Bacterial biofilms were observed on 69 of 75 catheters taken from patients undergoing long-term bladder management. Ten catheters were colonized by pure cultures of Proteus mirabilis. In each of these cases the bacteria formed layers on the catheter surface, underlying encrustations of struvite and hydroxyapatite which partially or completely occluded the catheter lumen. Encrustation was also apparent on catheters colonized by P. mirabilis plus other species, but was rarely seen on catheters colonized by non-urease-producing species. These observations support the hypothesis that catheter encrustation is brought about by the activity of urease-producing biofilms and confirms that the main target in the control of catheter encrustation should be P. mirabilis.

PLATES I-IVSWARMING by proteus on solid medium has been the subject of investigation since it was first described by Hauser (1885), as it is a phenomenon that does not occur to the same degree with other Gram-negative rods. Lominski and Lendrum (1947) suggested that swarming is a reaction by the organisms to toxic products of their own metabolism, a view supported by the work of Hughes (1957), who demonstrated that the forms in a swarm are morphologically indistinguishable from those produced by the damaging action of penicillin. A relation exists between the ability of proteus to swarm on a solid medium and the composition of the medium. It is common knowledge that the presence of bile salts inhibits swarming and a variety of toxic substances as diverse as chloral hydrate (Kramer and Koch, 1931) and sulphonamides (Holman, 1957) inhibit swarming when added to media. Naylor (1960) devised a medium that did not promote swarming of proteus, by omitting sodium chloride from the conventional formula of nutrient agar. Jones and Park (1967) found that growth but not swarming occurred on a solid medium based on the minimal medium of Fildes (1938), but these workers also claimed that swarming on this medium was promoted by the addition of various amino compounds.The present investigation is an attempt to define the components of solid media that are responsible for the promotion of swarming by proteus. MATERIALS AND METHODSOrganisms Proteus mirabilis NCTC5887 and 1 1 wild strains of Proteus mirabilis variously isolated from urinary infections, wounds, ulcers and rectal swabs were studied in the present investigation; the wild strains were identified as Proteus mirabilis by the criteria of Cowan and Steel (1965). The organisms were maintained in Robertson's cooked-meat broth at room temperature and were subcultured at monthly intervals. Before testing, the organisms were brought into the log phase of growth by inoculating them into 5 ml of nutrient broth and incubating for 5 hr at 37°C.