480 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS pounds for a given dosage level were examined simultaneously so that the activities would be relative within a given dosage level. Results of the bactericidal evaluation may be seen in Table 2. Those compounds dem- onstrating a 95 per cent kill or better are asterisked at the corresponding dosage level. The efficacy of Tween 80 as an antidote for hexachlorophene was ported by Lawrence (12). Antidote effectiveness in this investigation was determined by inoculation of roll tubes, with and without the bis-phenols,I with a known number of organisms. The level of bis-phenols in the agarl was 100/•g./ml. and therefore approximately 5 to 30 times greater than the• maximum expected carry-over from the actual test system. No significantl differences were observed between tubes with and without Ns-phenol, andl it was presumed that bacteriostatic effects were eliminated. Each active• bis-phenol was examined in this manner and the antidote agar was shownl to nullify completely bacteriostatic effects. It may be noted that the method employed a minimum of extraneous• organic material in the test system, due to the inoculum dilution. For• this reason the initial numbers of cells is relatively low compared to, forl example, the A.O.A.C. methods for evaluating disinfectant type productsl (11). This factor was purposely taken into account in this evaluation to increase the sensitivity of the method so that differentiation between the compounds would be possible. The frequent use of Ns-phenols as bacteriostatic agents in liquid and bar soaps prompted us to evaluate soap solutions containing the most l active bis-phenol (No. 3) and hexachlorophene (No. 1). We employed essentially the same technique as with the synthetic anionic surfactant however, in this case, the compounds were tested at several ratios of Maxine (Swift) soap to test compound. The level of test compound was held constant at 1 •g./ml. As may be noted in Fig. 1, bactericidal action was observed to decrease as the soap ratio increased. The superior lethal properties of the hexachlorophene isomer (No. 3) to that of hexachlorophene (No. 1) is apparent in Fig. 1. Interest in the reduction of bath water counts by hexachlorophene was previously reported by Ayliffe (8). We had previously drawn some general conclusions with regard to the relationship between the bacteriostatic properties of the Ns-phenols and their chemical structure (1). Although a certain pattern is evident in the case of bactericidal activity, the connection with chemical configura- tion is less definite. The specificity of the bis-phenols, i.e., greater potency against gram positive organisms than against gram negative ones, is again noted from the data shown in Table 2. An isomer of G-11, 2,2'-methylenebis-(3,4,5- trichlorophenol) (No. 3), demonstrated the greatest bactericidal activity against S. aureus. This same compound was inhibitory at the lowest level

CHEMICAL STRUCTURE AND ANTIMICROBIAL ACTIVITY 481 (0.2 •g./ml.) as was shown previously in the bacteriostatic evaluation. There, it was also the most effective substance against E. co/i here, other compounds are superior to it. The other isomer of G-11, 2,2'-methylene- bis(4,5,6-trichlorophenol) (No. 2), is unexpectedly inactive at the test levels G-f1 (No.1) itself falls between the two. 2,2'-Methylenebis(5,6- dichlorophenol) (No. 9), which was a much weaker bacteriostat than G-5 © (No. 8) and G-11, is outstanding against both S. aureus and E. co/i in the present series. It is also surprising that a compound with the methylene-linkage in the 4,4' position, such as 4,4'-methylenebis(2,3,6- Controls •2- o 25 Compound No.I I I 4 I Compound No. 3 x,•5 I I I0 20 :$0 I0 20 30 Exposure time (min.) Exposure time (min.) Figure 1.--Bactericidal activity of compounds 1 and 3 against S. aureus in 100, 50 and 25 •g./ml. Maxine soap solution at 40øC. Level of compounds is 1.0 •g./ml. trichlorophenol) (No. 42), is as bactericidal against S. aureus as its isomer G-11 bacteriostatically, compound 42 was found to be much inferior to GAl. In addition to compound 9, the most active compounds against E. coli are bis-phenols which have only one halogen substituent in each ring, such as Nos. 6, 15, 20, 30 and 31. It has been shown that the germicidal potency of such bis-phenols against gram negative organisms is greater than that of Ns-phenols with four to eight chlorine atoms in the rings when tested by a phenol coefficient type of test (2). We observed, as we did previously, that no compound without halogen in the phenol rings is active. The thio-linkage appears to lead to somewhat more active compounds than the methylene-bridge. Replacement of the phenols by naphthols results in loss of activity, with the exception of compound 39.