34 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS streams passing through chambers attached to the axillary skin. The moist gas streams passed into two separate moisture meters where electrolytic decomposition of the water converted phosphoric acid into non-conducting phosphorus pentoxide and the instruments indicated the changes in con- ductivity. The electrical signals representing changing moisture levels in •ach axilla were then passed to an x-y recorder, which plotted the two values and provided a trace of the axillary ratio. Retraces were run for 15-30 min and from them an accurate slope for the ratio of sweat in each axilla was obtained. FACTORS AFFECTING SWEATING A wide variety of factors influence sweating and, if meaningful sweat studies are to be undertaken these factors must be eliminated, or allowed for, in the test procedure adopted. Some of the more important factors are: (i) A high relative humidity promotes over-production of sweat (unless air movement is high) because impaired cooling of the body surface allows the internal temperature to rise (29) this is made use of to aid promotion of sweating in experimental studies. (ii) A variable time is required for an equilibrium sweating rate to be attained in some subjects several hours may be necessary (29). Measurements should not be made until the subject has been allowed time to equilibrate with his surroundings. (iii) Emotional or mental stimulation causes a rapid onset of sweating in thermal studies this may constitute an uncontrolled variable (29). (iv) The position o• the trunk during sweat collection affects sweat pro- duction in various parts of the body, including the axillae. Localized unilateral pressure will cause a reduction in sweating on the part of the body where it is applied and a corresponding rise elsewhere (29). This effect may occur if a bulky collection device is used in the axilla, or if the subject sits sideways in his chair. (v) If a subject falls asleep during the period of a test, perspiration may become quite erratic (28). (vi) Variations in sweat production are observable in different skin areas corresponding to variations in the numbers of glands on different parts of the body (61). Comparative assessments of materials must always be carried out on equivalent skin areas. Skin temperature has a definite localized influence on sweating a high local temperature leads to an increase in sweat output. Skin temper- (vii)

MEASUREMENT AND CONTROL OF PERSPIRATION 35 (viii) (ix) (x) atures over 33øC show this effect (t32). Techniques involving air flowing over the skin must not cause local cooling, which will have the converse effect of reducing sweat secretion. Under different conditions the same stimulus may elicit differing sweat gland responses this is referred to as 'conditioning' of the eccrine sweat gland (10). Environmental temperature is a 'condi- tioner' sweating induced by a heated, artificial environment, or by drugs is less pronounced in winter than in smnmer (131). There may be sex differences in the period required for sweat to appear on the skin surface following stimulation of the glands, although in the axilla similar amounts of sweat are found in both sexes (133). The period in females is longer than in males, and the proportion of males and females must be equal in different treatment groups. Metabolic rate, which is influenced by exercise, thyroid function, fever, intake of food, etc., affects the quantity of sweat produced (29). If tests are being carried out on a number of subjects over several weeks, it is always desirable to make collections at the same time of day on any one subject. Subjects receiving hormone therapy should be excluded from the tests. ANTIPERSPIRANT MATERIALS Until the introduction of an aluminium chlorhydrate complex (13), the most popular antiperspirant materials were aluminium sulphate and aluminium chloride. Since its introduction aluminium chlorhydrate has been, and still is, widely used. In aqueous solution the aluminium salts usually are acid, and this may prove irritant to skin and is certainly damaging to textiles in contact with them. Aluminium chlorhydrate yields a less acidic solution than the other aluminium salts. Formalin is an effective anhidrotic (used mostly on the feet), but owing to its presence in crease-resistant fabrics many people are becoming sensitised to it (11). Zirconium salts (11, 36) have not been in general use as antiperspirants since 1956 when it xvas discovered that they may lead to granuloma in the axillae this was considered to be an allergic reaction which could be potentiated by the use of soap and hexachlorophene (11). In 19139, Alphin, Vocac, Saunders and Ward (18) reported the results of tests, carried out on cats, to assess the effect of lignosulfonates on pilo- carpine-induced sweating. Two of the materials tested produced a signi-