EYESWEET AND COLOUR SCIENCE IN COSMETICS 5O5 much like a little camera, with a lens in front and a screen at the back on which the lens can form an image (Fig. 1). The screen is immensely com- plicated and has not yet yielded up all the secrets of its structure and function, but we know enough to realise that it is a fine mosaic of light- sensitive elements, minute photocells, each of which is eventually con- nected to the brain by a complex system of nerve fibres (Figs. oe-4): light of different wavelengths may be supposed to affect these photocells to different extents, so that they can send impulses to the brain which depend upon the wavelength and upon the particular photocell stimulated. There is still nothing in this which could be called "colour", only an ab- sorption of radiation and an initiation of a nerve impulse, purely physical and electrical processes. The nerve impulses travel along the optic nerve to more and more distant stations in the brain and there, for the present, all definite knowledge has its boundary. We can only say that somewhere in the brain there occurs what we call perception of colour, perhaps in minute particular localities, perhaps in ill-defined patches, perhaps all over the brain as in the holographic concept recently put forward by Longuet- Higgins (1). To the tidy-minded scientist this is a distressingly vague ending, but fortunately the various factors which precede it are amenable to quite precise description and measurement, and to this extent the "perception of colour" is well understood. COLOUR VISION We have seen how light can be analysed into its component wave- lengths which finally excite different colour perceptions in the brain. When the original mixed light falls upon a surface it is in part reflected back, the rest being lost by absorption within the surface. A white surface is one which reflects equally and almost completely all wavelengths of light a grey surface reflects equally but only partially a coloured surface reflects both partially and unequally. The appearance of colour results from the loss of parts of the original white light: what is not lost is reflected back (Fig. 5), and can be seen by the eye. This selection of spectral intensity is the only function performed by a surface, or by a transparent substance such as stained glass, in the production of colour. Nevertheless, the property of "colour" is commonly ascribed to surfaces, pigments, glasses, liquids, etc., as if it belonged solely to them. In fact, the most intricate and myster-

506 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS ious part of the process of seeing colour comes only when light has travelled from source to surface and from surface to eye. As soon as the light has been absorbed, wholly or partially, by the receptors of the retina, its role is finished. It is transformed into nervous impulses which travel to the brain. In some way these impulses must carry the information, previously carried by the light, coded in such manner that the colour information in the light is perceived by the brain. There is little doubt that the primary coding at the level of photoelectric transformation in the receptors is followed by many re-codings, cross-codings and group- codings. The surprising thing is that, through the intricate maze of our higher nervous system, perception finally occurs in such perfection, crystal- clear and, usually, unambiguous. The disentangling of the various stages of nervous transmission and coding is one of the great preoccupations of the sensory research worker at the present time, but the nature of the primary, receptor coding has been well established for many years. It is described as the trichromatic theory of vision, first proposed by Young (2, 3), who was unconciously following up an idea proposed by Wiinsch (4) it was forgotten for many years, then later revived by Helmholtz (5) and given an experimental basis by Max- well (6). The experimental fact which provides the basis for this theory is that the colour of any light stimulus, whether coming to the eye direct from a source, or after reflection at a surface, can be matched in appearance by a mixture, in appropriate proportions, of not more than three chosen stimuli, i.e. the three chosen stimuli are all shown upon the same area of a screen and can then be adjusted to match in appearance any other light stimulus. This principle is illustrated by Fig. 6. It must be admitted that this state- ment is an over-simplification, but the only extension we need consider is that one or other of the three chosen stimuli, called the primary stimuli, must sometimes be imagined as having a negative value. This is, of course, physically impossible: what happens in practice is that certain stimuli must have one of the primaries mixed with them in order to match the mixture of the other two. It is rather like an algebraic equation in which Test q- Primary A = Primary B q-Primary C This is physically possible, but on paper can be algebraically transformed to Test = Primary B q- Primary C--Primary A which is physically impossible but preserves the basic concept of the trichromatic theory. These primaries, A, B, and C, are physical stimuli, but they can be transformed algebraically into three fundamental sensations (7)