762 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Although the exact mechanism for the enhanced impedance of dehydrated skin to an alternating electrical current is obscure, Allenby et al suggested that the impedance measurements were an indication of the composition of the stratum corneum, the forces binding the components in the keratin and the state of hydration of the tissue. When the skins of rats fed on a low fat diet were examined, epidermal lipids were depleted, water binding was reduced and the impedance (A/C) and resistance (D/C) were increased. Histologically these skins exhibited a state of gross spongiosis (191). Measurements of the electrical resistance to a direct current or of the impedance to an alternating current may be carried out, either using excised skin samples, or in vivo on convenient areas of the skin of patients. Blank and Finesinger (181) used a simple direct current with zinc electrodes which were brought into contact with the human skin for long enough to obtain a reading for cutaneous resistance. The palms of the hands and the arms were suitable areas for this work. They found that the apparent skin resistance decreased as the size of the electrode increased. Also, the effective size of the electrode is not indicated by its overall size, but by the area of skin actually in contact with the electrode paste. A further problem en- countered in their work concerns the effect of liquid film on the electrode paste. This has the effect of increasing the overall size of the electrodes and thus leading to misleading results being given. The effect of the electrodes and the paste used to provide contact may be a source of error, due to their ability to increase the state of hydration of the skin. In examining the effect of topically applied substances on the electrical conductivity of the skin, it would also seem important to use subjects only after periods of rest. In this way the effects due to natural sweating could be minimized. In a study of the effects of organic solvents on the electrical impedance of the skin samples, an AC half bridge circuit was used (187). The electrodes were of stainless steel and constant contact with the skin was maintained using beryllium-copper springs. In this system, a reduction in the electrical impedance of the skin due to the action of some carboxylic acids, was present 1-2 h after treatment. With dimethyl sulphoxide, however, which greatly enhances the penetrability of the stratum corneum, a drastic reduc- tion was evident after only a few minutes. Although the changes in the electrical impedance or resistance of skin treated with solvents or other substances may give some information on the extent of damage present in the stratum corneum, at present the value of the method is open to speculation. Until now only limited experience has been gained in using them and no positive conclusions are possible as to

APPRAISAL OF METHODS FOR DETECTING PRIMARY SKIN IRRITANTS 763 the correct interpretation in terms of tissue damage. The conclusion reached by Allenby et al (187) that the impedance of the skin varies according to the permeability to water, suggests that probably the techniques are of greater value in detecting changes in the permeability of the skin due to topical applications of substances, rather than to detecting damage which is not evident to the naked eye. Collagen swelling in vitro Choman (192) suggested that since normal skin contains approximately 70}/0 collagen, changes in dermal collagen might be useful in evaluating the response of the skin to irritants. He exposed discs of animal and human skin collagens to a series of alkyl sulphates in vitro and noted changes in the thickness of the discs and their ability to stain with indigo carmine dye. With these surfactants, consistent changes were evident after 24 h immersion. Swelling or contraction of the discs varied according to the nature and concentration of the chemical and the temperature. Choman (192) regarded the magnitude of swelling as a measure of the conformational changes in the structure of the protein. Sodium lauryl sulphate, which showed the greatest effect in this in vitro assay, was also the most irritant of the agents in in vivo studies on human skin (193). It would seem that a correlation exists between the magnitude of collagen swelling and absorption of the surfactant at the protein-solution interface. The tissue changes would be a reflection of the weakening or rupture of hydrogen bonds and salt linkages which are largely responsible for holding together the collagen fibres. In these studies the staining characteristics of the collagen discs with indigo carmine correlated well with the swelling observed. Greatest dye uptake was observed in discs exposed to surfactant solutions which produced greatest swelling. It was not found possible to remove the dye taken up for quantitative measurement of the tissue responses. In a subsequent study, Choman (194) used skin collagen swelling to compare the effects of aqueous anionic, cationic and nonionic surfactants. Whereas anionic soaps caused swelling and inhibition of dye uptake, cationic and nonionic agents caused contraction. Where swelling was less than 50•o structural changes were not evident histologically. The lack of tissue changes with nonionic surfactants is consistent with Putnam's view that nonionic agents are not reactive with proteins on account of the absence of strong polar groups on the molecule (195). However, this concept would not seem adequate to explain the lack of