Three experiments involving threshold measurement and brightness scaling were conducted with ten dark-adapted subjects. It was demonstrated (1) that the experimentally determined absolute threshold rises with the level of stimulation prevailing in the particular experiment, (2) that threshold values extrapolated from the psychophysical brightness function agree closely with thresholds experimentally determined under proper conditions of stimulation. Finally, it was shown (3) that the psychophysical brightness function is approximately linear for low luminance levels, and this unexpected finding was supported by a re-analysis of data reported by Jameson. © 1968.