270 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS is illustrated by the alkyl sulphates. The favoured chain length for the alkyl group in these compounds is C12, and the so-called lauryl sulphates contain dodecyl as the predominant alkyl group. They also contain appreciable amounts of the homologues with 10 and 14 carbon atoms which tend to modify the properties. Solubility is one property which is easily demonstrated. Commercial sodium lauryl sulphate, although initially soluble in water in low concen- tration, will usually throw out a deposit when even a 0.002 molar solution is kept for a few hours, whereas a similar solution of pure sodium dodecyl sulphate will remain clear and bright for weeks, or even months. In deter- mining the lauryl sulphates by titrating with a standard solution of a cationic, the effect of the different species present becomes noticeable. In a commonly used method, an aqueous solution of a cationic dye (usually methylene blue) is added to the anionic solution, followed by chloroform. On shaking the mixture, the alkyl sulphate forms a chloroform-soluble complex with the dye, and the blue colour becomes concentrated in the solvent layer. The addition of the cationic solution decomposes this complex, and the liberated dye mi- grates back to the aqueous layer, the end point usually being regarded as the stage when the depth of colour in the two layers is equal. The stoichiometric end point, however, does not necesssarily correspond with the visual end point. When titrating with pure sodium dodecyl sulphate the end point ob- tained does correspond roughly with the presence of equimolecular propor- tions of the two materials, but with increasing chain length of the anionic com- pound, less and less cationic is required to reach the end point. On the other hand, if we use anionics of chain length C10 or less, no definite end point is ob- tainable. It follows that in a mixture of anionic homologues the exact point of occurrence of the visual end point is a matter of some doubt, and the ques- tion arises how to standardize the cationic solution with which one titrates. A number of methods are in use for the standardization of cationic solutions. The British Pharmacopoeia uses potassium ferricyanide to esti- mate Cetrimide, the Toilet Goods Association of America uses potassium dichromate, and Lincoln and Chinnick 24 recommend the precipitation of the cationic as the phosphotungstate, which is then dried and weighed. The last method gives results which are independent of the chain length of the quaternary compound, but if ce•,l trimethylammoniumbromide is used as the standard cationic to determine alkyl sulphates with 12-14 carbon atoms, then it appears that either the phosphotungstate or dichromate method will give figures which are in reasonable agreement with those determined by methylene-blue titration against a standard anionic of the same chain length. Care should always be taken in the presentation of results that the precision implied is not greater than that actually attained, and in some

CHEMICAL ANALYSIS IN THE COSMETIC INDUSTRY 27I cases it is advisable to give the probable limits of error. Many young assistants (and some who should know better) err in this respect. It is not uncommon to see results given to two or even three places of decimals when the method employed makes it obvious that one can only be confident of the first place. The interpretation of results is an extremely important part of analytical work and a very fascinating one. In some cases, the real work of an analysis only begins after the bench work has been completed. It is not intended to convey the idea that the examination of unknowns is the only way in which the analyst can contribute to the cosmetic industry. This particular part of his work has been chosen because little seems to have been published about it. The importance of quality control analysis has already been referred to, and the analyst can frequently help his colleagues in the research and formulation sections by tracking down the offending material in faulty batches, by identifying and characterizing new materials, and by keeping a check on the efficacy of mixing, or the progress of a reaction when new products are formulated or new raw materials made. IReceived ß 15th April 1959j REFERENCES Smellin, C. F., Hartmann, L., and Stetzler, R. S. J. Am. Oil Chemists' Soc. 35 (1958) 179. Desbusses, J., and Desbaumes, P. J. Soc. Cosmetic Chemists. 8 (1957) 380, Franks, F. Analyst. 81 (1956) 390. Williams, 141. A. Oils, Fats and Fatty Foods--Their Practical Examination. 3rd Edn. (1950) 106. (J. & A. Churchill, London). Ravin, L. J., Meyer, R. J., and Higuchi, T. J. Am. Oil Chemists' Soc. 34 (1957) 261. Quinlan, P., and Weiser, A.J. Ibid. 35 (1958) 325. Newburger, S. H. J. Assoc. Offic. Agr. Chemists. 41 (1958) 664. Green, T., Harker, R. P., and Howeft, F. O. Analyst. 80 (1955) 470. Root, M. J., and Maury, M. J. J. Soc. cosmetic Chemists. 8 (1957) 92. Teitelbaum, C.L. Ibid. 316. Methods of Analysis of Offs and Fats. British Standard 684 (1958) (British Standard Institution, London.) Godbole, N. N., Fette u. Seifen. 43 (1936) 155. Toms, H., Analyst. 53 (1928) 69. Levy, G. B., and Fergus, D. Anal. Chem. 25 (1953) 1408. Seris, G. Ann. fals. etfrauds. 47 (1954) 29. Rosen, M.J. Anal. Chem. 27 (1955) 787. Morgan, P.W. Ind. Eng. Chem., Anal. Ed. 18 (1946) 500. Etienne, H. Ind. chim. beige. 22 (1957) 1175, 1287. Siggia, S. Anal. Chem. 30 (1958) 115. Dean, E. W., and Starke, D. D. Ind. Eng. Chem. 12 (1920) 486. Bidwell, G. L., and Stirling, W. F. J. Assoc. Offic. Agr. Chemists. 8 (1925) 295. Jones, J. M., and McLachlan, T. Analyst. 52 (1927) 383. Tare, F. G. H., and Warren, L. A. Ibid. 61 (1936) 367. Lincoln, P. A., and Chinnick, C. C. T. Ibid. 81 (1956) 100.