler shift due to wind in the 90–100-km region are briefly described.