Although auricular fibrillation is one of the commonest and most serious disorders of the human heartbeat, its mechanism in man has not yet been demonstrated. Direct observation of the arrhythmias has heretofore been confined largely to that done on experimental animals. Flutter and fibrillation were produced in dogs by a variety of methods, including electrical, chemical, and mechanical stimulation.1Some of these methods showed disorders that appeared to involve a circus movement2; on the other hand, some instances of flutter and fibrillation were found to result from a rapidly discharging ectopic focus.3Thus the mechanism of arrhythmias observed in laboratory animals may vary with the method of production. Because of this limitation of the experimental approach, the only reliable means for determining the nature of clinical auricular fibrillation is to observe the disorder as it actually occurs in human hearts. Evidence obtained directly from the fibrillating auricles