Although a good many selective culture media have been described for the isolation of Staphylococcus aureus, none of them has proved useful for the examination of cultures from air. This is partly because, in examining such cultures, one is trying to select the S. aureus from among a great variety of other members of the Micrococcaceae. Media depending on a high concentration of sodium chloride (e.g. Chapman, 1944), which are principally inhibitory for enterobacteria and similar Gram-negative bacilli, are not, therefore, of much use. Other media, such as those containing potassium tellurite (Zebovitz, Evans & Niven, 1955) or polymyxin B (Davis & Davis, 1965), though often satisfactory for examination of material from infected lesions or carrier sites, prove too inhibitory for the isolation of staphylococci from air. It has been a common experience that bacteria are often difficult to cultivate when collected from the airborne state. The recent studies on the taxonomy of staphylococci and micrococci published by Baird-Parker (1963, 1965) suggest a new approach, for he showed that S. aureus is able to grow at a higher temperature than most of the micrococci. Moreover members of the genus Staphylococcus are able to grow under anaerobic conditions, while those of the genus Micrococcus cannot, and other studies have shown that a substantial proportion of the airborne cocci fall into Micrococcus (Corse & Williams, 1968). We therefore explored the use of anaerobic culture at 41°C. as a selective method for the isolation of S. aureus. After a number of preliminary experiments with pure cultures, which need not be discussed here, all our studies were made by the exposure of culture plates containing the media under test in the wards of St Mary's Hospital.
METHODS
Air samplingCulture plates 14 cm. in diameter were exposed in hospital wards for periods varying between 8 and 16 hr. The plates were stacked in the holder illustrated in Plate 1, which provides about 5 cm. separation between plates. Preliminary experiments showed that there were no significant differences in the numbers of colonies developing on the four plates in the stack. The culture plates with the different media were placed at random on the four layers of the stack. One or more stacks of plates were exposed each day but each stack included at least one 'control' plate.The culture media used are listed in Table 1. 3-2