52 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS The restraining collar was a thin (0.25 mm) card disc, 10 cm diameter, in which was cut a central hole to fit around the animal's neck. The disc was opened by a single radial cut and placed around the animal's neck. The cut disc was then stapled up slightly overlapping the cut edges to form a shallow cone similar to a large ruff. This type of collar was successful in preventing small rats (up to 150 g) from grooming the treated area for up to 12 h after treatment. The non-occlusive protective patch used in this study was similar to that described by Noakes and Sanderson (10). The treated area of skin was covered with a triple layer of surgical gauze approximately 1 cm larger in each direction to the treated area of skin. Over the surgical gauze a stainless steel gauze (100 mesh), approximately 0.5 cm smaller in each direction to the surgical gauze, was placed and 'Sleek' surgical strapping (Smith & Nephew Ltd, Welwyn Garden City, Herts), which had been punctured to give some 10 X 1 mm holes/cm a over the treated area, was wrapped around the animal. This has been found to be effective in preventing grooming of the treated area of skin for 2 days and for some animals up to 4 days after treatment. The effect of prewashing the skin on the penetration of the [x4C] soaps was examined by washing groups of rats either once or three times with a non-radioactive, 300 m• model soap solution (i.e. 60 m• solution of each of the five soaps studied). This soap solution (2 ml) was lightly lathered over the backs of rats for 1 min, left in contact for a total of 15 min, copiously rinsed with distilled water and dried with paper tissues. 2 h later the animals were treated with either 0.1 ml of the [•4C] soap solution over 7.5 cm • of skin or rewashed with the inactive soap solution twice, at 2 h intervals before treatment with the [•aC] soap solution as described above. The topically-treated animals were treated similarly to the injected animals described above. Before homogenizing the carcasses however, the protective patch was removed and the treated area of skin was excised and frozen between glass plates. Punch autopsies (1 cm diameter) from the frozen skin were monitored for •aC by solubilizing in 'Soluene' and counting. Further samples of treated skin were sectioned histologically for auto- radiographic analysis as described by Rutherford and Black (11).

PERCUTANEOUS ABSORPTION OF ANIONIC SURFACTANTS 53 RESULTS Penetration in vitro of [•4C] surfactants through human epidermis and rat skin A summary of the results from the experiments performed with isolated rat skin and human epidermis is presented in Table I. The results show no measurable penetration of SDS, SDI, DOBS or the C•s: 0 soap through rat skin up to 24 h after application, but 0.2 [tg/cm •' of the C•6: o soap had penetrated at 24 h. Some 7.5 [tg of the C•o: o, C•.: o and C•4: o soaps had penetrated per cm •' at 24 h but the results were not significantly different for the three soaps. For the three soaps which pene- trated the skin there was a lag time of 1 h before any measurable penetration occurred, but after this the rate of penetration steadily increased. At the end of the experiment, i.e. 24 h after application, between 60 and 70}/0 of the applied [•C] soaps and [x•C] SDI were rinsed from the skin and 30-40}/0 was associated with the skin. The [•4C] SDS and [•4C] DOBS were less easily rinsed from the skin as only 30•o was recovered in the rinsings and 70•o remained associated with the skin. The results from the human epidermis experiments showed no measurable penetration of the [•4C] DOBS and no measurable penetration of the [•C] SDS until 24 h after application when the rate of penetration was rapidly increasing so that at 48 h, 87.24-24.1 pg/cm •' had penetrated. The [x•C] SDI showed a steadily increasing rate of penetration up to 48 h. The penetration of the [•C] soaps from the model system showed different rates of penetration which ranked C•.: oC•o: oC•4: 0C•6: 0C•s: o. All of the surfactants which penetrated the epidermis showed increasing rates of penetration over the duration of the experiment which probably reflects the surfactant/ stratum corneum interaction and the breakdown of the barrier properties. This effect was most marked for the SDS where no penetration was detected during the first 8 h of contact. It should be noted that all of the epidermal samples showed some degree of swelling after 48 h contact but this was most marked with the SDS treated samples. The amount of [x4C] surfactant adsorbed to the epidermis was highest for the [•C] SDS where some 75•o of the applied'X•C was not removed by rinsing. For the other surfactants 30-50• of the applied x4C was retained in the epidermis after rinsing.