The original and subsequent early reports of aneurysmal dilatation of the left atrium (Owen and Fenton, 1901; Jones, 1901;Owen, 1902;Lutembacher, 1917Lutembacher, , 1918 Emanuel, 1923; Shaw, 1924;Schott, 1924; East, 1926; Cowan, 1926) were collected together and analysed by Bramwell and Duguid (1928). They concluded that two main factors were concerned in the aetiology of this condition. A "passive factor" was gross fibrosis of the wall of the left atrium: this was usually, but not always, due to chronic rheumatic heart disease that had predominantly affected this chamber. An "active factor" was increased flow of blood into the inelastic fibrous left atrium commonly from a hypertrophied left ventricle through an incompetent mitral valve. They stated that " one cannot fail to be impressed by the fact that the condition is so constantly associated with great ventricular hypertrophy," and this applied in their collected series of cases not only to the left ventricular hypertrophy associated with mitral incompetence but also to right ventricular hypertrophy. Of the 20 cases in their series the state of the right ventricle at necropsy was specified in 13, and in 10 of these it was said to be hypertrophied. Wood (1954) also stated that "mitral incompetence is a more important cause of left atrial dilatation than mitral stenosis ". In the face of these data and opinions we thought it would be of interest to see how many of our cases coming to necropsy, with extreme dilatation of the left atrium, had mitral incompetence, right ventricular hypertrophy, and pulmonary vascular disease. SUBJECTS AND METHOD Necropsy material was available from nine patients with extreme dilatation of the left atrium who had died of chronic rheumatic heart disease. Five were men (Table).At necropsy the thickness of the right ventricle and the size of the mitral valve orifice had been measured. Paraffin-embedded blocks of lung were available. Thin sections of lung were stained to demonstrate elastic tissue by the Lawson modification of the Weigert-Sheridan method and counterstained for collagen and muscle with van Gieson's reagents. Measurements were made of the external diameter of the media and average thickness of the media of pulmonary arteries up to 300t in diameter, which had been sectioned as nearly as possible transversely. Only vessels that were virtually circular in transverse section were measured; this method of selection decreased the number of arteries included in each study but avoided error in the measurement of the vessel diameter. Approximately 15 to 40 small muscular pulmonary arteries were measured in each case. The external diameter was taken as the mean of two measurements, at right angles to each other, of the distance between diametrically opposite points on the external elastic lamina. The medial thickness was estimated as the mean of four measurements taken at points approximately equally spaced around the vessel wall. From these data the thickness of the media was expressed as a percentage of the external diameter ...