82 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS The commercial nonionic G-3720 was used throughout this work. It is a polyoxyethylene (20) stearyl alcohol. A 2 per cent concentration was used. Tween 80 was used in a concentration of 2 per cent in some tests. It is a polyoxyethylene (20) sorbitan oleate. Methyl p-hydroxybenzoate 0.2 per cent was the only preservative tried. Varying amounts of glycol were used as the test required with due adjustment of water content to compensate for the glycol (1). The polyols used were propylene glycol of commerce 1,3-propanediol dipropylene glycol 1,4-butanediol hexylene glycol sorbitol and glycerol. Discussion. In this work, results with molds cannot be considered on the same basis with usual bacteriological tests. Thus, it is found that Jaag medium preserved with 0.1 per cent methyl p-hydroxybenzoate showed no growth of,4. niser for four months. At the end of seven months, a very heavy growth was obtained. Too often such tests are reported on a seven-, fourteen- or thirty-day basis. Our tests with molds are con- tinued for at least a year. If the number and position of hydroxyl groups on a polyol have an effect on preservative inactivation, then it is desirable to examine the configuration of the polyols used. The propylene glycol of commerce is largely 2,3-propanediol with a small amount, rarely over 5 per cent, of 1,3-propanediol. Hexylene glycol of commerce is 2-methyl 2,4-pentanediol having the following configuration: HO OH I I HaC--C--C--C--CHa H H HCa Hexylene Glycol H H H H--C--C--C--H HO H OH 1,3-Propanediol H H H I I I HC--C--CH H OH OH 2,3-Propanediol In hexylene glycol, the two hydroxyl groups are not attached to adjacent carbons as they are in ordinary propylene glycol. This configuration could account for the difference in behavior. At the same time if this line of reasoning is valid, then such commercial compounds as 1,4-bu- tanediol, 1,3-butanediol, hexanetriol-l,2,6 and others may have similar effects, but sorbitol and glycerol would be comparatively ineffective. Results. We find that 8 per cent propylene glycol when added to the

INTERFERENCE OF EMULSIFIERS WITH PRESERVATIVES VIII 83 medium containing G-3720 or Tween 80 and methyl p-hydroxybenzoate showed no growth after six months using the four bacteria and z/. niger. In view of the effect of hexylene glycol previously mentioned, a series of tubes was set up using various concentrations of the glycol with G-3720 and methyl p-hydroxybenzoate. The results showed that 4 per cent hexylene glycol is as effective as 8 per cent propylene glycol in preventing growth of the four bacteria and z/. niger for four months. Hexylene glycol as the sole preservative failed to prevent growth of both bacteria and z/. niger in 2, 3 and 4 per cent concentrations. In a concentration of 5 per cent the Pseudomonas and Proteus bacteria were still able to grow as did the •spergi//us. These results are at variance with those mentioned earlier (1). Concentrations of glycerol and sorbitol ranging from 2 to 12 per cent failed to prevent inactivation of methyl p-hydroxybenzoate by G-3720. While growth periods varied, all samples grew actively within two months. In screening 1,4-butanediol, dipropylene glycol and 1,3-propanediol, arbitrary concentrations of 5 and 10 per cent were added to the appropriate medium containing methyl p-hydroxybenzoate and G-3720. In all three cases the 5 per cent concentration failed to prevent growth of only Pseu- domonas aeruginosa, but a 10 per cent concentration of either glycol inhibited the growth of all the organisms. Further tests are necessary to find the more exact concentration of glycol that will prevent growth of these microiSrganisms in a nonionic medium containing 0.2 per cent methyl p-hydroxybenzoate. G-3720, % 0.1 0.2 0.4 0.6 08 10 20 4O 6O 100 Ratio Growth Months G-3720: MP* + 10 1/2:1 + 13 1:1 + 9 2:1 -1--1--1- + 2 3:1 ++ 2 4:1 ++++ 2 5:1 ++++ 4 10:1 +++ 11/2 20:1 +++ 11/2 30:1 ++ ll/= 50:1 * Methyl p-hydroxybenzoate. Figure 1.--Growth z/. niger in Jaag Medium with G-3720 and 0.2% MP EFFECT OF RATIO OF NONIONIC TO PRESERVATIVE Pisano and Kostenbauder (3) have found that 5 per cent Tween 80 in water bound 78 per cent of the methyl and 90 per cent of propyl p- hydroxybenzoates, using a dialysis technique. Somewhat later it was found that these figures are increased by bacteriological media.