510 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS each arm, and covering one arm with an occlusive dressing while the other arm was protected by an elevated, perforated (non-occlusive) guard, it was found that the occluded arm showed vasoconstriction with glucocortico- steroid concentrations 100 times less than those required to give equivalent vasoconstrictor responses on the non-occluded arm. A similar conclusion was reached by Scoggins and Kligman (172) in seven patients with dermatitis, using plasma cortisol levels as an index of' the suppression of' pituitary adrenal axis by the percutaneously absorbed synthetic steroids. The authors applied a weighed amount of a commercial cream containing asteroid mixture of known composition (prednisone, triamcinolone, fluocinolone acetonide, methylprednisone acetate and hydrocortisone) on areas of dermatitis and covered a number of these areas with an occlusive dressing, the rest were left uncovered. Under these con- ditions absorption of corticosteroids through diseased skin covered by an occlusive dressing is considerably higher than that from similarly affected but uncovered skin. The authors observed that: 'without an occlusive dressing, systemically significant amounts of these substances are absorbed only if the dose applied is very large'. Occlusion of the site of application in humans was found by Feldmann and Maibach (173) to increase 10-fold the amount of hydrocortisone absorbed. In these experiments, occlusion of the site increased the percutaneous absorption irrespective of the type of vehicle employed. No comments were made as to the possible mechanism responsible for the increased absorption from the occluded site. Covering the skin with an impermeable dressing prevents evaporation of sweat and of insensible perspiration resulting in an increased hydration of the stratum corneum. One would presume that this increase in the water content of the stratum corneum is responsible for the increased absorption in accordance with the results of the other authors already mentioned. The site at which water is retained in the stratum corneum is not clearly defined. Recent work (174) suggests that it is retained in the intercellular space in association with macromolecules possibly of a mucopolysaccharide nature. SPECIES VARIATION The rabbit, guinea-pig and rat are the most commonly used species for studying percutaneous absorption but the mouse and pig are used on occasions.

PERCUTANEOUS ABSORPTION 511 According to Draize (23) 'the rabbit is the most popular choice because it is generally more susceptible than man to the action of most substances'. This opinion was also expressed by Brown (34). He considered the rabbit more sensitive than other species and cited the work of Stevenson (175) with Telodrin in support. The percutaneous LD50 for a xylene solution of, Telodrin was 25 mg kg 4 in rats but only 5.8 mg kg 4 in rabbits. There are however, important exceptions to this general rule. For example, the per- cutaneous LD50 for parathion is lower for rats than it is for the rabbit (176). Because of their low toxicity, it is doubtful whether the LD,0 estimation is applicable to the cosmetic chemicals commonly in use. In any case, the wide difference in species sensitivity to toxic substances makes the LD,0 an unreliable index of percutaneous absorption. The experiments conducted by McCreesh (48) in which he compared the sensitivity of a variety of species to a toxic dose of two organophosphorus compounds with their passage in vitro through the epidermis of the same species serves to illustrate this point. McCreesh (48) determined the percutaneous LD50 of two organo- phosphorus compounds (formula not given) in the rabbit, pig, dog, monkey, goat, cat, mouse and rat. He found that the rabbit and cat were the most sensitive followed in order by the dog, the goat, the monkey, the mouse, the pig and the rat. The dog, goat and monkey were about two to four times less sensitive than the rabbit and the cat, the mouse about six times, and the pig and rat about 10 and 20 times respectively. Measurement in vitro of the penetration of the labelled organophosphorus compound through the excised skin from dorsal thorax of these eight animal species revealed that there was no correlation between penetration and toxicity. The fastest rate of penetration was recorded in the case of rabbit and the rat, followed by that of the guinea-pig, cat and goat, monkey, dog and slowest of all, that of the pig. If one assumes the diffusion rate across the pig's skin to be unity, the rates of diffusion of the skin of the respective species are 9, 18, 18, 18, 27, 35, and 35 (rabbit and rat) times that of the pig. For an objective assessment of the rate of absorption, in vivo tests are essential although valuable data may be obtained from in vitro work. A number of authors have conducted comparative tests of this sort. Employing the technique of measuring radio-activity in blood, urine and selected internal organs, N6rgaard (177) found that when 10 [tl of an aqueous solu- tion of*7Ni was applied to the shaved skin of rabbits, and guinea-pigs, the degree of radio-activity after 24 h in the blood, kidney, and liver of the rabbit was approximately 5, 10 and 2 times those of the guinea-pig respec- tively. When 56Co was applied in the same way it was found that the radio-