COSMETIC CREAMS AND LOTIONS 111 was included on the basis that some studies have been made with evalua- tion of axillary odor from that area of a T shirt. Direct sniffing of the axilla has not yielded a level as high as overpowering but clothing such as a T shirt has. At this time, these 5 levels are suggested as covering the neces- sary range and also permitting evaluation on clothing if interested parties consider that this should be included in a test procedure. First consideration should be given to odor levels obtained by direct sniffing of the axilla. One laboratory found a correlation between direct sniffing of the axilla and that obtained from a 1-inch cotton ball swab of the axilla stored in a 1-oz. ointment jar with the cap liner removed from the cap of the jar. On repeating this work more than 5 years later, it was found that the odor on the swabs dissipated rapidly so they could not be used. The nature of the swab material varied from the original used and can account for the trouble. At the present time, Zobec pad material seems to retain axillary odors and correlate with direct sniffing of the axilla. What- ever is used must give comparable data with direct sniffing. 3-5. These paragraphs are self-explanatory. 6. Laboratories have found like odor intensities in L and R axilla when no deodorant preparation has been used. Therefor% the control axilla must have a definite characteristic axillary odor in order to judge the effectiveness of a product used in the treated axilla. For checking between laboratories, the bench mark preparation, mentioned under Antiperspirant Test is suggested. SUBCOMMITTEE ON DEODORANTS AND ANTIPERSPIRANTS G. W. Fredell, Chairman Dr. J. Longfellow Report on EVALUATION OF GOSMETIG GREAMS AND LOTIONS CREAMS ASD •,OTtOSS may be tested for safety, for efficacy or for attractiveness. Methods of testing for safety have not been discussed by this subcommittee, since they are very much the same for a wide variety of cosmetic products, and should be left for the consideration of the Com- mittee on Laboratory Methods for Product Testing as a whole. It may be worth while to remark in passing that an event greatly to be desired is the development of suitable in vitro tests to replace or supplement the lengthy and tedious animal tests now used to screen possible new ingredients. Advances in tissue culture techniques and the use of mast cells to assess

112 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS several pharmacological effects of chemicals are straws in the wind which may encourage the belief that safety testing may eventually be a less for- midable and a less empirical procedure than it is now. Testing for efficacy has not been undertaken by our subcommittee. Creams and lotions are designed to produce diverse effects, many of which are extremely difficult to assess. An effect frequently sought is that of increasing water content of skin to which the cream or lotion is applied. Papers recently read before this Society may point the way to a practical method for determining how effective a preparation is in moistening the skin, and the standardization of such a test may well be a future under- taking of this subcommittee. The attractiveness or esthetic appeal of a cream or lotion has so many facets that it must be considered piecemeal. The color, gloss, opacity, surface finish, texture, consistency, odor and feel are all qualities which may affect esthetic appeal and some of them are very difficult to measure or evaluate objectively. Our subcommittee has made a modest start on this complex task by tackling the problem of setting up specifications which will result in reproducible viscosity readings on liquid emulsions. Samples of each of three liquid cosmetic emulsions were supplied to four laboratories by S. J. Strianse, a member of our subcommittee. Detailed instructions were issued for the viscosities of these emulsions to be read on given days, using Brookfield viscometers at given speeds for given lengths of time at a given temperature, etc. A very considerable amount of work was done by our subcommittee members in making the periodic readings stipulated, tabulating the results and reporting them. In spite of this ef- fort, all that we can now report is that our specifications were lacking in some of the details essehtial for consistent results. We shall have to spell out additional conditions before we have a standard method. It is to be expected that viscosity readings on liquid emulsions of the type we are considering should be affected by many factors. Not the least is the structural nature of the system, which carries with it the consequence that every time the system is made to flow, its structure and therefore its viscosity is altered. This results in different apparent viscosities for dif- ferent cylinder speeds. The theory of structure of soap-type gels, and the mechanism of forma- tion of such structures, appear not to have been adequately worked out. One possible approach is that of Bourgoin and Joly who developed a theory ofgelation for gelatin solutions.* By a process of elimination they reached the conclusion that flow during the first phase of sol-gel transformation involves deformation of a structure and breaking of bonds. They derived equations for the relationship of structural bond breakage and reformation * Bourgoin, D., and Joly, M., Kolloid Zeit., 146, 121 (1956).