258 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS with 2øC. is an adequate (although quite conservative) range. It is also desirable that the product should be able to withstand periods of exposure to It may be considered that the temperature ranges quoted are narrower than are likely to be encountered in practice, but it should be remembered that the tests involve stringent conditions in the form of regular, rapid temperature changes covering the extremes of the range. If, under such conditions, no phase separation, flocculation, or crystallization occurs for a one- to two-week period, then it is probable, but not certain, that the prod- uct will be stable under actual conditions of use. The fact has to be recognized, however, that, particularly in the case of emulsion products and other colloidal systems, the behaviour of a new product cannot be reliably established without a field trial. In any case, a new product should be studied by laboratory tests for Several months, as some types of changes (for example of consistency) can be much delayed. Further, with certain creams, such as lather shaving creams, a greater number of alternating temperature tests covering various ranges are neces- sary, as consistency changes are frequently specific to a particular range. MORE INDIRECT TECHNIQUES In addition to accelerated shelf tests, there are available more indirect techniques which may be regarded as rather specialised analytical proce- dures. Such tests are best illustrated by quoting examples. Emulsion Stability Subjecting an emulsion to a few minutes of centrifugal action is a useful method of assessing proneness to phase separation. In some cases of emulsion creams and pastes, particularly for comparing alternative formu- lations, dilution of an emulsion with an equal weight of water and observ- ing rates of sedimentation or creaming is a useful test. Emulsions may break down as a result of evaporation losses any such tendency can be assessed by a simple oven test using a tightly closed vessel. Closure The efficacy of closure of a jar or other container to prevent drying out can be examined by enclosing a suitable volatile solvent and determin- ing weight losses at intervals, and similarly, a desiccant can be used in the container to estimate efficacy to prevent ingress of moisture. Corrosion Many products packed in collapsible tubes involve a corrosion hazard, leading to breakdown of the tube, product, or both. Corrosion propensities and the efficacy of inhibitors, can be assessed by measuring corrosion po- tentials against a standard electrode employing a suitable potentiometer.

ACCELERATED STORAGE TESTS 259 Preservation The liability to microbiological spoilage can be explored by inoculation experiments with cultures and experiments with special media can be used to assess preservatives. TESTS RELATING TO INDIVIDUAL PRODUCTS Lastly, may be considered tests relating to individual products. Gener- ally such tests have very specialised application and for that reason are rarely, if ever, found ready made in the pages of a textbook. The true experimenter loves to establish a correlation between a parameter, which he can devise tests to measure, and an important property of his product. If this property relates to shelf stability, such special tests can save time, provide a basis for routine control of regular production batches of the product, etc., and also serve as an invaluable tool in the further investiga- tion and development of the product. Illustrations can only be quoted from within one's own experience such examples include: 1. A stability test based on dilution and a mass sedimentation end point for a bulk solid, stabilised, emulsion paste. A quantitative estimate of the effect on stability of changes in formulation led to a better understanding of the physicochemical system involved, and greatly increased the tempo of development. 2. An electrolysis test for aluminum collapsible tubes in connection with a product particularly prone to cause localised perforation of tubes. The test was correlated with the results of long term shelf storage trials and enabled a specification for tubes to be developed and incoming supplies of tubes to be tested. 3. A product comprising a clear solution was subject to crystalline deposition at elevated temperature. Investigation revealed that the de- posited material was a decomposition product associated with loss of car- bon dioxide. A test for dissolved material and gas content was introduced to minimise deposition. CoNcLtss•oN It may be considered that, in this short paper, a somewhat wide inter- pretation of the term "storage tests" has been taken. The scope does not go beyond the preliminary definition of "procedures to assess shelf life characteristics," however. The aim has been to be provocative, rather than comprehensive, and in this spirit I end with two problemsmone specific and one general: 1. Is there an accelerated test for syneresis? 2. At what point do accelerated tests break down for example, can exposure to 100% relative humidity be regarded as of any value for pre- dicting longer term behaviour at exposure to a 75% relative humidity ?