PRESERVATIVES FOR PHARMACEUTICALS 719 (28) Hibbott, H. W. and Monks, J. Preservation of emulsions--p-hydroxybenzoic ester partition coefficient. J. $oc. Cosmet. Chem. 12 2 (1961). (29) Bean, H. W., Heman-Ackah, S. M. and Thomas, J. The activity of antibacterials in two- phase systems. J. $oc. Cosmet. Chem. 16 15 (1965). (30) Bean, H. S. and Heman-Ackah, S. M. Influence of oil: water ratio on the activity of some bactericides against Escherichia cell in liquid paraffin and water dispersions. J. Pharm. Pharmacol. 16 Suppl. 58T (1964). (31) Bean, H. S., Kenning, G. H. and Malcolm, S. A. A model for the influence of emulsion formulation on the activity of phenolic preservatives. J. Pharm. Pharmacol. 21 Suppl. 173S (1969). DISCUSSION MR. R. J. MCBRIDE: It is good to see you advocating 100K mortality level within a fixed time limit as a standard for a preservative suitable for use in preserving ophthalmic solutions. We believe that the time period of 3 h is too slow, however, especially for multidose solutions used in eye clinics, and we prefer the time of 1 h, suggested by Kohn et al (10). Our own work has shown that this 1 h sterilization time can be achieved against high inocula of Pseudomonas aeruginosa in preserved fluorescein sodium, pilocarpine hydrochloride, atropine sulphate, physostigmine sulphate, or salicylate and sodium bicarbonate ophthalmic solutions. T•E LECTURER: I suppose for ophthalmic solutions it is a question of toxicity I really do not know the optimal time. I would make the point that if you do not get 100•o mortality you are liable to have subsequent multiplication. For injections we have been obtaining sterilization times 1 h by incorporating 0.1• chlorocresol, and there have been no criticisms. It was only when we had 0.1• chlorocresol which did produce roughly a 10 min kill in ophthalmic solutions, that the problem of toxicity arose we now know that this was due to an unfortunate use in ophthalmic surgery. The major problem is one of toxicity rather than producing sterility. DR. R. M. E. RICHARDS: The use of mixtures of antibacterial agents for the preserva- tion of pharmaceuticals has not often been reported in the literature. Nevertheless mixtures of antibacterial agents are in frequent use in commercial preparations. Accord- ing to Remington (32) they now account for almost 50•o of the preservative systems in commercially available ophthalmic solutions produced in the U.S.A. Remington also predicts the increase of the use of mixtures of antibacterials in extemparaneous com- pounding. Would you please comment on the parameters to be observed in choosing a mixture of antibacterials in pharmaceutical preservation. THE LECTURER: AS far as I see, the advantages are that a mixture broadens the spectrum of micro-organisms which may l•e dealt with and it may increase the capacity of the preservative system. For example, if one uses shall we say, two of the parabens-- methyl paraben, propyl paraben, the probability is that with the more active compound (propyl parabort) the cell-aqueous phase partition coefficient will be in fayour of much of the propyl compound going into the cell, leaving the system deficient in activity. If one would also include the methyl compound this would tend to increase the capacity. There are great theoretical difficulties in handling these mixtures of preservatives--if they work, use them. I think the reason that so little has appeared in the literature is the (32) Remington, P. J. Pharmaceutical scieaces 14th edn. 1570 (1970) (Mack Publishing Co.. Easton, Pennsylvania).

720 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS difficulty of predicting precisely what they will do. Garratt and Brown (33) were able to produce a very simple kinetic explanation if you try to apply that to most antibacterials it just does not work. One can get a long way with mixtures within one chemical class. For example, if you are going to mix phenol and if you like chloroxylenol, and if you know the activities of the separate compounds and their concentration exponents, you can get a long way towards calculating with reasonable accuracy (q-50•o) precisely what the activity of the mixture will be. For mixtures of different classes of compounds, for example phenyl ethyl alcohol together with a paraben or similar, heaven only knows where the mathematics goes. It is not only a question of mixture of the preservatives I think Richards and McBride (34) made this point. If you have two preservatives in an ophthalmic solution you have three components in the system. Now attopine and, I think, pilocarpine will tend to enhance the activity of many of these preservatives whereas fluorescein will actimt•ro•posite direction. Until we have much more knowledge it is thus very difficult •to predicf• •Wrhat the activity of a product will be. MR. G. S¾•coes: I wonder w}i•Iher Remington (32) means a real 505/0 in the terms of mixtures in which we think of t :•:!• .,. or whether a large proportion of that 505/0 is not groups of homologues of the paralSL•,•hich can be called mixtures. DR. R. M. E. RicH^v,rs: The 505/0 do not include the parabens--these are all kinds of combinations of antibacterials probably with different modes of action and sometimes used at quite strange concentrations. THoe Loec'ruv,oev,: Many insulins contain a mixture of preservatives and, I think, this mixture is not always there for preservative reasons alone, e.g. there may be analgesic reasons. (33) Garrett, E. R. and Brown, M. R. W. The action of tetracydine and chloranphenicol alone and in admixture on the growth of E. cell. J. Pharm. Pharma½ol. 15 185T (1963). (34) Richards, R. M. E. and McBride, R. J. The preservation of ophthalmic solutions with antibacterial combinations. J. Pharm. Pharrna½ol. 23 Suppl. 235S (1971).