TESTING THE EFFICIENCY OF DEODORANTS 109 Work of the nature discussed necessitates frequent identification of un- known materials by comparison of their spectra with those of known sub- stances. A large file of spectra of known materials is therefore essential. In the two years during which we have been working on the problem, we have accumulated a library of about five thousand spectra. Of these, about a thousand are of pure compounds. The value of such a file increases greatly as it approaches completeness, but so does the difficulty of using it. This problem is solved by the use of punched cards which sort the spectra mechanically. Two such systems are in general use one using Keysort cards and the other IBM cards. Each method has advantages. PROCEDURES FOR TESTING THE EFFICIENCY OF DEODORANTS* By EDWARD SAGARIN• Sagarin !nstilule for Olfactory Research, Inc., New York, N.Y. THE VERY •'IRS'r problem involved in determining how the effi- ciency of a deodorant should be tested is to define such a product. One cannot possibly test the quality of a material when there is no agreement as to what such a material should be. As the name suggests, it would seem more or less apparent at first glance that a deodorant would be any product that diminishes perceptible odor. Such a definition would not necessitate the abolition of such odor, although commercial products have been marketed with claims that they banish all odors. The above definition, however, is quite faulty in two respects. A deo- dorant, by its nature, should effectively diminish the specific unpleasant odor character against which it is utilized. That this can be, and often actually is, accomplished by the replacement with another and more acceptable odor, either through masking or by chemical transformation, does not make the product less of a deodorant. The second objection to the suggested definition is more fundamental. A material, to be utilized as a deodorant, must effectuate such diminution in comparison with a logical control product. Ordinary washing with plain water, without any soap or detergent or alleged deodorant ingredient, will diminish the odor by the mechanical removal of dirt and other decaying material. Would that justify the packaging and sale of a bottle of water as a deodorant? To ask the question is to answer it. * Presented at the December 11, 1952, Meeting, New York City. ]• Present address, Standard Aromatics, Inc., New York City

TESTING THE EFFICIENCY OF DEODORANTS 109 Work of the nature discussed necessitates frequent identification of un- known materials by comparison of their spectra with those of known sub- stances. A large file of spectra of known materials is therefore essential. In the two years during which we have been working on the problem, we have accumulated a library of about five thousand spectra. Of these, about a thousand are of pure compounds. The value of such a file increases greatly as it approaches completeness, but so does the difficulty of using it. This problem is solved by the use of punched cards which sort the spectra mechanically. Two such systems are in general use one using Keysort cards and the other IBM cards. Each method has advantages. PROCEDURES FOR TESTING THE EFFICIENCY OF DEODORANTS* By EDWARD SAGARIN• Sagarin !nstilule for Olfactory Research, Inc., New York, N.Y. THE VERY •'IRS'r problem involved in determining how the effi- ciency of a deodorant should be tested is to define such a product. One cannot possibly test the quality of a material when there is no agreement as to what such a material should be. As the name suggests, it would seem more or less apparent at first glance that a deodorant would be any product that diminishes perceptible odor. Such a definition would not necessitate the abolition of such odor, although commercial products have been marketed with claims that they banish all odors. The above definition, however, is quite faulty in two respects. A deo- dorant, by its nature, should effectively diminish the specific unpleasant odor character against which it is utilized. That this can be, and often actually is, accomplished by the replacement with another and more acceptable odor, either through masking or by chemical transformation, does not make the product less of a deodorant. The second objection to the suggested definition is more fundamental. A material, to be utilized as a deodorant, must effectuate such diminution in comparison with a logical control product. Ordinary washing with plain water, without any soap or detergent or alleged deodorant ingredient, will diminish the odor by the mechanical removal of dirt and other decaying material. Would that justify the packaging and sale of a bottle of water as a deodorant? To ask the question is to answer it. * Presented at the December 11, 1952, Meeting, New York City. ]• Present address, Standard Aromatics, Inc., New York City