Simultaneous observations of the vertical distribution of sodium density and the intensities of the OI 5577-¿, NaD, and OH (8,3) airglow emissions show that the nocturnal variations of the emissions are well correlated with those of the sodium at 91, 87, and 84 km, respectively. In the case of NaD and OH (8,3) these heights correspond closely to expected heights of peak airglow emission, although in the case of OI 5577 ¿, maximum emission would be expected from about 4 km higher. The correlated variations appear to be caused by atmospheric waves having vertical wavelengths between 10 and 15 km whose phases propagate downward with velocities between 1 and 2 km hr-1. The amplitude of the correlated nocturnal variations in sodium density on the bottom side of the layer is normally too large to be explained as the direct result of atmospheric density changes. These large variations could be the result of vertical motions of the layer or might be caused by changes in sodium and its oxide. On a number of occasions, large nocturnal changes in sodium density were seen to be closely correlated with changes in OH rotational temperature, suggesting a temperature effect on the chemical equilibrium. Enhanced sodium density, observed on days of exceptionally high OH (8,3) band rotational temperture, could either be the result of increased sublimation from sodium-bearing aerosols or again might be caused by a temperature dependence of the sodium photochemistry.