212 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS soluble sulfates or glucuronides and then liberated by hydrolytic enzymes of bacteria and/or the skin as volatile steroids. A gas chromatographic analysis of axillary sweat, performed parallel to a sniff test with perfumers, gave 20 different odorous substances. A few by these substances characterized only by retention times occurred in all persons studied (2). Moreover, amino acids with a characteristic odor have been identified in eccrine sweat (15). Sweat secretion, the bacteria population, and a moist environment are the three major components contributing to odor production by the skin (4,6,14,21,22). The bacteria flora of the skin vary within broad limits both qualitatively and quantitatively (5,14). In a review, the relationships between the bacteria population, the host, and the environment have already been described (7). From the foregoing it becomes understandable why it is possible to inhibit sweat odor by different mechanisms. These include: ß inhibition of sweat secretion by systemic administration of sedatives, ataractics, parasympatholytics, and saluretics ß application of topical antiperspirants, such as formaldehyde, glutaraldehyde (danger of sensitization), and formulations containing aluminum hydroxycholoride and tan- nins ß binding of odorous substances by mixtures of zinc ricinoleate and other zinc com- pounds that act synergistically (17) ß environmental control, e.g., by body hygiene (soap or surfactants), and by absorbent, loose underwear ß deodorizing by means of an antibacterial therapy with strong disinfectants such as halogenated phenol compounds or quarternary ammonium compounds that influence virtually all of skin flora in the same way (12) Inhibition of esterases is an alternative mechanism of deodorant action. Glyceryl triace- tate, triethyl citrate, and other rapidly saponifying esters represent substances acting according to this mechanism. However, most other esters are bacteriologically inert, i.e., have no measurable antibacterial effect, when tested according to conventional methods. The aryl sulfatases and [3-glucuronidases can also be inhibited by Cu + +- and Zn + + compounds. For example, even concentrations of 10-100 •x Cu or zinc glycinate have an effect (3). If proliferation of bacteria is prevented by antimicrobial substances, the production of skin odor caused by bacterial decomposition of sweat is also largely reduced. However, it is still possible that the deodorant effect could be a result of a regulatory role of the substance in the biochemical processes on the skin surface. For years the deodorant HGQ, which is recommended as a natural synergistic complex of active substances (9), has proved successful in everyday use. It is offered as a con- centrate as well as a 50% solution in dipropylene glycol or ethanol, and consists of three individual components: ß 34% farnesol, which has been identified in cotton bud oil, cabreuva oil, musk seed oil, neroli oil, tuberose blossom oil and other volatile oils ß 11% glyceryl monolaurate, found in the feather fat or marabous (Leptoptilos cru- meniferus) ß 53% phenoxyethanol, which occurs in tropical fruits, in Cichorium endivia and in Camellia sinensis (green tea)

GROWTH INHIBITION OF CORYNEFORM BACTERIA 213 In a concentration of 0.3%, HGQ completely inhibits the growth of Staphylococcus aureus, Staphylococcus epidermidis, and Propionibacterium acnes (10). The present study was designed to test the bactericidal properties of HGQ against wild strains of corynebacteria isolated from human axillary swabs. Based on biochemical characterization, 30 species of coryneform bacteria were identified from 530 human axillary swabs. These were tested for their sensitivity to HGQ. MATERIALS AND METHODS ISOLATION AND IDENTIFICATION OF CORYNEBACTERIA STRAINS Smears freshly taken with cotton swabs from axillary skin were plated on blood and endoagar. After incubation for 24 and 48 hours at 37øC, potential corynebacteria strains were isolated as individual colonies. They were identified by their typical macromor- phological appearance, being usually gray, opaque colonies, and by gram staining as gram-positive rods. Different strains were identified by examination of the following properties: catalase and oxidase production capacity, [3-hemolysis on sheep blood agar, nitrate reduction, pigment formation, ureases, gelatin hydrolysis, mobility, esculin hydrolysis, serum response, and glucose, lactose, maltose, rhamnose, arabinose, trehalose, saccharose, xylose, and manitol fermentation. Tests were performed and interpreted according to methods described by Coylectal and Lipsky (1), Lenette et al. (13), and Lipstick et al. (16). Reference strains in the tests were a C. diphtheriae Park William-8-strain and a C. pseudotuberculosis strain. The purity of the strains to be iden- tified was checked on blood agar plates before and after application of the inoculum to the test series.* TESTING THE SENSITIVITY OF IDENTIFIED SPECIES OF CORYNEBACTERIA STRAINS TO HGQ The only feasible method proved to be determination of the minimum bactericidal concentration (MBC). 10, 3, 1, 0.5 and 0.25% HGQ stock solutions were prepared in 70% ethanol. Nine milliliters of glucose nutrient broth (Berlin-WeiBensee Institute of Immunopreparations and Nutrient Media) were added to 1 ml of each of the HGQ stock solutions, giving a final HGQ concentration of 1 to 0.025% and an ethanol content of 7%. An inoculating loop was used to inoculate the 6-h preculture, which had been adjusted to 1 million viable microorganisms/mi. An inoculated glucose nutrient broth tube with 7% ethanol served as the positive control, and uninoculated tubes of each of the HGQ concentrations were the negative controls. To ensure a maximum homogeneous distribution of HGQ in the nutrient medium, the tubes were shaken at a moderate rate during the entire 18-h incubation period at 37øC. Aliquots from the shaken culture tubes were spread on blood agar (5% human blood) by means of an inoculating loop. After incubating for 48 h at 37øC, those smears in which less than 0.1% of the original inoculum could be counted as individual colonies were taken as the MBC. * We are grateful to Dr. Med. Lehmann of the Institute of Medical Microbiology and Epidemiology of the University of Leipzig for his assistance with species identification.