The experiments of Newton were accepted as correct by Huygens and Hooke, probably the two most eminent physicists of the time. But they objected to his definition of a primary colour and to his conclusions. They had found previously, by their own experiments, that certain pairs of complementary colours, such as a certain red and a certain blue, gave to the eye the same sensation of white as did clear sunlight. They had defined them as the two primary colours which, by different proportions of mixture, would produce all other colours, including white. They therefore objected that it was unnecessary and cumbersome to assume an infinite number of primary rays when two were quite sufficient. By no process of reasoning could these two opinions be either reconciled or controverted. They involved fundamentally different criteria. White produced by the combination of a continuous spectrum and the white produced by a combination of red and blue were one and the same to Hooke and Huygens because their criterion of identity was the sensation of sight, and it must be the same for all psychologists who deal with subjective light. To Newton, the two whites were altogether different. The one examined by a prism gives a continuous spectrum, and the other gives two separated bands of blue and red. What, then, to the physicist is a fundamental dissimilarity is to the psychologist complete identity. How then can psychological sensations be studied by physical methods? To show that this is not an isolated case, but that this gulf runs between the entire fields of physics and psychology, I can give an artificial example. While I deprecate the pseudo-scientific pictures, now fashionable, of the state of prehistoric man and of the condition of the earth before its habitability, I am able to imagine a world in which the eye had never developed, so that all life was blind. I am sure the word colour would never have been coined in such a world and that there could be no psychology of sight, but I also know that the blind race could develop a physical science of light because very many of its phenomena can be, and are now exclusively, studied by such apparatus as thermometers and electric galvanometers which can be read by touch. Or, to cite an everyday example, is not the mere fact, that the sensation produced by pepper on the tongue cannot be distinguished from that of heat, sufficient to make us hesitate before trying to study the sensations objectively, to synthesise the objective and subjective worlds, or to try to investigate them by the same method?