246 ]'OURNAL OF THE SOCIETY OF COSMETIC CHEMISTS surface, and their oily secretions, although relatively sparse, are thought to be a contributing factor in the formation of body odour (3). The eccrine glands are distributed over the whole body surface (4), they secrete a pro- fuse watery liquid which is responsible for the unsightly wet patches under the arms, which appear during hot weather or during moments of stress. Most people, when they are in conditions of moderate temperature and little or no anxiety produce very little sweat. For the purpose of the anti- perspirant tests, therefore, it is necessary to induce sweating artificially, so as to try and reproduce the sort of perspiration rates against which the products are likely to be used. This can be done in several ways, by emo- tional stimulation of the subject (5), by heat stimulation (6), or a combina- tion of the two such as encountered during strenuous exercise or even by pharmacological means (7). THE PURPOSE OF AN ANTIPERSPIRANT The aim of an antiperspirant product should be to keep sweating below the critical level where people are conscious of the problem and not neces- sarily totally to inhibit sweating. Since little is known about this critical level it is normal to measure either the reduction in sweating in absolute terms, as a percentage of the untreated response or to compare the effective- ness of two products directly, i.e. is product A better or worse than product B at redudng perspiration? The ultimate test of any antiperspirant must be the willingness of the consumer to buy the product and probably involves the user's subjective judgement about how effective the product is at stopping or reducing axillary wetness. Information on consumer preference is difficult to obtain and from overall preferences it is impossible to separate the individual product attributes into discrete units. The preference may be due to halo effects from perfume, effectiveness against body odour or non-functional attributes which tend to confuse the antiperspirancy evaluation. It is there- fore very important to the scientist that he should be able to isolate such a measurable attribute as antiperspirancy, quickly, efficiently and accurately. This communication shows how this can be done in the laboratory and how to avoid some of the misleading information that can come out of these tests.

EVALUATION OF METHODS FOR MEASUREMENT OF ANTIPERSPIRANCY 247 LABORATORY METHODS OF MEASURING EFFECTIVENESS OF ANTIPERSPIRANTS Any method that has been used to measure eccrine sweating can be adapted to the evaluation of antiperspirants. Some methods are more suitable than others and yet other methods have been suggested in which the antiperspirancy can be measured on model systems. The methods available can be roughly divided into four groups: (1) Model systems. (2) Visual methods. (3) Gravimetric methods. (4) Continuously recording methods. Model systems Various model systems have been suggested for the evaluation of anti- perspirants. These include the methods of Govette and de Navarre (8) included 'astringency' was measured by protein precipitation of egg albumin or shrinkage of frog skin. Other authors have described methods where sweat- ing was measured on the foot pads of mice (9), rats (9, 10) or cats (11). All these methods are open to criticism because the effectiveness of the products is not measured against sweating of the human sweat glands. In the case where 'astringency' is measured it is not clear what relationship there is between 'astringency' and antiperspirancy, and whilst there may be some relationship between them when aluminium salts are used, this pro- bably would not carry over to materials which operate by different mech- anisms such as anticholinergics or other anhidrotics. In the methods in which the materials are examined for their ability to decrease sweating in animal foot pads the same criticism applies, i.e. animal foot pads do not necessarily behave in an identical fashion to the human axillary sweat glands. In fact anatomically and physiologically they are much more similar to sweat glands found on the human palms and soles of the feet, and it is well known that these populations of eccrine glands are particularly insensitive to antiperspirants. The methods involving the use of models to measure antiperspirancy give particularly misleading results and as such have not found much favour amongst cosmetic chemists.