THE Departmental Committee on Sight Tests has recommended a method of classifying colour-blindness by measuring the luminosity of the colour sensations by means of the flicker method of photometry. The degree of abnormality is estimated by the ratio of red to green compared with the normal. This classification is absolutely erroneous. Light perception and colour perception are quite distinct—that is to say, there may be considerable defect of colour perception without defect of light perception. The first two cases of colour-blindness (dichromics who confused red and green) examined by me on the method suggested by the Committee had a ratio identical with the normal, whilst a man who had not the least defect of colour perception had an abnormal ratio. Prof. A. W. Porter and I examined one of the above-mentioned colour-blind men by another method, and we could not detect the least defect in the perception of luminosity in any part of the spectrum. We ascertained the point of extinction and the point of reappearance of light from all parts of the spectrum.