20 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS et al. that complete lack of volatility is necessarily to be preferred. A good cosmetic humectant should, first and foremost, exhibit a high degree of hygroscopicity, and particularly dynamic hygroscopicity, under the normal conditions of use. It should be of low volatility, and cause minimum viscosity changes with variations in temperature and water content. Other desirable features are: lack of toxicity, dermatological suitability, a wide spectrum of compatibility, reasonable cost, and good colour, odour and taste. It is the nicely balanced combination of these and other more general properties that has led such authorities as Kalish to state that "the humectant powers of glycerin are outstanding and account in large measure for its predominant position in cosmetic formulation." A great deal of useful information is contained in the now well-known paper by Griffin, Behrens and Cross (loc. cit.), but in view of the admittedly "erratic" character of some ot the recorded weight losses, it would be as well for chemists not to place too much reliance on them. Even ruling out the ever-present possibility of errors in manipulation or measurement, one is faced, in this admirable piece of supplier-sponsored investigation, with (a) data based on cream formulae that are themselves full of variables and (b) deductions that, while unexceptionable in themselves, do not fully present the story as told by the authors' own tables (i.e., of per cent weight loss in relation to the type and concentration of humectant present). I mention this because it has become almost a fashion to quote these authors' paper as the standard reference to the comparative utility as cosmetic humectants of glycerin, sorbitol and propylene glycol. V•lon pointed out in 1051 •' the high rate of loss of diethylene glycol from a 75 per cent aqueous solution when dried in an oven at 105 ø ½. and examined at intervals of 3, 5, 11, 16, 20, 28 and 36 hours. After the lapse of 36 hours all the diethylene glycol had evaporated. The loss of glycerin, exposed to similar conditions, was also high, but it was nevertheless much more gradual --ranging gradually from 25 per cent after 3 hours to a maximum of 42 per cent after 36 hours. It may be argued with good reason, however, as with so much published data concerning humectants, that this type of experiment bears little relationship to humectaat performance in face creams, where the hygroscopic agent is unlikely to be exposed to the air for long periods or to very high temperatures, and is usually present in proportions not exceeding 10 per cent of the total formula. Also, as Harry truly observes, •-' "the majority of cosmetic and toilet products are packed in air-tight containers, and the humidity inside the container is then purely a function of the temperature to which the container is subjected." In the tabulated results of their numerous experiments, Griffin et al. state that a soap-type cream containing 10 per cent of glycerin showed a loss of 0.24 per cent by weight after 1 hour at R.H. 50 per cent as compared with 0.55 per cent loss shown by sorbitol and 0.48 per cent by propylene

GLYCERIN AS A CONSTITUENT OF COSMETICS 21 glycol in comparable experiments. In a non-ionic cream 2 per cent of glycerin showed a loss of 0.40 per cent after 1 hour at R.H. 70 per cent as compared with 0.55 per cent loss with sorbitol and 0.72 with propylene glycol. After 24 hours at R.H. 70 per cent the non-ionic cream containing glycerin had lost only 5-34 per cent of its weight, as compared with 6.04 per cent for the comparable sorbitol-containing cream. I do not quote these results in support of any claims that can be made for glycerin as the cosmetic humectant of choice, but merely by way of comment upon the rather misleading generalisations put forward on pages 17 and 29 of the paper under reference. • • A further important aspect of humectancy that has not so far been touched upon is the humectant action of the film of cosmetic deposited on the skin. Glycerin has long been utilised in pharmacy, medicine and cos- metics for this purpose. The osmotic action of glycerin and dried magnesium sulphate is utilised, for example, in Morison's paste to withdraw water from boils and carbuncles. Glycerin is similarly used in cataplasma kaolini, where it also helps to prevent the poultice from drying out during storage. The traditional softening "emollient" action on the skin of diluted glycerin, as in glycerin and rosewater and glycerin-containing lotions and o/w creams generally, is, of course, largely due to humectancy. Some later published work of Irvin H. Blank • relates to the prevention or relief of dry skin conditions by the application of glycerin solutions, designed to "help attract water to the stratum corneum from the environment." The subject of humectancy cannot be left, however, without at least a passing reference to the fact that this property can sometimes prove unde- sirable in cosmetic practice if utilised to excess. Excessive concentrations can prove too searching to the skin. The concentrated material is rather unpleasantly sticky, though not so markedly so as sorbitol syrup. Excessive use of glycerin in creams and lotions may cause what Florence Wall •5 aptly calls "polka dot" spottiness in face powder or make-up films subsequently applied over it. Finally, it must be conceded that glycerin is not always to be preferred to the other humectants. In certain formulae the use of sorbitol, propylene glycol or even hexylene glycol may, for one reason or another, p•rove more satisfactory. Thus I have used hexylene glycolin in a liquid hand cleaner as a combined humectant and solvent. Bandelin and Tuschhoff have recommended the use of a mixture of glycerin and sorbitol as the most effective means of minimising evaporation losses in ascorbic acid syrups during storage. Many more examples of this kind will come to mind. Another important property of glycerin, from the cosmetic point of view, is its versatile behaviour as a solvent, coupling agent and vehicle. It is almost certainly the most cosmetically useful of water-miscible substances, being miscible in all proportions with both water and ethyl alcohol. It has limited solubility in ketones and esters. The miscibility of aliphatic and