682 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS back than on the lower back, upper arm, forearm and thigh. With allergic reactions the variations were less marked but the response was greater on the back than on the arm, and least on the thigh. When patch testing with cosmetics, the cosmetic itself is applied and, if it is known, the ingredient which is likely to be the allergenic component. The tests are removed and read at 48 hr and read again at 96 hr this latter assessment is important, because it is not unusual for the test to become positive during the second and fourth day. Due to the enhancement of penetration by the closed patch test, a toxic reaction may be produced by an ingredient, which is not an irritant under the conditions of normal use. The appearance of such a toxic reaction may be a typical toxic soap reaction but it can also completely simulate an allergic response. To evaluate a doubtful reaction the open test is used, the preparation being rubbed on to an area of normal skin, usually on the forearm, three times a day for two days and the site is then exami- ned. The preparation is also applied to normal controls. A positive open test is considered to be significant, the cosmetic is fractionated and the patient tested to the ingredients. Tracing the allergenic, or toxic component of a cosmetic can be difficult. The patient may be reluctant to return for tests, which are only of academic interest it is occasionally difficult to ascertain the composition of a particular preparation or in some instances the patient may react repeatedly to the whole substance but to none of the individual components. Such a reaction has been reported by Rieger and Battista (3), who found that a positive reaction to a preparation could not be traced to a single component but on testing to a combination of the mineral oil and the sodium alkyl sulphate in water a positive reaction was obtained. The mineral oil was changed in the preparation, and no further irritant reactions occurred. Lipstick is the commonest cause of cosmetic dermatitis and may be diagnosed by the patient herself with some certainty. During the past six years, 38 patients have been seen with dermatitis of the lips due to sensiti- vity to eosin. Their ages ranged from 21-75 years occasionally the patient relates the swelling and soreness of her lips to the buying of a new lipstick but in general the condition develops for no obvious reason. Sulzberger et al (4) reported that it was the halogenated fluoresceins which are the usual sensitizers in lipsticks, and Calnan (5) confirmed their findings that the allergen is an impurity in eosin which can be removed by repeated crystallization. Patch testing with the patient's lipstick is unreliable for the detection of eosin sensitivity. Eosin binds with keratin and the amount present in an ordinary lipstick is too small to saturate the keratin and allow

CONTACT DERMATITIS FROM COSMETICS 683 eosin to penetrate through the horny layer. This is overcome by increasing the concentration of eosin to 500//0 in a standard lipstick base or in Vaseline, and by rubbing the lipstick or eosin stick directly on to the skin, thus ensur- ing that an adequate amount is applied. Patch testing of the patient's lipsticks is always done to detect sensitivity to a component of the lipstick other than eosin. It is noticeable that the incidence of lipstick dermatitis is decreasing whereas 29 patients were seen during the three years 1960-1962, only nine patients were seen in the years 1963-1965. This is probably a reflection of the gradual replacement of the halogenated fluoresceins. Wilmsmann (6) has reported the successful use, in lipsticks, of the FDC & DC azo dyes in the form of free sulphonic acids in this form, these dyes stain the lips and give a better range of colours than the eosins. During this period five patients were seen and found to react to com- ponents of their lipstick other than eosin. In two, the allergen was the lake D & C Red No. 31 (calcium salt of $-hydroxy-4-phenylazo-2-naphthoic acid) one of these was also sensitive to D & C Red No. 19 (3 ethochloride of 9-o-carboxyphenyl-6-diethylamino-3-ethyl-imino-3-isoxanthine). A third patient, reported by Calnan (7), was sensitive to the barium lake of D & C Orange No. 17 1-(2,4,dinitrophenylazo)-2-naphthol, and a fourth to the perfume in a lipstick. The fifth patient did not re-attend for further investigation. In 1961 a patient was seen who, knowing that she was sensi- tive to ordinary lipsticks, tried to use a "barrier" lipstick and developed soreness of the lips. On patch testing she reacted to the barrier lipstick and on testing to its components was found to be acutely sensitive to azulene (cyclopentacycloheptene). The azulene was presumably incorporated as an anti-irritant. Allergy to the aryl sulphonamide formaldehyde resin in nail varnish is the next most frequent type of cosmetic dermatitis that occurs. In this six year period, 33 patients were seen, their ages ranging from a girl of 15 years to a woman of 67 years. The face and sides of the neck are practically always the site of the dermatitis, in particular the eyes are often red and itchy. The nail beds are not affected because, in general, the nail-plates are impermeable to the lacquer. Patients are tested to their nail varnish, which is allowed to dry on pieces of lint before being applied to the skin, and they are also tested to the aryl sulphonamide formaldehyde resin, 10% in Vaseline. As the dermatitis is so remote from the responsible nail varnish, this condition can easily be missed if the clinician is unaware of the pattern of dermatitis which nail varnish causes. Patients never suspect their varnish and may be frankly sceptical of the possibility until proven