84 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Our tests, using G-3720, in various ratios and a uniform 0.2 per cent concentration of methyl p-hydroxybenzoate in Jaag medium showed growth at all concentrations and ratios of the nonionic to the preservative the heaviest growth occurring at 0.6, 1.0, 2, 4.0 and 6 per cent of the nonionic. Heavy growth is first noticeable when the ratio of nonionic to preservative is3:l. However, using 2 per cent Tween 80 and concentrations of methyl p-hydroxybenzoate from 0.2 per cent to 0.5 per cent, an oil (complex?) eventually separated from the medium when the concentration of methyl p-hydroxybenzoate reached 0.3 per cent. In a similar experiment using Myrj 49, no oil or crystalline separation took place, but it required 0.5 per cent of methyl p-hydroxybenzoate to prevent growth of •/. niger in Jaag medium. In a similar series it required 0.4 per cent of methyl p-hydroxybenzoate to preserve 2 per cent G-3720 against .4. niger. While at 1 per cent of methyl p-hydroxybenzoate and 2 per cent G-3720, a crystalline compound settled out. Igepal Co Growth Moles ETO Growth Months 210 q- q- 1.5 Carbopo1934 d- q- 6 430 q- q- 4 Cellosize - 8 630 d- d- d- d- 10 Tragacanth d- d- d- 5 730 + 15 PVP K-30 + + 5 850 d- d- d- d- 20 Methocel 1500 d- d- 2 880 + + 30 Figure 3.--Growth of .4. niger in Jaag Figure 2--Eleven Month Growth .4. niger Medium Plus 0.2% MP* Plus 2% Gum Jaag Medium, 0.2% MP* and 2% Igepal * MP = Methyl p-hydroxybenzoate. ß MP = Methyl p-hydroxybenzoate. ESZYMIe ACTIVITY We have found that polyoxyethylene ethers of the type of G-3720, particularly in the presence of dehydroacetic acid are apparently not only split by a possible "etherase" but that the prod.uct of hydrolysis is either selectively split from the nonionic liberating lower molecular weight alcohols, from the mixture comprising commercial stearyl alcohol or that the stearyl alcohol is somehow degraded. This splitting is also found occasionally in mixtures preserved with sorbic acid. A similar splitting has been observed in mixtures of sodium alkyl sulfate with G-3720, Tween 60, Myrj 49 and ethoxylated glyceryl monostearate. Oddly enough, no mold growth has been found in any of these tests. It is obvious that in terms of a commercial emulsion, any splitting of the emulsifier by microbial esterases could lead to a separation of the emulsion phases.

INTERFERENCE OF EMULSIFIERS WITH PRESERVATIVES VIII 85 SORBIC Acre TESTS It has been thought that the discrepancy between our results and those reported by others using sorbic acid, could be in the medium employed. Both magnesium and ferrous iron have been suggested as forming sorbates less soluble than the free acid. However, magnesium sorbate is far more water soluble than sorbic acid itself. In experiments made with Jaag medium with and without these two ions, no difference in growth was observed. When deionized water served as the medium, the addition of these ions in the concentration found in Jaag medium showed no effect on preservative action. Though Lach (4) has found sorbic acid to be less complexed by poly- glycols than the p-hydroxybenzoates, our work to date, using mainly niger as the test organism does not confirm this. Furthermore, one of us (M. G. deN.) has found that at a pH above 5.0, sorbic acid is a poor pre- servative even in the absence of nonionics in Jaag medium. against niger. 0.1% cationic only* 0.1% cationic plus 2% G-3720' 0.1% MPt 0.15% MP 0.2% MP Ethyl Hyamine Roccal Cetab 10X ++++ ++++ ++++ + ++ _ _ ++++ Figure 4.--Twelve Month Growth of •/. niger in 2% Mixture of Cationic (20%) in G-3720 with Methyl p-Hydroxybenzoate * No other preservative. t Methyl p-hydroxybenzoate. 12 Months Tween 80 q- q- q- G-3720 q- q- G-3810 G-1425 q- q- Figure 5.--Growth •/. niger, in Jaag Medium, 2% Nonionic 0.2% MP* * Methyl p-hydroxybenzoate. MISCELLANEOUS OBSERVATIONS Of the many preservatives tested, we found benzoic, sorbic and dehydro- acetic acids very sensitive to pH in their antimicrobial activity. Ethyl vanillate and the parabens had a fairly wide pH spectrum. In general, we found that the greater the number of moles of ethylene oxide in the ether or ester chain, the greater the inactivation of pre- servative. These data are based on tests of a series of ethoxylated nonyl phenols, glyceryl monostearate, stearic acid and cetyl alcohol. We found •lspergi//us growing slightly in 0.2 per cent formaldehyde (100 per cent basis) in Jaag medium containing 2 per cent G-3720. Similar