A theory of colour vision

Abstract

At the present time there is no generally accepted theory of colour vision, and there is considerable divergence of opinion as to the tests which should be applied for colour blindness. The young-Helmholtz theory certainly explains the facts of colour mixing, but it is doubtful whether we can deduce from these facts the existence of three primary colour sensations. For, if we represent colours by particles on a plane, making the masses of the particles proportional to the luminosities of the colours, and represent mixtures of these colours by the rule for determining the centre of gravity of the particles, as is done on Newton's colour diagram, then, if any triangle be drawn enclosing all the possible points, its corners may be taken as representing the three primary colours, and these corners may take up an infinite number of positions. The usual procedure is to make two sides of this triangle fit the curve representing the spectrum as closely as possible, and take two corners of the triangle on the ends of the curve. Two of the primary Colours are then definite colours of the spectrum, while the third is a sensation which can never be produced pure. But there is a certain amount of arbitrariness about this proceeding.