8O2 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS vent invasion also retard the antimicrobial activity of the biostats, so that a W/O emulsion which is more difficult to infect also appears to be more difficult to preserve. Other factors affecting the activity of pre- servatives in W/O systems are those controlling the availability of the preservatives in the aqueous phase, in particular, the O/W partition co- efficient of the preservative, the phase-volume ratio, and the tempera- ture. Antimicrobial compounds with high oil-water partition coeffi- cients concentrate in the oil often with insufficient quantities being dis- tributed in the aqueous phase to inhibit microbial growth. The parti- tion coefficient of most biostatic agents will be higher for vegetable oils than for mineral oils, making vegetable oil creams more difficult to pre- serve than those prepared from mineral oil. Knowing the concentration of a biostat, the dissociation constant, the concentration of nonionics and the partition coefficient, the amount of the agent dissolved in the aqueous and micellar phases at a given pH can be calculated. Evans (22) found, for example, that a 0.5% concentration of p-hydroxy benzoate in an aqueous 6.0% Tween solution gave a concentration equivalent to 0.1% of the benzoate in the water phase. Noble and Savin (23) reported a hospital pseudomonas outbreak which illustrates these concepts. A well preserved steroid cream was diluted with lipids. The diluted cream containing 0.1% chlorocresol was found contaminated with P. aeruginosa. Although 0.02% chloro- cresol in aqueous solution was sufficient to prevent growth of organism, the chlorocresol migrated into the oil constituents of the cream leaving the aqueous portion with insufficient biostat. Examination of the clini- cal records suggested that the cream had caused minor infections over a period of several months. Similarly diluted creams from two of eight other hospitals were also found contaminated with pseudomonads. In preserving a product one frequently overlooks the nutritive qual- ity of the preparation. Cosmetic and pharmaceutical formulations often range from water solutions to high protein soups. The growth po- tential of the product itself is a prime agent influencing the ease or diffi- culty of preservation, for a nutrient-free preparation is easier to preserve than one containing nutriles. The "magic ingredients" now current in many formulations can on occasion be nutrients, stimulating the growth of organisms and obstructing the action of preservatives. Existing screening tests in agar or broth which determine a com- pound's inherent antimicrobial activity are helpful only in eliminating completely inactive materials. They will not predict, with any reliabil- ity, performance in the actual cosmetic or pharmaceutical emulsion.

PSEUDOMONADS IN COSMETICS 803 The preservation of formulations depend on many factors and the only reliable way to determine the resistance of a product to organism growth is to inoculate and test the product itself. In the author's laboratories a minimum incubation period of 13 weeks is utilized to determine the possible loss of preservation potency with time. Re-inoculation of the samples during incubation often demon- strates declining effectiveness of the preservation system. On one occa- sion during a prolonged product inoculation test the preservative re- pressed the inoculum for 17 weeks, but shortly thereafter vigorous growth and deterioration of the preparation ensued. Adaptation of pseudomonads can take place quickly, or very slowly over several weeks, and it is the latter which one has to watch (6). This may account for the sporadic spoilage reported in the field when control samples on shelves are in perfect condition (19). Preservation is interpreted by some as preventing the growth of micro-organisms. Does that definition fully protect the consumer? When pseudomonads introduced into a product survive, but do not multiply, can one really say, from what is known of the organism, that the preparation is adequately preserved? Kohn et al. (24), consider a preservative too slow-acting to be used in ophthalmic solutions if it could not sterilize a given pseudomonas suspension within one hour. Preparations are placed on preservation study to determine the ability of the product to withstand consumer use and abuse. The numbers of pseudomonads one finds in a product may have been introduced as incre- ments during use or developed as a result of growth. Regardless of the origin, the presence of a large number of pseudomonads over a significant period of time predicates a quality, health, and legal hazard. A small number of organisms surviving within a preparation may in time adapt to and proliferate in the product. It therefore follows that the only ade- quately preserved preparation is one that is essentially self-sterilizing. The sterilizing time however need not be measured in hours but in days or even weeks. The test organisms which the author employs have been taken from working formulations which were at one time contaminated, so that the screen contains strains adapted to and encountered by the laboratory's preparations. Screening methods should not be fixed since new strains of pseudomonads continually are isolated from domestic and foreign sources. It must be accepted that unforeseen and exceptional organisms or conditions may turn up during routine manufacture to inactivate a