JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS The effect of reflection at a surface is to reduce the amount adsorbed. Molecular polarity, molecular non-symmetry and surface contamination by other materials very considerably reduce the amount adsorbed to below that calculated because of the reflection of the gas molecules due to these causes: e.g., determinations show that 96 per cent of water vapour molecules hitting a water surface are reflected. It is a reasonable assumption that molecules which strike a surface at a position already occupied by a molecule of the same species will be either reflected or, at most, held for an exceedingly short time at the surface. PART III THE MECHANISM OF OLFACTORY ]½,ESPONSE It is generally agreed that smell is the result of the stimulation of the olfac- tory receptors in some manner, so that a message is sent through the olfactory nerve to the brain. The mechanism of conduction of the message along the nerve is the same as that for any other message along any other appropriate nerve, but the receptor mechanism is unique to smell. We have seen that all sensory stimuli are converted by receptors into electrical impulses which nerve fibres transhilt to the brain. In the case of light, the energy received is first absorbed by substances which undergo chemical change, generating electrical energy at the same time. In the case of sound, the energy of vibration of gas molecules is converted by the movement or stretching of hairs into an electrical impulse. The working principle of the olfactory receptor is not yet fully known, but something of its construction has been established and we should be able to infer its modus operandi. If we postulate that smell is a chemical sense analogous to light, we should expect there to be primary odours (there are three primary colours) from which any odour could be produced: in the same way that a colour can be produced from its primaries, so could an odour be produced. When blue and red lights are projected on to a white screen, the eye responds as if only green were projected on to it and is quite incapable of distinguishing the red and blue lights individually. If smell is a physical sense analogous to hearing, we should expect there to be a multiplicity of primary odours just as there are notes of recognisably different pitches, a law similar to Ohm's law should apply to compound odours and, just as the ear can appreciate the whole of an orchestra as a unit but can also pick out the various instruments, so should the nose be able to smell a perfume as a whole and also to pick out its ingredient odours. Obviously smell is closer to hearing in this respect.

SMELL--THE PHYSICAL SENSE 59 When we compare the structure of the olfactory organ with those of vision and hearing, the closer approach to the latter is again evident. The olfactory receptors terminate in a hairy structure and they have to convert energy, brought to them by rapidly moving gas molecules, into an electrical impulse. In both of these characteristics they resemble the auditory receptors closely. Let us postulate that the olfactory receptor works in a physical way: that odoriferous gas molecules are adsorbed on to the hairs of the vesicles at the end of the nerve fibres and that the energy represented by the heat energy liberated by the adsorption is converted by the vesicle to an electrical impulse. If these postulates are true it should be possible to explain the various phenomena of oilaction by reference to the known properties of nerve systems, of the adsorption of gases and of the construction of the olfactory area. The observed phenomena of oltaction given in Part I have there been arranged in such order that they are easily assigned to sections whose fundamental basis is similar. We have groups concerned with the properties of the nervous system, the properties of the sensory receptors, the properties of gases, and the properties of solutions. Referring to these under the same reference letters and in the same order as given in Part I, we have the following explanations: (a) Adaptation and Fatigue. Adaptation has been found to be a character- istic property of the brain cells, of the nerve fibres and of sensory end-organs (particularly of those of the skin, where it is much more rapid than in the nerve fibres). It is reasonable to expect the olfactory receptors to resemble other end- organs, and, like them, rapidly to come to an equilibrium state with any given level of stimulation by a particular odour. Provided this level is below that giving maximum olfactory response, further increase in odorant concentra- tion should make the substance perceptible once more. This is often made use of by a perfumer when studying a compound in order to copy it: he lets the smell diffuse over the olfactory area, mentally indexing the odours as they are perceived one after the other. During this performance, the more volatile smells are recognised first because they more rapidly reach their maximum concentration then, as the olfactory area comes into equilibrium with, and adapted to, them and they are no longer recognised, the next set of not quite so volatile materiMs are now recognised until adapta- tion sets in after they have reached their maximum adsorption, and so on. This process of adaptation is essential for the recognition of the odour ingredi- ents. If the perfumer now wishes to check his assessment, he clears his nose by passing air over the olfactory area and starts again, this being possible because the clearing has removed the stimulants from the olfactory area, which is then once more in a "reactive" state.