NEW APPROACH TO ODOR CLASSIFICATION tion. The eye makes use, ap- parently, of different perceptive mechanisms for different nerves, so that different individual nerves re- spond to different stimuli. The number of "classifications" is in- herently thus limited, at most, to the finite number of different nerves in the retina. Note the difference. The ear nerves are most probably all activated by the same mech- anism at the perceiving end, the analysis being pre-performed ac- cording to the different placement of the nerves and the mental inter- pretations due thereto. The eye, having no analytical mechanism dependent upon physical or spatial placement comparable to the coch- lea of the ear, must instead de- pend upon different types of sensing for different nerves, of finite num- ber. The eye, however, divides by two the number of different types of sense-impressions necessary, by utilizing the nature of color, and the involvement of more than one nerve of the same discriminating type. If an individual nerve is "set," as it were, for detecting red, by what- ever mechanism, it will also trans- mit the sensation green, since a negative sensation of red is a posi- tive sensation of green in ordinary daylight, with intermediate com- binations which are thus seen to be a matter of intensity of impression. The color attribute of "brightness" we shall leave alone, for the present, relying on the total sensation throughout the retina to give us that illusion. To further limit the number of different classifications of nerve- endings to be postulated, it is not necessary to assume as many classi- fications, divided by two, as there are conceivable colors. Many of our conceivable colors are scientifi- cally inadmissible to the spectrum. It is logical to assume that it is the spectrum upon which color vision depends, color per se having no physical attributes not also at- tributable to any portion of the spectrum. For instance, there is no "brown" in the spectrum. How- ever, since brown is a mixture of different spectrum colors in dif- ferent proportions, the eye can re- solve the total color into a com- pound sensation involving x groups of nerves. It does not seem conceivable that odor-perception is analogous to auditory perception. There would not appear to be any anatomical mechanism available for this "posi- tional analysis." This premise need not, however, rest entirely on this mere anatomical convenience. If it is demonstrated that the nerve- endings in the odor-detecting area are different histologically, or if there is the possibility of anesthe- tizing small areas, this necessity to differentiate between the two broad differences in method of per- ceiving can be decided experimen- tally. The concept of "complementary" odors, necessary if we accept the visual mode of sense perception, is intriguing. We have a known anal- ogous phenomenon in the sensing of odor to visual-fatigue. In fact,

24 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS this attribute is even called "odor fatigue." Is it possible that delib- erately fatiguing the sense to one "type" of odor and later smelling other "types" and noting differences between fatigued and unfatigued odor perceptions may furnish a valuable and definitive clue? Is it possible further that spraying the nasal passages with an effective antihistaminic or alternatively with an effective anesthetic may serve to differentiate between sense impres- sion and pain or trigeminal impres- sions? Could it be demonstrated that just as taste (defined colloqui- ally) has odor components, that the reverse is true? Could "fragrant" have a "sweet" taste associated with it, so that "fragrant" is therefore not one of the "basic elements of odor," as I believe it is not? This could be decided by anesthetizing the sense of taste and smelling dif- ferent types of odor, before and after. Should another component of taste analogous to the pain-in- volvement in odor be postulated, and possibly be called "metallic," which "metallic" component might further function as a perceiver of, say, "burnt" odor ? The complexity of the whole sub- ject can thus be narrowed down to a few principles, by observing the results of a few definitive experi- ments. The guiding postulate can be, in all cases, that a single nerve can detect or transmit nothing but different intensities of only one sensation. We must therefore strive to eliminate as many of the complex sensations, and sensations which do not properly belong to the odor sense per se. The first major step is to determine whether the eye-type mechanism is involved, or the ear-type. One case will show us that an odor-classifying system is possible, after sufficient work, and ear-type will show that a continuous scale of some underlying funda- mental is involved, which we can- not, at this time, define. The next step, assuming the eye-type, will be to find the physiological "basic elements" of odor. Then, maybe, we shall have the key.