260 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS plus reaction on the challenging application was not regarded suflqcient to designate allergic reaction. Moreover, prior reactions were of the same grade. Consistently, re-test two weeks later was negative. The fatiguing or dermatitis potential of some waterless hand cleansers in actual use is in relation to degree and duration of exposure and other circumstances already discussed the presence and amount of solvent usu- ally deodorized kerosene (contained in some). It is interesting that five subjects showed poor tolerance. The result of this study was more favorable than patch test study of several waterless hand cleansers reported by Goldman, Preston and Scheen (i3). Goldman observed primary irritant action from 8 per cent dilution and some allergic reactions. Table 3: Since sorbic acid is used as a bactericide 0.1 to 0.2 per cent in some cosmetics it was desired to know if it is allergenic and in what con- centration does it exert (under the test conditions) primary irritant action. On account of sparing solubility in water, hydrophilic petrolatum U.S.P. was selected as the vehicle for testing. Pilot study suggested using one per cent concentration. Forty-nine of 50 subjects showed no evidence of primary irritant, fati- guing or allergenic action. Only one subject reacted on the third applica- tion for this individual one per cent concentration is a potential primary irritant, this is doubtless consistent with results (in pilot study) using 2 and 4 per cent. These results suggest that in actual use, irritant potential is above one per cent and that there is a wide margin between its use con- centration in cosmetics and irritant potential. Table ¾: This study on 100 subjects of an antiperspirant fbr ten consec- utive applications showed unfavorable results. Fatiguing reaction ap- peared early and for some increased in intensity (Number 94 in table) this necessitated moving the site of the next application immediately ad- jacent to the prior one. A negative reaction at this new site (Number 94) opposed the 2 plus prior reaction as an allergic one. Other untoward reactions were folliculitis and dry wrinkled skin as already discussed and shown in Fig. 4. These reactions invariably eventuated into frank pri- mary irritant reactions. Negative reactions on challenging applications did not suggest allergenic action of the tested substance and was consistent with the aforementioned additional evidence. Table 5: Study of this antiperspirant was also unfavorable. The untoward cutaneous reactions were similar to the one studied in Table 4 although the incidence was less and they appeared later in the course of the consecutive applications. In addition, the study showed allergic reac- tion (subjects 2, 28, 48). Subject Number 44 is an instance where fati- guing reaction is not too clearly defined from allergic reaction. This necessitated re-test some weeks after challenging application. This was

INSULT PATCH TESTS IN STUDY OF CUTANEOUS REACTION 261 negative and opposed allergic interpretation of the 2 plus reaction of this subject on challenging application. Table 6: Study of this insecticide in contrast to the preceding tables showed negative reactions up to suddenly appearing, strongly positive reactions on the twelfth application. The reaction of one subject is shown in Fig. 2. Additional evidence of allergenic action was the delayed ap- pearance of the reaction about forty-eight hours after removal of the patch. On subject Number 37, the site of the twelfth application was adjacent to that of the eleventh application the site of the latter "flared up" concomi- tantly with the positive reaction of the twelfth application. This evidences high degree of allergenicity. Consistently, re-test (conservatively per- formed) of one subject two months after challenging application showed delayed 2 plus reaction to three drops (open patch test) on the arm of the subject. Prior tests were conducted on the back. SPECIFICITY OF PATCH TEST STUDIES It would be desirable to know the comparative value of the aforemen- tioned methods in determining allergenicity. I know of no such reported study. An approach to this was reported by Traub, Tusing and Spoor (2). They studied three lotions and three ointments by their aforegoing described method and also by the prophetic patch test but not to its second phase---usage test. They stated the opinion that this phase has been deemphasized in practice. On of the ointments showed allergenic action that was not shown by the prophetic patch test. The allergic index (incidence of allergic reactions) of a known cutaneous allergen in terms of percentage in an X number of persons is not known. It can be extrapolated from a limited amount of data. As Rosentberg pertinently stated, there are here two variables. A biological variable and a statistical variable. The two are independent. Knowledge of the incidence of allergic reactions of known cutaneous allergens is gleaned from experience, from the literature and frequency of its contact with the skin. This is expressed in comparative terms such as, very rare, rare, not common, frequent. On this basis one can estimate that, for example: nickel is more allergenic than dichromate the latter more allergenic than mercury paraphenylenediamine more allergenic than all of the aforementioned. In discussing specificity of patch test study prophetic test employing 200 subjects and the repetitive test employing 50 to 100 subjects, I would estimate that such tests would begin to detect allergenicity at about the midregion of a scale at one end of which allergenicity is very rare (lanolin), at the other end it is frequent (paraphenylenediamine). Detection of allergenicity would increase in ratio to the increase in the allergic index of the tested substance.