SUMMARYThe initial low frequency component of the first heart sound, 'M', has been studied in normal subjects, and in patients with various prosthetic mitral valves and with mitral stenosis, using simultaneous low frequency phonocardiography, echocardiography, and apex cardiography. These techniques showed 'M' to have a constant morphology in preisovolumic systole. In mitral stenosis, 'M' and the preisovolumic 'presystolic' murmur appear to be the same phonocardiographic phenomenon. While 'M' was present in sinus rhythm, augmentation of this normal vibration occurred particularly during the short cycles of atrial fibrillation. Leaflet coaption and movement of the ventricular wall as detected echocardiographically do not appear to play a role in its pathogenesis but the sound could emanate from the ventricular wall as it tautens and decreases its compliance at the onset of systole.The initial low frequency component of the first heart sound has been studied for more than 150 years (Orias, 1936;Eckstein, 1937;Smith et al., 1941;Lakier et al., 1970;Luisada and Argano, 1971;Lakier et al., 1972a, b;Armstrong and Gotsman, 1973). We hope to show that the sound is more than just an abstruse phonocardiographic finding but has clinical significance in that it is responsible for the short crescendo, so-called 'presystolic', murmur of mitral stenosis in atrial fibrillation, as suggested by Tavel and Bonner (Bonner et al., 1976;Tavel and Bonner, 1976). This sound, which occurs near the peak of the R wave of the electrocardiogram and at the crossover point of the left atrial and ventricular pressures, has been termed 'M' by Lakier et al. (1970) and '0' by Luisada and Argano (1971) (Fig. 1). In contrast, the later higher frequency component of the first heart sound (Ml) occurs at the time of the left atrial C wave (Lakier et al., 1972a, b).The cause of 'M' remains controversial. The sound has been attributed to initial coaption of the mitral valve leaflets and the sound 'Ml' to subsequent tensing of the mitral cusps and chordae