MORE TRUTH ABOUT ODOUR A few years later, Bijtel and Van Iterson (Arch. Neerland. Physiol., 1925-26, 10, 447-448) published a brief report on the psychogalvanic reflex produced by olfactory excitation. Using camphor, citral, caproic acid, glacial acetic acid, ammonia, cinnamic aldehyde and cigar smoke, they found that quantitative differences could readily be recorded but that there was considerable uncertainty about qualitative differences. These British and Continental researches were followed by an American psychological study in which one hundred adolescent children were examined, fourteen odorous substances being used. The investigators, Coombs and Shock (Psychol. Bull., 1935, 32, 531) divided the odours into three groups-- pleasant, least unpleasant and most unpleasant--and also studied odorous strength. The tentative conclusions which they reached were: first, that the affectire quality of the stimulus must be considered as well as the affectire strength as a factor determining the magnitude of the galvano- metric deflection and, second, that for equal degrees of affectire strength, unpleasant stimuli tend to give greater deflections than pleasant stimuli. Psychogalvanic procedures seem to have been adopted on at least one other occasion although there is some doubt about the scientific status of the "investigations" concerned. The subjects were film stars from Hollywood and the instruments employed were the "psychogalvanoscope" and the "electro-oleo-convector." Perhaps the chief value of the experiments con- ducted with these devices lay in the delightfully ironical comment which they elicited from the French writer who described them in Parfumeri• Moderne (1930, 24, 335 337). "Nous serions f•ch•s de jeter une ombre, m•rne l•g•re, sur ces illusions juveniles l'Europe n'est pas mfire pour les r•v•lations de l'•lectro-psycho- olfacto-galvanoscope, pas plus que pour celles de l'•lectro-oleo-convector. Les Francaises, notamment, n'ont besoin d'aucun outillage • lampes d'amplification pour vibrer plus ou moins agrdablement sous l'influence de tel ou tel parfum qu'aucun appareil ne leur fait choisir' leur go•t affin• leur suffit." DISSOLUTION IN AIR Continental workers on olfaction sometimes refer to the "solubility" of odorous substances in air and, indeed, it has even been claimed that, in order for a substance to be odorous, it must possess such solubility. Although it appears certain that there can be no specific solubility of one gas or vapour in another, it is perhaps not impossible that odorous substances behave towards air in a way not shared by materials which are devoid of odour. In this connection, the experiments of Erdmann (J. prakt. Chem., 1900, 109, 225-228) are most interesting. He decided to study the solubility of odorous and inodorous substances in liquid air and found that whilst ether, alcohol, water and carbon dioxide are practically insoluble in that material (they solidify and can be removed by filtration), odorous substances such as c itra 1 209

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS ionone and rose oil are appreciably soluble. Apparently his proposal to continue the experiments was not carried out and no confirmation of his results has been recorded. On the other hand, his evidence has not been refuted. THE OSMOCEPTOR THEORY At the conference on "Basic Odor Research Correlation" held in New York in April, 1953, the eminent American biochemist, James B. Sumner, very briefly outlined his new theory of olfaction (Ann. N.Y. Acad. Sci., 1954, 58, Art. 2, 68-72). Here, it is interesting to recall the rather similar theory advanced by another distinguished chemist, Ruzicka, about thirty- five years ago (Chemiker-Zeitung, 1920, 44, 93-94 129-131). His observa- tions led him to conclude that, in addition to volatility, specific solubility in air and ability to react with certain constituents of the olfactory mucous membrane are the two properties most essential for power to stimulate the olfactory nerves. These supposed constituents he named osmoceptors. He further supposed that there are two kinds of osmoceptor, namely, the primary and the secondary. Any odorous substance, on reaching the membrane, combines first with primary osmoceptors only. If there are sufficient primary osmoceptors to combine with the quantity of odorous substance present, or if such osmoceptors can be re-formed quickly enough by the organism, the quality perceived remains constant. But if all the primary osmoceptors are saturated, the substance may then react with secondary osmoceptors and the quality of odour may change, the change being slow if the concentration of the substance is low, and rapid if it is high. If the change should be very rapid, a "mixed" odour is perceived. Substances which react with primary but not with secondary osmoceptors cease to be perceived when all the primary osmoceptors are saturated, and the fatigue thus caused lasts until free primary osmoceptors are again available. Fatigue for one odorous substance presumably extends to substances having similar "osmophoric" groups but not to other substances. MODELS OF THE OLFACTORY SYSTEM Useful models of biological systems have been constructed from time to time. Perhaps the only one of these regarded as capable of illustrating the mechanism of olfaction is that of De Jong and Saubert which was described shortly before the Second World War (Proc. Sci. Section, Kon. Akad. Wetensch. Amsterdam, 1937, 40, 302-306, and Protoplasma, 1937, 28, 329-344). This colloidal model can be said to be not too remote from nature. Its designers began by supposing that the first stage--the only stage which they considered--in the process is a reversible change (possibly in permeability) in the olfactory cells resulting from penetration by gaseous molecules of odorous substance. Although the cell constituents probably chiefly affected are phosphatides--the phosphatides are both hydrophilic and lipophilic 210