DISCUSSION 37 CHAIRMAN: Thank you very much. MR. D}•NAvARRI•: I think that Mr. Chaleyer hit on something there. The authors agree that there is a fundamental need for a system of classification. Per- haps we are not as articulate as we might be in defining the different classes. For ex- ample, if you use the classification of acidity, I dare say you would get 100 per cent variation in this audience, most of whom like to think of themselves as being some- what adept at classifying odor. So perhaps the task is to try to define a language for the description of these sensations. In my own observations, I find that if you expose a fragrance to a group of people you can get a report of almost every type of sensation that you can think of. Some find it spicy some find it is heavy, others find it is flowery, and so on. I am sure that all of the group get the same odor sensation but they are inarticulate in describing what they smell. MR. CROWr)eR: It seems to me that basic reasoning behind all this is lacking--that we have no standard conditions. I do not be- lieve that the Crocker-Henderson system has tried to standardize such things as degree of light in the room, the color of the room, the temperature, the humidity and even the sound. All of these things must have some effect. The color of the glass, or what one is thinking about, may even enter into it. So, I wonder if Mr. Crocker has done any work in ruling out all these outside influences. MR. CROC•CER: We were aware from the very beginning of the importance of illumina- tion, sound, comfort, and many other things--the length of time after your last meal--many other physical considerations came in. Actually, no attempt was ever made to control the degree of humidity or degree of illumination. Sound was kept down, and the observers were always seated under comfortable conditions, as nearly home-like as possible. Physical comfort comes into,the picture very much. You can't concentrate on such a difficult thing as smell unless you are comfortable. The points you brought out are all correct. We find, also, that the psychological factors at which you hinted, but didn't develop, are very important, such as the person's ex- perience with things in the past. It was the custom of many of the GI's during the war to refer to bitter experiences. Well, that "bitter" thing in their mind is taste. Bitter taste seemed to them the most un- ple•sa nt thing in the world, so they expressed unpleasantness as bitterness. Well, we are groping very hard to express pleasantness and unpleasantness in the various factors of the sense of smell, and to most of us caprylicness is unpleasant, but that isn't true of all of us. Some like limburger cheese, which is about the highest point in caprylicness. So, there is no agreement on what is pleasant and what is unpleasant. DR. FOURMAN: There is one thing that Mr. Sagarin said which I think is a bit extreme, viz., that the sole hope for odor classification is the establishment of a sound and proved theory of odor. I don't believe that it would be necessary to establish a sound theory of odor in order to have a good classi- fication of odor, because if we take one of the greatest contributions to classification in science--that of Linnaeus in botany--where an attempt was made to classify thousands of plants--there was no theory of plants in- volved at all. It was based on morphological differences of form, on observable differences. So I believe that we could have a classifica- tion without firat having a theory, especially if the classification is of a qualitative nature rather than of a quantitative nature. CHAXRMAN: Mr. Sagarin will comment on the subject a little bit toward the end of the discussion. MR. KLExN: There are several points that I would like to take up in this discussion. First, the point made by Mr. Sagarin on the. scientific invalidity of odor classificationt Now I agree with that bald statement, bun all who work with aromatic materials sool, find the need for a practical, though persona odor classification. This classification is unscientific but it catalogues aromatics into a useful reference file. This "rule of thumb" classification may list an odorous substance as flowery, spicy, herby, woody, balsamic, or more specifically associate it with com-

38 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS monly known odors, such as, rose, jasmin, or cedar-like. Most perfumers, with similar experience in the realm of odor, are in rather good agree- ment in their personal &scription or classi- fication of specific odorous substances. Of course there •are normal differences in indi- vidual &scription or opinion, but there is usually a basic and fundamental agreement. For example sandalwood, cedarwood, and vetiver are regularly termed woody odors jasmin, rose, and cassie are regularly termed flowery odors. Now there are large differ- ences between each of these woody and flow- ery odors, yet perfumers will invariably group them together even though their secondary description will vary in fine points of differen- tiation of one from another. This cataloguing of odors is helpful and practical, even though it is unscientific. On another point: I believe that in May I was in communication with Mr. Crocker on his projected odor directory. At that time I inquired into the matter of reproducibility of different subjects using the Crocker-Hender- son method of odor classification. As I was not entirely satisfied with Mr. Crocker's reply I decided to spend some time with this method. As you know, this assumes all odors as comprised of four distinct com- ponents, each with eight degrees of intensity. These components are Fragrant, Burnt, Acidic, and Caprylic. Each odorous sub- stance is assigned a definite odor number. After adequate study of the method, I assigned a four digit number to each of the series of 32 odorous materials in the odor standards set. I found that my closest approach to the numbers assigned by Crocker-Henderson was within 1 number up or down on a single component. In no case was I able to check that closely on even two digits. For example, if the assigned C-H number was hypothetically 7777, I would possibly obtain 6923. In subsequent trials, I found that I could get closer to my own assigned figures than I could to the Crocker- Henderson figures. However, I was so far from accurately checking myself that I con- cluded that the C-H Method could neither be scientific nor practical for classification of odorous substances. At the same time I can see some value in the Crocker-Henderson method from the viewpoint of introducing people to odors and to some ideas concerning them. On another point: this was discussed by Mr. Chaleyer and others. This question of odor-language, and a committee to study it. Recently a perfumer attempted to or- ganize a committee of his colleagues for just this purpose. The attempt flopped com- pletely. The problem was stated, the com- mittee selected and...no meeting was held. The idea just expired. So you see, the problem of odor-language is so ghost-like that even experts don't feel that they have sufficient community of thought, or even data relative to odor description or classifica- tion to make such a meeting practical. Nevertheless perfumers will discuss and describe odors to each other and the terms used will convey quite a good meaning. One will say that methyl salicylate is spicy, an- other will add that it is sweet and piney, and so on. All of these terms fit the material. Here is an example of shades of opinion. I cannot recall any perfumer listing methyl sali'cylate as flowery. Thus the odor de- scription will vary from expert to expert, but the differences will be minor compared to the important fundamental but unscientific agreement. Dr.. Bo•evcr: It seems to me that in con- sidering this whole problem, there is one thing that has defeated a great deal of our efforts to solve it, and that is the nature of the ex- perimental animal you are using. Unfor- tunately, we have no satisfactory standard. No two humans are exactly alike. You have not only psychological elements, but you have physical elements as well. In other words, the variable nature and the variable conditions of the experimental animal we are using become important factors impeding satisfactory results. Unless and until we know more about what odor really is, and what takes place on oilac- tion, I don't see how the experimental animal--no two of which are alike--can give similar •es. ults.