Evidence for the statement that patients with subacute bacterial endocarditis may recover has hitherto been the fact that those with a typical clinical picture and from whose blood organisms are grown on cultures do survive and become free from symptoms and bacteria.1 Libman2 reported at least 3 per cent of recoveries in the usual type of the disease. He was also convinced that many more recoveries occur in persons with a mild form of the disease which is often overlooked. Pathologic confirmation of healed subacute bacterial endocarditis is, however, largely indirect and inferential in nature. For example, Weiss and Rhodes1 examined a group of hearts obtained at autopsy and selected 3 as being examples of completely healed subacute bacterial endocarditis. Others are described as suggesting a "healing" endocarditis. Hamman3 similarly described 4 cases, 2 of which showed evidence of "almost complete healing of vegetations," while in