THE continuance of an ineffective artificial pneumothorax attempted over a long period of time is useless. The dangers of extension of the tuberculous process from the open cavity into the affected lung, as well as cross-firing into the so-called good lung, are great. Hemorrhages cannot be prevented, and spontaneous pneumothorax and subsequent empyema are especially menacing in cavities localized just below the surface of the lung which are held open by adhesions. The high-pressure pneumothorax, with and without artificial fixation of the mediastinum, is sometimes successful in such cases if administered with small doses of air given at short intervals, but its possible dangers cannot be underestimated. A tear near the thoracic wall, followed by uncontrollable hemorrhages into the pleural sac, or a tear into the lung tissue, followed by spontaneous pneumothorax, cannot be prevented, even if the best technical skill is assured. Therefore, electrosurgical severance of pleural adhesions should be much more often practised than heretofore, in order to convert an ineffective artificial pneumothorax into an effective one. The open method by thoracotomy is dangerous. The percentage of pleural empyemas in such cases is high, and the shock to the weakened tuberculous individual is, in many instances, too severe to justify such a procedure. Since thoracoscopy is relatively harmless, intrapleural closed pneumolysis should be the treatment of choice in such instances. What Are the Roentgenologic Indications for This Type of Operation? Of What Aid Could the Roentgenologist Be to the Thoracoscopist? Every experienced phthisiologist knows that pleural adhesions are present in almost every case of artificial pneumothorax. They are almost physiologic, and disappear or give way in many instances even under negative intrapleural pressure. But the picture changes as soon as the adhesions become organized, forming fibrous strings, cords, bands, or fans, which prevent the tuberculous cavity from collapsing, or which, by contraction, cause a re-expansion of the collapsed lung and reopening of the cavity already closed by an effective collapse. Here, the aid of the roentgenologist is invaluable. It is up to him to study carefully large series of chest films taken at short intervals, and to describe every single adhesion and its character, often resorting to lateral films taken at different angles, in order to differentiate between cords and fans. These latter frequently prove to be unsuitable for closed intrapleural pneumolysis on account of the danger of lung tissue or cavities extending into the adhesions. He will often be able to call the phthisiologist's attention to the fact that, in his attempt to stretch pleural adhesions, he had accomplished nothing but a mediastinal hernia with its consequences, such as disturbance in the lesser circulation, downward pressure of the diaphragm, etc.