In a previous paper mention was made of a girl with cyanotic congenital heart disease and a pulsatile tumour under the umbilicus (Astley and Parsons, 1953). This child died in Birmingham Children's Hospital in 1950, but the condition was thought to be a chance monstrosity until 1952, when Dr. Helen Taussig referred in a letter to another example of this condition. In 1955 both Dr. Catherine Neill and Dr. James Brown drew my attention to descriptions of other similar cases. CASE REPORT M. A., the fifth child of a healthy family, was born after a pregnancy complicated by a small hkmorrhage at the third month. A pulsating swelling at the umbilicus was noticed at birth; intermittent cyanosis was first observed six months later. Cyanotic attacks became increasingly severe and frequent, especially when feeding or crying, and she was unconscious when, at the age of 18 months, she was admitted to hospital for the last time in one of these attacks. She was quite well-developed, and uniformly cyanosed, with moderate clubbing of fingers and toes (hkmoglobin, 14-7 g.; red blood corpuscles, 6-8 million). The venous pressure was raised, the liver enlarged, and the lungs slightly congested. Arterial pulses were normal, and the blood pressure in the arms 120/?100 mm. Apical impulses could be felt in the fourth left intercostal space just internal to the mid-clavicular line, and in the fifth right space in the mid-clavicular line; the chest was not deformed. A systolic thrill was felt at the right sternal margin in the second and third spaces and a harsh systolic murmur was heard all over the front of the chest, and faintly between the scapulk. A systolic thrill and a continuous murmur could be detected over a smoothly-rounded, pulsatile swelling immediately deep to the umbilicus. A straight cord, which also pulsated, ran under the skin between the epigastric notch and the upper border of the tumour. The depth of cyanosis was not affected by compressing the tumour or its connections. X-ray examination supported a diagnosis of dextrocardia, but an electrocardiogram, using only the standard limb leads, showed peaked P waves in lead II, marked left axis deviation, and rather wide ventricular complexes, suggesting the possibility of tricuspid atresia with a diminutive right ventricle. No further investigations were possible until after death, when radioopaque material injected into the superior vena cava demonstrated a communication between right and left atria, and outlined the pulmonary veins. An injection into the sub-umbilical tumour showed connections with the aorta and its branches through a ventricle.Necropsy. The only abnormalities were those related to the heart. Externally (Fig. 1) the aorta and pulmonary artery appeared to arise from two ventricles separated by an interventricular groove; the right atrium was dilated. The apex of the heart was fusiform and continued downward as a muscular tube " like an anteater's snout " to an ovoid tumour (27 x 18 x 10 mm.) lying deep to the umbilicus. Unfortunately the reflections of th...