A double apical impulse has been described as a physical sign in hypertrophic obstructive cardiomyopathy (Mounsey, 1959; Braunwald et al., 1960;Boiteau and Allenstein, 1961;Pare et al., 1961;Benchimol, Legler, and Dimond, 1963;Cohen et al., 1964;Gillam, Deliyannis, and Mounsey, 1964;Tafur, Cohen, and Levine, 1964). In this study we 45 degrees. They were taken when the breath was held at the end of quiet expiration and a simultaneous electrocardiogram and phonocardiogram were recordedJ as reference tracings. In every patient the form of the impulse cardiogram corresponded closely to the clinical impression on palpation of the prmcordium.Retrograde left ventricular catheterization was performed in all of the patients. In addition, in 9 patients in whom right heart involvement was suspected, right heart catheterization was also performed. Bi-plane angiocardiograms, using an Elema roll-film changer, and also one-plane cine-angiocardiograms were taken during the same investigation.The pressure records thus obtained were available for detailed study. Pressure gradients between the left ventricular cavity and the aorta were measured. Where right heart catheterization had been performed, any pressure gradient between the right ventricle and the pulmonary artery was noted. Three moments in diastole were selected for pressure measurement: diastasis (or the period immediately preceding atrial systole); atrial systole, when the height of the "a" wave was measured; and end-diastole, between the "a " wave and the beginning of ventricular systole. These measurements were selected as affording evidence of increased force of atrial contraction and of diminished compliance of the ventricle.In order to gain an impression of changes in cavity volume throughout the cardiac cycle, three of the cineangiocardiograms were selected for detailed analysis and sequential changes in ventricular cross-sectional area were measured. The three cine-angiocardiograms selected were free from ventricular ectopic beats, tachycardia, or technical imperfections which made the other cine-angiocardiograms unsuitable for detailed study. The cine-films were taken at a speed of about 40 frames a second, the exact speed in each run being arrived at by dividing the number of exposed frames into the duration of x-ray exposure, which was recorded alongside a simultaneous electrocardiogram. Individual frames could thus be accurately related to the electrocardiogram and, in a more general way, to hemodynamic events recorded during the same investigation, the