The two cases that are the basis for this report exhibited numerous fleeting paroxysms of abnormal auricular action closely resembling auricular flutter. The mechanisms of these paroxysms, as will be pointed out, may be regarded as forms of slightly "impure" flutter. The remarkable never failing brevity of paroxysms, their rapid and at times almost regular recurrence, furnished striking clinical pictures which, as it seemed to us, deserve emphasis as a clinical entity. Furthermore, these cases are of interest since numerous onsets and offsets of clinical circus movement, of which hitherto there have been but few recorded, were obtained. It has therefore been possible to make certain new observations regarding the onset and offset of clinical circus movement and the effects of vagus stimulation and certain drugs on its incidence and duration. REPORT OF CASES Case 1.—J. S., a Polish coal miner, aged 50, was admitted to the Medical Division of