102 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS particularly to the aromatic and cosmetic industries, should not suffer from fundamental neglect. In order to carry out gas-liquid partition chromatography efficiently, we have to bear in mind the fact that the principal constituents of essential oils and almost all aromatics are polar substances, and that most of them are able to form associations by means of hydrogen bonds. Therefore, analogous substances are to be preferred for the fixed phase. Even if it is possible to work at pressures going as low as 40 cm. of mer- cury at the inlet and 10 cm. at the outlet, by using sufficiently sensitive catbarometers, the working temperature should lie between 120 ø and 220 ø. This limits considerably the choice of the fixed phase. At such temperatures, most of the heavy substances generally used in gas-liquid partition chromatography have to be removed. Thus the n- octyl, n-nonyl, isodecyl and tetrahydrofurfuryl phthalates are no longer suitable above 150-170 ø, nor are octadecyl stearate or orthocresyl phos- phate at slightly higher temperatures. The safety margin for using these products is still further lowered by the use of certain carriers. Certain polyethylene glycols or polypropylene glycols having a mo- lecular weight between 800 and 1000 or their esters (e.g., polyethylene succi- nate), Or their ether derivatives resulting from partial methylation, can be used from 180 ø to 190 ø, and alkyl (or aryl) siloxanes can be used up to 220 ø . At such high temperatures, however, most of these give off gases during the first few hours of use. After six to eight hours this drawback disappears and the columns can be used for a long time. Silicones of diverse origins (Silicone Elastomers DC 123 or DC 152, Silastomers DC 132 and DC 156, Wacker's Silicone gum DC and Apiezon M grease) can be used. It is advantageous first of all to wash some of these products with ethyl acetate. In the case of nonpolar substances, the polythenes of molecular weight 30,000 to 40,000 or perhydrosqualene may be used. The use of ph thalic or succinic esters or of some silicones is to be avoided in the case of easily dehydrated constituents of essential oils or aromatics. The choice of the carrier is also of great importance. Diatomaceous earths lower the temperature at which the esters used as fixed phase are altered, and give rise to dehydration, condensation or other reactions of the constituents under examination. The same thing happens with the powders obtained by crushing refractory bricks (e.g., Sterchamol). Sea sand of the size between sieve 44 and sieve 52 (British standards) is often preferable, in spite of its poor retention of the fixed phase. With glass powders this disadvantage becomes too grea. t• Sodium chloride passing through sieves 30 to 40 can be used, as it retains 5 per cent silicone grease. The reactions which are apt to occur at the resistances in the catharom- eter must not be neglected. Thus dehydrogenation of the alcohols has

LABORATORY METHODS FOR PRODUCT TESTING 103 been observed at as low as 140 ø in thermal conductivity cells with a plat- inum wire. The phenomenon may be observed through the appearance of negative peaks. In any case it is wise to test a reference sample under the same conditions befbre questioning the purity of a product on the basis of gas chromatog- raphy. LABORATORY METHODS FOR PRODUCT TESTING* By DONAI. D H. PowEP, S Lambert-Hudnut Division, F//arner-Lambert Pharmaceutical Company, Morris Plains, N. •. PERHAPS THE greatest need for the cosmetic industry and certainly the greatest need of the cosmetic chemist is for accurate methods for evalu- ating cosmetic products. A recent advertisement for a new cosmetic prod- uct certainly cost at least $75,000 for the single advertisement and yet nowhere in that ad was there any clear indication or measure of what man- ner and to what extent the product performed. In the defense of the adver- tising agencies we must point out that they have turned to glamor and indefinite, vague promises of beauty only when the chemists could give them no accurate data and information on performance. LABORATORY TEST METHODS IN OTHER SOCIETIES The importance of laboratory test methods in the role of a chemist is best demonstrated by reviewing the latest Yearbook of The Textile Chem- ists. In the latest issue you will find that they have developed or are cur- rently working on 83 different and distinct laboratory methods for evaluat- ing textile fibers and fabrics. This number is growing each year and hun- dreds of chemists are cooperatively working on them. I am sure you are all acquainted with the excellent publications of the American Society for Testing Materials whose entire object is to establish reliably accurate methods for testing all types of materials from steel to cork and from soap chips to lubricating oils. At some future date we should work with A.S.T.M., but we must, first, outline, indicate and study the basically important prob- lems. The justly famous Marburg lectures of A.S.T.M. are classics in reviewing the progress and problems in evaluating and testing products from con- crete to plaster. It is still years ahead before such a review should be con- sidered in the cosmetic industry. * Presented at the September 19, 1957, Seminar, Chicago, Ill.