In 1950, a diaphragmatic defect was suggested as the basis for a large pleural effusion suddenly developing in a patient with cirrhosis and ascites. At that time, Williams 1 demonstrated that diagnostic pneumoperitoneum quickly resulted in the development of a pneumothorax in such a patient. In 1955, Emerson and Davies2 demonstrated a small opening in the tendinous portion of the diaphragm at the autopsy of a similar patient. In 1964, Okuda et al3 found similar defects at autopsy in two such patients, and added one more autopsy case in 1967.4 In 1966,5 we published evidence that the large pleural effusion of cirrhosis was not caused by movement of ascites through diaphragmatic lymphatics, as had been suggested for the pleural effusion of Meigs' syndrome.6 We concluded that an opening existed in the diaphragm that allowed ascites to escape into the pleural cavity. These openings were detected