RAPID ASSESSMENT OF COLOUR FOR ROUTINE STORAGE TESTING (2) Judd, D. B. and Nimeroff, I. in Kolthoff, I. M. and Elving, P. J. Treatise on Analytical Chemistry õ 2873 (1964) (Interscience, London). (3) Kornerup, A. and Wanscher, J. H. Methuen Handbook of Colour (1965) (Methuen, London). (4) Arny, H. V. Proc. Intern. Congr. Appl. Chem. g{I $19 (1912). (5) Arny, H. V. and Ring, C. H. J. Franklin Inst. 180 199 (1915). (6) Arny, H. V. Ind. Eng. Chem. 8 309 (1916). (7) Collins, J. F. Brit. J. Appl. Phxs. 1{1 527 (1965). (8) Henderson, S. T. and Hodgkiss. D. ibid 14 125 (1963). DISCUSSION MR. K. McLAREN: The measured values of the colour temperature of northerly skylight reflected from a matt white laminated plastic are very much lower than one would expect. Collins obtained a mean value of 6,950øK in Teddington (7) Hender- son and Hodgkiss 7,290øK in Enfield (8). There are severa! possible reasons for this discrepancy of which the most likely one is probably the reliability of the colour temperature meter. We have often found identical meters varying by more than 500øK and have consequently adopted the practice of having them checked at frequent intervals. There is also a danger in assuming that a correctly calibrated meter will be reliable when used to measure the colour temperature of a combined source such as F221/45, one component of which is a fluorescent tube, because its spectral power distribution curve departs markedly from both a black-body radiator and from day- light. B.S.I. are shortly to publish a revision of B.S.950, Artificial Daylight for Colour Matching and Colour Appraisal, and fluorescent tubes conforming to it have been available since last summer. The standard is based on the recently adopted C.I.E. source D.6,500 (colour temperature of 6,500øK) which is universally accepted as the most representative of average daylight tubes conforming to it must surely be ideal for judging cosmetics. MR. J. D. CHESHIRE: We are quite aware of the limitations of colour temperature meters, and we are actively considering purchasing an emission spectrophotometer, so that there can be no mistake. However, for the colour temperatures that we reported, although they might have been rather low because of the inaccuracies in the colour temperature meter, we were concerned only with matching the colour in the testing laboratory which had been used hitherto. We did not want to change the light rendering in the middle of a series of experiments, and the temperatures we quote were for this particular sampling room, which faced north. There were surrounding buildings and the colour of the walls was, perhaps, not ideal for this sort of work. That probably explains the variation in colour temperature of the illumination we were using and of northerly skylight. I am glad to hear about this new light source that is being arranged by the B.S.I. because, after we had published this paper, we found out that the A.E.I. colour matching unit is no longer available. However, no source of light can surely be ideal for all colour matching of cosmetics, which are viewed under so many conditions. You have to face the fact that we have to view it in daylight, fluorescent, sodium and tungsten light, and I think these sources are going to be with us for a long time. Although this standard of 6,500 ø has been adopted, and one must adopt a standard, I really do not see that this is going to solve our problems.

16 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS MR. K. McLtR•,N: Your observations are absolutely true, and this particular standard emphasises this need. For the first time in a British Standard the word 'metamerism' is mentioned and defined so that any colour user will know exactly what it means. The virtues of this tube are solely that it does give average daylight, completely reproducible, any time and anywhere you want it, but it is only an imitation of average daylight. If you want to assess not only colour matching but colour appraisal, then clearly you have to assess it under entirely different kinds of illuminant. Tubes conforming to the B.S.950:1966 are made by British Lighting Industries Ltd. and sold as "Artificial Daylight" by both A.E.I. (Mazda) and Thorn (Atlas). They are at present available in 65/80 watt 5ft bi-pin tubes. This tube will be a marked step forward for colour viewing, both for matching and for appraisal, and of course, what is frequently overlooked is that daylight itself is extremely variable. MRs. H. BrJTL•,R: What is the life of this tube? I understand that our ordinary tubes darken in a very few months, and how would one test that it was still the standard? MR. K. McLAR•,N: The B.S.I. committee recognised that there is no method that the ordinary lamp-user can use to see if the spectral distribution of his lamp is still what it was when he installed it. The British Standard requires that the manufacturer will state the number of hours for which the tube can be used before its emission is outside the specification. The commercial varieties of the tube at the moment have a guaranteed life of 9,,000 hr, and it is known that after 9,,000 hr the only significant difference is in the uv output, which is important if you are assessing fluorescent compounds, as in the textile industry. But if you are just judging non-fluorescent colours you probably can use these tubes until they start flickering. MR. J. D. CH•,SHIR•,: We use a xenon arc lamp for exposing our samples. One of the reasons that we want to get an emission spectrophotometer is so that we can check the change in spectral distribution of this particular lamp with time. Miss A. E. YOUNG: I was interested in your comments abouts the fading of a product with one colour. Would you care to comment, considering that frequently most of the cosmetic products we use contain at least two, or often three or four different colours? Perhaps your method may not be quite as straightforward as it seems. MR. J. D. CH•,SHIRE: Our method will not differentiate between 2 dyes. However, this method was only intended to see whether there was any deterioration. Then it went back to the physical chemist to find out which dye is fading, and why. The tests are put on with the whole product and one dye, and the whole product and another dye, etc. This is only intended to be an assessment, a screening test if you like. Mr. J. C. BROWN: Could I return to a point raised by Mr. McLaren, regarding the stability of colour temperature of light sources? A colleague has recently been doing some ranges of dyeing, following a paper in J. opt. Soc. Am. He had two ranges of green dyes which varied in shade from a yellowish to a bluish green. One set was done with dyes which changed very little in artificial light, and the other with dyes which changed very rapidly. Apparently by matching up the two scales he was able to make a judgment of the colour temperature of a source of light.