IN A SERIES of articles in 1962, Dennis,'Senning,2 Hall,3 and Moreno4 described a method by which the left ventricle could be bypassed without thoracotoiny. A cannula passed down the right internal jugular vein perforates the atrial septunm and enters the left atrium; a simple circuit then allows left atrial blood to be pumped to the femoral artery. This method of reducing the work of the left ventricle has obvious possibilities of being clinically useful. This is a report of an investigation into the effects of such a bypass on circulatory dynamics and nmyocardial metabolism. Attempts to assess its value in experimental coronary shock and following aortic valvular operations are also reported.Methods Forty-five mongrel dogs (11 to 32 Kg.) were anesthetized with Pentothal, atropine, and nitrous oxide. Both pleural cavities were entered via short intereostal incisions. Pressures were nionitored in the ascending aorta, the right atrium, the pulmonary artery, and the left atrium. The arterial pressures were transmiiitted via strain-gauge transducers to a Grass four-channel recorder on which the dye curves and electrocardiogram were also recorded. The atrial pressures were monitored on saline manomiieters. All pressures were standardized to a level of 10 em. above the operating table.5 Esophageal temperature was kept between 36 and 39 C. by means of a controlled temperature blanket beneath the dog and a heat exchanger in the arterial line. After preliminary dissections were completed, heparin (2 mg./Kg.) was given.Cardiac outputs were estimated by the indicatordilution technique (indocyanine-green dye, Colson densitometer, and three-cycle logarithmic paper). Two curves were used in estimating each output.For right ventricular outputs, the injection site was the superior vena cava and the sampling site the pulmonary artery; for left ventricular outputs,