7O JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS meaningful full statement of the composition of a cosmetic can only rarely be made, for the addition of perfume to the cosmetic generally adds a complex of many known and many unknown chemicals, making a full statement of composition well nigh impossible. This example has been cited to point out to the reviewing scientist that proposed legisla- tion must be carefully, adequately and completely scrutinized. A New Cosmetic Application, according to proposed legislation, would call for a complete description of the manufacturing process. This would include the kind and quality of ingredients used a detailed description of the manufacturing process, including the packaging and labaling details of the control procedures employed to assure that the ingredients meet specifications, that the manufacturing process is ade- quately controlled, that procedures are employed to prevent mislabeling of the product, and finally, that the finished product is subjected to ade- quate control procedures. Provision should be made so that minor changes in manufacturing, packaging and control procedures would not necessitate amending the petition. It certainly would be particularly burdensome if every time one made a change in processing, control methods or labeling, he would have to wait for an amended petition to be approved. As proposed, a New Cosmerle Application would require the sub- mission of data establishing the safety of the cosmetic, when used as directed or in the customary or usual manner. Certainly the cosmetic chemist developing a new product wants to know all about the physical, chemical and toxicological properties of the various ingredients he in- tends to use in formulating this new cosmetic. He knows that the finished product must be subjected to safety evaluation. It is, of course, a fallacy that one can make a list of hundreds of approved cos- metic ingredients and assume that any and all combinations of those ingredients will result in a safe product. The safety of a new formula- tion can only be assured by evaluating the product. This will undoubtedly result in pharmacological evaluation of the product on animals, followed by clinical evaluation on man. The tests to be employed will be governed to a large extent by the nature of the product and its intended use. The final test, in my opinion, should be carefully planned sales in a very limited geographical area. May I say a few words on the methodology of safety evaluation. As you know, I am supervising a pharmacological evaluation program covering some 25 certified colors and which has to date, cost about half a million dollars. Four species are involved: mice, rats, rabbits and

FIFTEENTH MEDAL AWARD 71 dogs. A detailed discussion of the methodology cannot be gone into here, but suffice it to say research should be undertaken in an effort to develop new methods that are shorter, more meaningful, and less costly. Methods that will permit us to forecast with much greater accuracy the effect that may be expected in man, not excepting carcino- genesis. We need better methods for determining the effects of ma- terials topically applied to the human skin. We need more cutaneous toxicity conferences like that held last October in Washington, D. C., and more research that makes such conferences possible. Before leaving a discussion of safety evaluation, I would like to point out a factor which is sometimes overlooked in assessing the "in use hazard." I refer to exposure. Exposure data are very significant for, if exposure to the substance under consideration is minimal, then the effect is minimal. In order to determine the exposure of the consumer to trace materials in cosmetics, it is necessary to determine the per diem use of several types of cosmetics. The experimental method may be summarized as follows: Weighed packages of the cosmetic are given to the testers with instructions to use it in their customary manner for one week. The loss in weight is determined and is regarded as having been applied to the person of the user during the test period. This is actually loss from the container and disregards spillage, etc. .Per diem use so determined may be regarded as maximum. In these studies the individuals chosen were normal users of the cosmetics to be tested, and they were supplied with the shades they were currently using. We were interested only in maximum exposure therefore, we took the maximum per diem use and not average use. For example: of the 222 women who participated in the lipstick study, we considered as the maximum per diem use the one person who used 73 rag. of lipstick per diem. We assumed that the lipstick contained 12.97% pure dye (about 1.5 times the average). We also assumed that she ate all she applied to her lips. It may be calculated that about 6 parts per million of her daily diet would be color. We have found exposure data extremely useful in evaluating the hazard that may arise from the use of cosmetics. I will not go into the details here, but perhaps a few examples are in order. Let us suppose there is under consideration a limit of one part per million of mercury in color additives. We are able to show that in the ease of a lipstick containing 43% color additives about 0.015 rag. of mercury would be in- gested per year, and that in the case of a dusting powder which we will