PRINCIPLES OF CORROSION OF METAL CONTAINERS 27 the metals present in the can, and the corroding environment. The effect of inhibitors on the solution/metal system can be assessed by polarisation resistance measurements (16). Some of the points mentioned in the paper are illustrated by the following examples. Fig. oe shows the bottom seam of a tinplate aerosol can. The position of the perforation within the seam of the can is in an area of restricted oxygen access, and it seemed likely that the failure was due to a differential aeration effect. This was confirmed by T-cell tests, when it was found that, although tin was normally anodic to steel in the solution, if the liquid round the steel was deaerated, the steel became anodic to the tin. When an inhibitor was added, it was found that deaeration of the steel no longer caused it to become anodic to the tin. It was confirmed in practice that use of the inhibitor prevented corrosion at the bottom seam. Fig. $ shows a section through the valve cup and can bead of an aerosol container. The cup was made from tinplate and the can from aluminium. It can be seen that perforation has occurred in the aluminium adjacent to the valve cup. It was found that in the product aluminium was cathodic to tin, but anodic to steel. It was therefore concluded that initially tin was dissolved until an area of steel was exposed. This then acted as the cathode for the anodic dissolution of aluminium. The mechanism was confirmed by the appearance of de-tinned areas on the valve cup opposite to the perfora- tions. CONCLUSIONS In this paper it has been possible to give only a superficial summary of the many aspects of corrosion which may occur in packaging. Much work remains to be done in areas where our knowledge is incomplete at the moment, especially in the use of inhibitors, and modifications to formula- tions. However, by use of some of the methods described above, we have been able to gain insight into several corrosion problems, and to make considerable savings in terms of container cost, and also size and duration of tests. (Received: 16th January 196'8) REFERENCES (1) Evans, U. R. The Corrosion and Oxidation of Metals (1060). (Arnold, LoneIon). (2) Pourbaix, M. Thermodyna•nics of dilute Aqueous Solutions (Translation by J. N. Agar). (1040). (Arnold, London). (3) Matsuda, S. and Uhlig, H. H. J. Electrochem. $oc. 111 156 (1064). (4) Pourbaix, M. Atlas of Electroche•nical Equilibria in Aqueous Solutions (1066). (Peragmon, Oxford).

28 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS (5) Stern, M. and Geary, A. L. J. Electrochem. Soc. 104 56 (1957). (6) Stern, M. and Weiserr, E. D. Proc. Am. Soc. Test. Mater 59 1280 (1959). (7) Iamm, G. G. and Willey, A. R. Corrosion 17 77 (1961). (8) Hancock, P. and 3/layne, J. E. O. J. Appl. Chem. Lond. 7 700 (1957). (9) Tomashov, N. D., 3/l[kha[lovsky, Y. N. and Leonov, V. V. Zh. fiz. Khim 85588 (1961). (10) Antropov, L. I. First International Congress on Metallic Corrosion 147 (1962). (Butter- worths, London). (11) Brasher, D. M. ibid. 156. (12) Rosenf'ld, I. L. ibid. 243. (13) 3/lindowicz, J. and Puchalska, D. Corros. Sci., 4 179 (1964). (14) Tomashov, N. D. Corrosion 9,0 7 (1964). (15) Koehler, E. L. Corrosion 17 93 (1961). (16) Legault, R. A. and Walker, 3/1. S. Corrosion 19 222 (1963). (17) Antropov, L.I., Corros. Sci. 7 607 (1967). (18) Neufeld, P. Corros. Sci. 4 245 (1964). Introduction by the lecturer The polarisation resistance method described is quite useful for assessing the probability of corrosion of metal wads in a product. We have used it to decide whether wads faced with tin foil or aluminium foil would be suitable for a product, and also to compare the effect of different formulations on the facing material. We have not used, or done very little work on, potentiostatic polarisation. This is acknowledged to be a powerful method for the study of corrosion problems. One reason is that potentiostats are rather expensive instruments, and another that they are mainly used for studying active/passive systems, such as the corrosion of stainless steel. They are very useful in such contexts but stainless steel is used very rarely as a packaging component, and the metals which are used in packaging usually do not exhibit active/passive behaviour in the environments in which they are used. DISCUSSION MR. E. K. CLARKE: I gather that most of your work has been done in essentially aqueous systems. How have you made measurements in non-aqueous systems, in particular alcohol and propellant? T•E L•CTUR•R: This is certainly a difficult subject, and the measurement of potential and pH is not easy. A very high impedance voltmeter has to be used for measuring potential and the one which we have has an input impedance of over 10 ' 5 ohms. We also have to be very careful in screening all our leads to prevent drift from stray current. In this manner we are able to take measurements of a compara- tive nature, and we do not pretend to be able to measure the absolute potentials of metals. What we do is to compare several different solutions and say whether one is better than the others. DR. J. J. MAUSNER: Can you suggest a quick way of finding pin-holes in metal containers ? THE LECTURER: Pin-holing of tinplate occurs when one has a solution in which tin is cathodic to iron. Under these conditions the only solution is usually reformulation. Luckily, although tin is usually above iron in the electrochemical series, due to the