COUNTING AND TESTING FOR MICRO-ORGANISMS 69 to an appropriate liquid or solid medium. Filters are of two types, Seitz or membranes composed of cellulose esters. The former is now rarely used having two main deficiencies. Firstly, the difficulty of washing the pad free from adsorbed inhibitory material secondly, the possibility of micro- organisms growing unobserved in the thickness of the pad. Membranes do not suffer from these deficiencies and have been widely used in conventional sterility testing (33, 34, 37). In the examination of products for microbial contamination the mem- brane-filtration technique has been successfully used by several workers (7, 13, 24, 25, 39) especially if the test sample is soluble in water or can be solubilized in an inert solvent to render it amenable to filtration. The method is applicable also to oils and ointments (42) using isopropyl myristate as the solubilizing agent and to insoluble powders using dimethylsulphoxide (11). 'Sterility test type' methods Total bacterial and fungal counts are normally carried out using non- selective media by the methods previously described. Opinion among microbiologists is divided on the value of such counts when applied to the microbial control of non-sterile pharmaceutical products or raw materials. There is no division of opinion, however, as to the need to exclude harmful organisms such as coliforms, particularly Escherichia coli, Salmonellae, Pseudomonas spp. and Staphylococcus aureus. For the detection of these specific organisms, or group of organisms, a 'sterility test type' method is carried out with the aid of selective media. As in conventional sterility testing, the methods used are either direct dilution of the test sample in media or by the filtration technique. Coliforms and Escherichia coli The methods used vary in detail but basically the test involves pre- incubation of the test sample in a nutrient broth followed by subculture into a MacConkey or lactose type broth with further incubation. Acid and gas production is indicative of the presence of coliforms. Further subculture from tubes showing acid and gas production, into fresh MacConkey broth and peptone water at 44øC identifies the presence of E. coli (type I) if acid and gas are produced in the MacConkey broth and indole in the peptone water. Further confirmation can be carried out using biochemical tests. Most workers have used MacConkey or a similar lactose medium but others (3) prefer the use of eosin-methylene blue medium or brilliant green

70 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS broth or violet red bile agar (8). Some have used pre-enrichment techniques (17, 18, 37) while others (19) used direct addition of the prepared sample on to a selective medium. Salmonellae The test used for the detection of Salmonellae appears to have been more standardized and is based on a pre-enrichment stage in nutrient or lactose broth followed by sub-culture into either a selenite or tetrathionate broth or both. Further confirmation of positive reactions is carried out on triple sugar iron, brilliant green, desoxycholate citrate or bismuth sulphite agar. Liquids suspected of contamination with Salmonellae have been recom- mended to be tested by the filtration method using brilliant green agar and triple sugar iron and selenite agar (43, 44). For further confirmation of Salmonellae, biochemical tests and serological techniques have been used. The British Pharmaceutical Codex (1968) (45) uses Rappaport's medium for the detection of Salmonellae in carmine. Pseudomonas All workers have used a cetrimide containing agar or broth for the primary detection of the pseudomonas group of organisms. For the identi- fication of Pseudomonas aeruginosa, some of the workers (8) have carried out this test only while others (17, 19, 25, 37) have supported their identi- fication with the oxidase and/or biochemical tests. In one study (22) exhaustive biochemical, serological and phage typing tests have been used for confirmation together with successive growth of the organism at 42øC. Staphylococcus aureus Sodium chloride is used in selective media in concentrations of 8-10•o, for the isolation of Staphylococcus aureus. Media of this type has been used: salt meat broth (17, 18, 20) Chapman-Stone medium (8, 19, 25). Other selective media containing lithium chloride and/or tellurite have also been used: Vogel-Johnson media (16, 25, 37) and Baird-Parker medium (16). For confirmation, the coagulase test has been recommended (37). SAMPLE PREPARATION Methods of isolation and enumeration of micro-organisms usually require some form of treatment to release into the diluent or medium those organisms which may be embedded or trapped within the bulk of the