IN an orthopaedic hospital there is a range of staphylococcal infections varying in site and severity, some remaining localised in the skin, others entering the blood stream and setting up abscesses in bone and elsewhere. This study was undertaken to find out whether strains which cause the more severe infections in man, particularly those MATERIALS AND NETHODS Strains were originally selected by the slide method of Cadness-Graves et al. (1943), but all were subeequently tested on plttsms-ww plates. Human plasme from blood rejected by the blood bank wa8 mixed Coagduae teekr. J . PATE. BACL-VOL. Ispm (1954) 431 43% C. H. LACK AND D . Q. WAILLING Spertruin with nutrient agar in a concentration of 12-15 per cent.(v/v) of plasma a t a temperature of 46" C. The mixture was poured into Petri dishes and spotinoculated. Coagulaae production was indicated by the development of opacity around spots after over-night incubation. All slide-positive strains were also plate-positive. One hundred of these strains were ah0 tested by Fisk's tube method (Fisk, 1940) and found t o be tube-positive. Fibrinolysin tests. A human plasma-agar mixture similar to that deeclribed above wm heated at 56" C. for 20 minutes before pouring. This produced an opaque medium and fibrinolysh was judged by the zone of clearing around colonies after 48 hours' incubation (Christie and Wilson, 1941). Separation of fibrinolysis by plmmin from that by protease was made by incorporating soya-bean trypsin inhibitor in the medium. This inhibited fibrinolysis by p h r n i n but not by staphylococcal protease.