Abstract
Conclusions from earlier investigations indicating an average luminance of 2 cd/m 2 as necessary for dense traffic were confirmed by appraisals of luminance level and visibility trials. The colour of the light influences the visibility as well as the appraisal of luminance level, pattern and discomfort glare. In all these respects sodium lamp lighting provides more favourable results than those obtained with colour corrected mercury lamps under otherwise identical conditions. At higher values of the road luminance a lower degree of uniformity in the luminance distribution is acceptable. A formulae, derived from results of earlier investigations, is found to be a suitable criterion for discomfort glare.