When we undertook to relieve angina by following the suggestion of Franck as carried out first by Jonnesco,1 we were fully conscious that all that had been accomplished by Jonnesco's first operation was to relieve the substernal anginal pain and the associated referred pains in his patient, by a removal of the left cervical sympathetic system and the upper thoracic ganglion. The functions of much of the removed parts were the subject of some important differences of opinion, and certainly many tracts transmitting a variety of impulses were severed in Jonnesco's operation. It seemed to us wise, therefore, to simplify if possible this complex situation by beginning at one end of the tract and severing as few connections as possible, continuing to interrupt paths between heart and brain and cord until we found just what brought the relief. Our first four patients2 were operated on by severing merely the main