Mechanical trauma to the valves of the heart has long been known to be the most reliable experimental method of production of bacterial endocarditis.1 The first attempt by this method is attributed to Rosenbach (1878),2 who in dogs and rabbits succeeded in producing a bacterial endocarditis by injuring the aortic cusps with a nonsterile rod passed through the carotid artery. As the traumatizing agent, Ribbert (1885)3 used suspensions of staphylococci and starch granules from potato cultures; Fulci (1908),4 bacteria in carbon suspension; Panichi and Guelfi (1909),5 bacteria and carcinoma cells; Viti (1912),6 a medium composed of sterile flour and bacteria. Regardless of the nature of the traumatizing agent, a bacterial endocarditis ensued in the injured valves upon implantation of the bacteria. Equally successful was an experiment by Kinsella and Hayes (1927),7 who induced bacterial endocarditis in dogs by injuring the aortic valves with