284 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS 9.0 .c_ 7-0 'F,_, '• 6.0 'B 5.0 "• ,4.0 .• 3-0 • z.o I'0 o 4h I I I I I I I // I I 2 3 4 5 6 7 14 Time {doys) Figure 3. Influence of inoculum size and shampoo volume on the total popula- tion of E. cloacae in a model shampoo. Initial inoculum: --O---- = 50 bacteria in 10 g shampoo ---O .... 400 bacteria in 10 g shampoo • = 50 bacteria in 100 g shampoo ----• .... 400 bacteria in 100 g shampoo , l-- 50 bacteria in 1000 g shampoo ---1 - 400 bacteria in 1000 g shampoo. Isolation from mains water of bacteria capable of multiplying in a model shampoo Inoculation of shampoos with filters derived from filtration of 1000 ml of water or less did not result in contamination with bacteria capable of multiplying in the shampoo. Of five inoculations with filters derived from l0 1 volumes two resulted in contamination of the shampoo (Table I). Viable counts on the shampoos immediately after inoculation in every case failed to reveal the presence of bacteria. Despite the relatively low population which developed in the contaminated shampoos (2.7 x 105 and 3.6 x 103 g-•) contamination was associated with obvious visual changes, i.e. the production ooe turbidity and strings of slime, after only 7 days' incubation.

WATER-BORNE BACTERIA AND SHAMPOO SPOILAGE 285 Table I. Multiplication of bacteria isolated from mains-water in a model shampoo system Calculated no. of No. of shampoost Volume of bacteria in showing water filtered inoculum Growth No growth '1 ml 1.55 x 10: 0 5 10 ml 1.55 x 10 a 0 5 100 ml 1.55 x 10 • 0 5 1 1 1.55 x l0 s 0 5 101 1.34 x 106 2 3 * 1 ml volumes were not filtered but were added directly to the shampoo. Since the 10 1 volumes were not all filtered on a single occasion the number of bacteria in the inoculum expressed in the table is an average value. t Five replicates. Both contaminated shampoos were contaminated with a single bacterial species. First stage identification using the Cowan and Steel procedure (25) indicated that both organisms belonged to the genus Alcaligenes. Bio- chemically and morphologically both contaminants appeared to be the same organism. DISCUSSION The maximum volume of shampoo contaminated was 1000 g. However, it is clear that the organisms which survived and multiplied in this, because of their low numbers, did so as individuals and were in no way influenced by the presence of other organisms. In other words because of the enormous dilution factor involved it is unlikely that the initial growth and division of any single organism was assisted by the presence of metabolites or enzymes produced by other bacteria in the shampoo. Thus it is reasonable to postu- late that since 50 bacteria are sufficient to contaminate 1000 g of shampoo the same number would be capable of contaminating a whole factory batch --as much as 2000 kg. It is apparent that the minimum number of bacteria required to contaminate a shampoo need not be expressed in terms of numbers per gram but in numbers per batch. However, the number per gram is important in terms of detection. In products in which bacterial multiplication is possible contamination will only become detectable when