40 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS of aidehyde, alcohol, ester, or whatever the case may be, these are the properties by which he determines the quality of a product. The perfumer, on the other hand, is not a scientist, he is an artist. His analytical instrument is his nose, and the trained nose is much more sensitive than, say, a refractometer. It will detect traces of impurities which cannot be detected by the ordinary routine methods of chemical and physical analy- sis. It is true that such methods as ultra-violet and infra-red spectroscopy are recently being used more widely for the evaluation of aromatics, but their use is still far from being universally adopted, and we therefore have to depend largely on the perfumer's nose. As I said, the perfumer is an artist, and as an artist he is not logical. For whereas he expects one type of chemicals to be so pure that no impurities can be detected by the most sensitive of analytical instruments--the per- fumer's nose--he expects, on the other hand, other types of chemicals to contain impurities. In fact, he will only accept such chemicals when they are impure and will reject them in their pure state. Let me give you some examples of such chemicals, again using the classification mentioned earlier on. First of all, chemicals containing traces of starting materials. The odour of the Schiff's Base formed by the action of hydroxycitronellal on methyl anthranilate is, in my opinion, due to traces of both these starting materials. If one were to remove them this product would become useless to the perfumer. It is just this very proportion in which hydroxycitronellal and methyl anthranilate are present in this Schiff's Base, together with its fixative properties caused by its low volatility, which makes it such a valuable aromatic. Certain acetals may also be mentioned in this connection. There certainly is no doubt about it that there are many acetals which possess odours distinct from those of the parent aidehyde. But whenever I have the opportunity of smelling certain other acetals I cannot help feeling that the odour of traces of the parent aidehyde contributes to no small extent to the odour character of the acetal. Now we come to the second category, the chemicals which contain by-products and isomers formed during their preparation. It has been frequently stated that if one were to prepare certain products in an absolutely pure state one would obtain materials without any odour at all. The acetate of trichloromethyl phenyl carbinol is a case in point. Several years ago Dr. Mitchell published a paper in Perfumery • Essential Oil Record' in which he stated that on repeated crystallisation of the commercial product an odourless ester was obtained. Obviously the rose odour which the perfumer expects this ester to possess is due to some by-product formed as a result of a side reaction. It would be senseless to remove it and produce an odourless perfumery chemical.

SOME ASPECTS OF PERFUMERY CHEMICALS 41 A similar instance is that of peach aidehyde. The typical--shall we say-- "spicy note" of this aromatic is due to traces of dihydrojasmone produced as a by-product as a result of a side reaction. Peach aldehyde which has been so completely purified that the last traces of dihydrojasmone have been removed has a fatty, dull and uninteresting odour. In my opinion the coconut note in so-called coconut aidehyde is also produced by a trace of a by-product. When this is absent the lactone smells very similar to peach aldehyde from which the all-important trace of dihydrojasmone has been removed. Certain brands of coconut aldehyde have a peculiar plum-like note. This, too, is caused by a by-product, so that according to the method of working up or purification of the crude reaction mixture it is possible to obtain either a coconut-like, or a plum-like, or even an indifferent smelling aromatic. In the production of ionone the two isomers alpha and beta ionone are obtained. Both are used individually in perfumery, but their mixture, known in the trade as "Ionone 100 per cent," is an even more valuable aromatic. Alpha ionone is always predominant in this mixture, but the proportion of beta ionone depends entirely on the manufacturing process used for the preparation of the final product. This is the reason why 100 per cent ionones from different sources may have different odour characters. In some reactions even more than two isomers are formed. For example, four products are obtained in the manufacture of methyl ionone: alpha and beta methyl ionone and alpha and beta iso methyl ionone. The proportion in which these isomers are found in the final product will determine its final odour quality. Now let us turn to the last category, the chemicals which contain decom- position products formed during storage. The example which springs to mind here is alpha amylcinnarnic aidehyde. This aldehyde undergoes oxida- tion on standing, and the decomposition products are benzaldehyde and certain fatty acids. On one occasion I prepared a highly purified sample of amylcinnamic aldehyde which was stabilised to prevent oxidation. To me it smelled like pure jasmin. But when the sample was submitted to a highly competent perfumer his verdict was that it lacked in body. He was missing the familiar fatty almond note which is always present in amylcinnamic aldehyde together with the jasmin note. In his opinion I had produced a very nice product which, however, he considered to be a new aromatic and which he could not introduce into his compounds in place of the amylcinna- mic aldehyde he was used to. Why then does the perfumer use aromatics which are known to contain impurities and uses them because of the odour produced by the impurities ? In my opinion this is due to the fact that when these products were first introduced into perfumery they were not known to be impure. Most syn- thetics were introduced at a time when the technique of synthetic methods