Distinctive Characteristics of Individual Members of the Eisenmenger GroupPatent Ductus.- (1) The catheter passed through the duct more easily than through any other central communication between the two circulations. It did so in 90% of cases, and usually took an anticlockwise course to enter the descending aorta. (2) The shunt was wholly reversed far more frequently (50%) than with other defects. (3) Patent ductus was the only lesion which was occasionally associated with no shunt in either direction. (4 Differential oxygen desaturation between right brachial and femoral samples was pathognomonic of patent ductus, and occurred in all but two cases in which there was no shunt in either direction. The oxygen saturation of samples from the right brachial artery averaged 89.6% (range 81 to 96%), and from the descending aorta or femoral artery 77.1 % (range 65 to 88%/). Aorto-pulmonary Septal Defect.-(l) When the catheter passed through the defect it entered the ascending aorta from the pulmonary artery. (2) In other respects the findings were similar to those of patent ductus with bidirectional shunt but no differential desaturation.Eisenmenger's Complex.-(1) Passage of the catheter via the defect (28%) into the ascending aorta followed a characteristic medial course as in Fallot's tetralogy. (2) An appreciable left-to-right shunt at ventricular level, as described by Bing et al. (1947), could be detected in all but two cases, in both of which the diagnosis was subsequently proved at necropsy. (3 (1) The physiological situation was similar to that in aorto-pulmonary septal defect with the addition of a second left-to-right shunt at ventricular level. As the catheter passed from the right atrium, through the right ventricle, into the truncus, the oxygen saturation of respective samples rose first by 10 and 12.5%, and then by a further 10 and 14%. (2) The catheter entered the pulmonary arteries from the " aorta," instead of vice versa as in aorto-pulmonary septal defect. (3) Samples from the " aorta" and pulmonary arteries were identical.Atrial Septal Defect.-(I) The catheter passed through the defect in two-thirds of 18 proved cases. (2) The pulmonary artery pressure was less than the systemic pressure by an average of 27/34 mm. Hg in 70% of the cases; in 10% it ,was higher; and in the remalnder about the same. When mean pressures were similar the pulse pressure was usually greater in the pulmonary artery, the systolic pressure being higher and the diastolic lower than in the systemic arteries. Agents affecting pulmonary or systemic flow or resistance selectively usually altered the pressure relationship between the two circulations. (3) A direct shunt at atrial level was demonstrated in 84% of the cases, right atrial samples being on the average 13 % more saurated than samples from the superior vena cava (range 6 to 21%). (4) The oxygen saturation of left atrial samples averaged 81 % in the eight cases in which they were obtained (range 72 to 88). Left ventricular samples were similar when secured, and arterial s...