The rhythm of the heart was investigated in 123 cases of subacute bacterial endocarditis, 109 of which were in the active or bacterial phase of the disease, and 14 in the bacteria-free or healed stage. Auricular fibrillation was encountered but once among the active group, an incidence of less than 1 per cent, and in this case the rhythm had been normal until three days before death. In three other cases in which fibrillation was noted and in the single instance of flutter, the endocarditis had already progressed to the bacteria-free stage. These findings demonstrate how exceptional auricular fibrillation is in the active or bacterial period of the disease. On the other hand, this irregularity is observed in the bacteria-free stage with perhaps the same frequency as in chronic rheumatic valvular disease. Observations upon a large series of patients with fibrillation have convinced us that once this arrhythmia is permanently established, there is little danger that subacute bacterial endocarditis will supervene. The observation that auricular fibrillation and active subacute bacterial endocarditis are mutually exclusive (save in an exceptional instance) is of diagnostic value.

If one were to be asked what knowledge we now have concerning the course of, and the prognosis in, subacute bacterial endocarditis, one could answer in brief, as follows: 1.1. There is a small number of spontaneous, complete recoveries in the type of case with which we first became acquainted.2.2. There is a surprisingly large number of cases that come under observation with the sequelae of a former attack of the disease, either just before or directly after, the infectious stage has terminated, or, more commonly, without any clinical recognition of the attack.3.3. There occur mild cases of short or long duration.4.4. The disease may exist in recurrent form.5.5. In the course of healing, the lesions of subacute bacterial endocarditis play a rôle in the development of chronic valvular disease.