382 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS pyogenes, Proteus vulgaris and Ps.aeruginosa using inocula of about 103-104 cells/mi. Tests were made at 28 ø in the presence of numerous ophthalmic drugs in 75 formulations. Counts were made of recovered cells, and the preservative was regarded as satisfactory if the count remained static or was reduced. In most instances sterility was achieved within the shortest test period of one day. Increased counts occurred with both serum and methylcellulose formulations in the presence of 0.005% chlorhexidine. Growth occurred in the presence of lignocaine in one out of two formu- lations. It was suggested that growth in the presence of pilocarpine was due to incompatibility with the 0.005% chlorhexidine. Chlorhexidine is not in fact used in the Australian Pharmaceutical Formulary for eye drops containing pilocarpine (50). The experimental procedures of these workers are open to question. They did not use conventional inactivators in their recovery medium but attempted to inactivate the agents tested, including chlorhexidine, using a washing procedure, the efficiency of which apparently was not demonstrated. They recovered chemically treated Pr.vulgaris cells in medium containing 0.1% phenol which might well reduce recovery as well as swarming. One other unusual aspect of their experimental pro- cedure was an 18 hr recovery incubation period. Without evidence to the contrary, this would seem to be a short time for the recovery of chemically treated organisms. Hugo and Foster (32) found that inocula of Ps.aeruginosa (100/ml) were sterilized in aqueous suspension at 18 ø by 0.005% chlorhexidine, and at 30 ø by 0.002%. Kohn et al (25) tested 51 chemicals not previously employed widely in ophthalmic solutions as preservatives. Thirteen strains of Ps.aeruginosa (2 x 106/ml) were used, and in vitro and in vivo methods employed. Six out of 37 quaternary ammonium compounds were found to possess equal or superior activity to benzalkonium chloride which they used as a standard. All of eight amphoteric surfactants were found to be in- adequate while three iodophors were found to possess sterilizing times of less than 1 hr. Chlorhexidine and Colistin were each found to sterilize in less than 1 hr. Richards (35) found that chlorhexidine was less active in final eye drop preparations at acid pH than it was in simple solution. Wiseman (51) attempted to increase the resistance of several gram- positive and gram-negative species to chlorhexidine by training in its presence. He failed to increase the resistance of gram-positive organisms but found that gram-negative species showed a substantial increase in resistance. In particular the minimum inhibitory concentration against

THE PRESERVATION OF OPHTHALMIC PREPARATIONS 383 Ps.aeruginosa increased progressively from an initial 4 •g/ml to 64 •g/ml after 8 subcultures in the presence of chlorhexidine diacetate. Eronopol (2-bromo-2-nitropropane-1, 3-diol) is a new antimicrobial agent shown to be active against numerous strains of several gram-positive and gram-negative species including 22 strains of Ps.aeruginosa (52). Pre- liminary experiments showed that bactericidal concentrations are not irritant to the skin or mucous membranes. The authors failed to increase the resistance of Ps.aeruginosa and Staph.aureus to Bronopol by repeated subculture in its presence. SUMMARY AND CONCLUSIONS Consideration of the available published evidence reveals that at present there is no single ideal preservative for ophthalmic preparations. The capacity of strains of Ps.aeruginosa or closely related species to become resistant to quaternary ammonium compounds (11), chlorbutol (39), phenylethyl alcohol (36) chlorhexidine (51) and even to decompose phenol (53) and phydroxy-benzoates (9) would suggest caution in accepting any new agent supposed not to have this defect against this enigmatic organism. Many of the results in the literature are conflicting. Some of the discrepancies may be accounted for by reason of variation in such important factors as reaction temperature, composition of reaction mixture, and recovery medium. Importance must be attached to the results of those workers who have clearly defined their experimental conditions, and who have related the efficiency of the inactivators used in their recovery media to the results of in vivo tests. The capacity or not to cause disease would appear to be the ultimate, objective assessment, particularly when activity is measured in a final ophthalmic formulation. An appraisal of the scientifically acceptable data on the phydroxy- benzoates would suggest that nearly saturated solutions are capable of sterilizing heavy inocula (108) of Ps.aeruginosa in several hours. Such concentrations are too irritant for normal ophthalmic use. A concentra- tion of about 0.1% mixture may be useful in some eye preparations excluding, in particular, fluorescein. The capacity of Ps.aeruginosa to utilise (9) these compounds as a source of carbon would seem to preclude their recommendation for widespread use. The available evidence suggests that the concentration used in Solution for Eye Drops B.P.C. (0.0229% methyl, 0.0114% propyl) is unlikely to sterilize, and may possibly allow growth of Ps.aeruginosa.