28 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS solution the grey-blue to blue-black colour reaction represented sites of active sweat gland secretion. A paper which had been in contact with the sweating skin was treated with a 10% silver nitrate solution and the resulting precipitated silver chloride, when reduced by uv light, provided a picture of the sweating pattern {41). Plastic impressions and replicas Impressions of the skin have been obtained by use of a 2-4% solution of polyvinyl formal in ethylene dichloride containing 1% butyl phthalate as plasticizer. Viewed microscopically, holes seen in the plastic film corres- ponded to the sites of sweat duct orifices {42). Sarkany (43) used Sifio Dental Plastic to produce a skin impression. Metal shadowing of a replica taken from the impression allowed the sweat duct orifices to be seen under the microscope. Alternatively an impression obtained in this manner has been treated with silver nitrate: reduction of the precipitated silver chloride made the sites of sweat secretion visible (45). H ygr op hotogr ap h y This is a system based on the reversible colour changes which a silver- mercuric-iodide salt undergoes when exposed to the consecutive effects of light and humidity (44). An orthochromatic photographic emulsion is treated successively with a metol-hydroquinone developer, sodium thio- sulphate, mercuric chloride and potassium iodide. The end result is a yellow film which is then blackened by exposure to light. If this blackened plate is put into contact with sweating skin a reversible colour change occurs and the sweat droplets produce a yellow pattern. GRAVIMETRIC TECHNIQUES Visual methods of assessing sweat secretion are both useful and adequate for evaluating numbers of glands, or for assessing antiperspirant products where an 'all-or-nothing' effect is sought (as in the case of anticholinergic agents). However, most of the available materials only partially inhibit sweat production and often relatively fine distinctions need to be made between experimental products: in these cases fully quantitative techniques are required.

MEASUREMENT AND CONTROL OF PERSPIRATION 29 Capillary collection methods These methods have been described (5, 4t3) but are really only of interest if small quantitites of sweat are needed for analysis, as only single glands are studied. Collection chambers Collection devices attached to the skin have been used by several workers. Basically, these consist of a ring, with a flange which supports a disc of pre-weighed filter paper {2.5 cm diameter) and a close-fitting cover to prevent evaporation of sweat absorbed by the paper. The open side of the chamber is held in contact with the skin by straps, or by cement (Duco Household Cement, Durofast, Weldwood Contact Cement). If cemented to the skin the collection device can remain in position for 2-3 days. Quantities of about 0.01-0.1 ml of sweat can be absorbed on the filter paper discs and collections are usually made over a period of 10-30 min. Schwartz, Thaysen and Dole (47) used an aluminiun unit in preference to an earlier plastic unit which warped on repeated usage (48). One of the prob- lems encountered with this method is leakage of sweat into the collection area from the surrounding areas Sato and Dobson (40) overcame this by the design of a chamber which had a knife edge to contain sweat in the collection area and drainage grooves in the base to drain sweat in the sur- rounding areas away from the collection site. Such methods, whilst being convenient as a means of collecting sweat for chemical analysis, present problems if considered as routine test methods. Ideally the axillae should be utilized for collections as this is the site in which the antiperspirant product is expected to exert its inhibiting effect in practical usage. The shape of the axillary vault makes the attach- ment of collection devices difficult and the time required for a cement to dry may be as long as 20 min. Absorbent pads A gravimetric method which has found fayour (29) on account of its simplicity of operation is the 'pad' method. Tared, absorbent cotton pads, about 65 X 140 mm in size, fit comfortably into the axillae and the quantity of sweat produced in a given time is readily determined. MacMillan, Relier and Synder (26) left the pads in position for 10 min and then immediately transferred them to tared, covered, containers and placed fresh pads in the axillae for a further 10 min. Control sweat collections were generally made