502 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS usually, is, accepting human variability, whether in physiology or manual aptitude. Furthermore judgement scales need to be tuned to real-life situa- tions and the nature of the promise 'twice as much lather' must mean as used with water at the recommended or likely temperature and generated by the hand of the user. The development chemist on the other hand in seeking improvements needs to identify trends so as to know quickly whether he is going uphill or downhill. He needs quick, highly discriminatory tests which above all will tell him whether he has a hope. He needs controlled conditions, no varia- bility, replication, speed and, preferably, economy. A further distinction is that whereas the whole product evaluation is customarily performed by a group established for this purpose--an evalua- tion section, which should stand as independent from the product developers as Internal Audit from Accounts--the assessment of development progress is usually done by the development scientist himself. He can accept a degree of artificiality, this depends on how well he under- stands the system. These tests are sometimes very dependent on assump- tions of mechanisms. But if a toothpaste with caries-reduction properties can only be evaluated by a 3-yr clinical trial, how can formulae be optimized within a lifetime? Only by making one or more hypotheses on best possible scientific bases concerning the mechanism, thus giving rise to techniques for assessing the relative effectiveness of formulae by discriminating between, e.g. fluorine uptakes on enamel or reduction in acid solubility of enamel or changes in plaque composition. Such tests can be quite rigorously standard- ized but the real criteria of value lie not with the standardization, however good, but with the validity of the hypotheses. An apparently simple area, also in the dental field, is still a matter of dispute, that is the abrasivity of dentifrices. The very hypothesis that there is an optimum abrasivity rests itself on a well-backed, yet still sometimes contested, view that some abrasion is essential to remove stain plus a view that very high abrasivity is potentially damaging. But abrading what? Enamel? Exposed dentine ? Acquired pellicle? Nevertheless a measure of agreement on assessment techniques (4, 5) has enabled progress to be made and, when coupled with cleaning assessment-- another very difficult subject--has resulted in the creation of improved dentifrices. Such assessment techniques have enhanced value and impor- tance to-day when it may be necessary to introduce new raw materials to formulae and retain the product properties desired. There are many other types of assessment process lying in this area of

THE PROMISE AND THE PRODUCT 5O3 support to the development chemist. In fact proliferation of techniques is one of the problems. Quot homines, tot sententiae certainly applies here. There is also the problem of what happens to the Mark I technique when Mark II arrives unless the results are carefully correlated, earlier work becomes unrelatable and the temptation is always to continue with both. It is also an area of steady publication and I believe that something approaching a standard methods handbook could well be now timely if we could ever reach agreement. This is one of the declared aims of the I.F.S.C.C. (6) which has survived unchanged from the original draft put up by the US Society on 13th August 1958. Screening techniques Even simpler concepts based on even more comprehensive assumptions must be used to screen materials as opposed to assessing products. In this field there is greater inter-laboratory agreement and also, or because of, very many fewer problems of commercial security. I am thinking of such tests as foam tests for a detergent raw material, minimum inhibitory concentration of germicides, abrasivity of abrasives, Sward hardness of resins. Many of these tests are so standardized now that they will be handled by the raw materials section of the analytical laboratory and quoted by potential sup- pliers as part of their materials specification. Nevertheless there are some more specialized and local techniques in this category which will be deeply rooted in a locally-generated hypothesis, perhaps of a very fundamental nature and certainly a matter of commercial security, at least until the patents are solid! Summary Physical evaluation is thus largely a question of measurement of properties. Some of the properties we wish to quantify are easily measured and in some circumstances attention is mostly required to determine that it is indeed the right quantity that is measured. Viscosity measurement, for example, is not a technical problem, but selection of what method of measurement, under what conditions, does in fact measure the property most relevant to the attributes may be a difficult decision. It is the properties of the product-in-use that have to be quantified