Since 1921 when Friedmann1 reported that artificial pneumothorax was a valuable procedure in the treatment of lobar pneumonia, isolated reports of various observers have been published that substantiate his claims. Recently Lieberman and Leopold2 reviewed the results of fifty published cases and found a mortality of only 6 per cent. More striking than the low mortality figure is the dramatic symptomatic relief which is consistently produced by this measure. Coghlan3 reported that with pneumothorax he was able to produce a crisis which simulated a natural crisis. The improvement was maintained by the employment of subsequent refills. Support for these optimistic clinical results is presented by Lieberman and Leopold, who carried out experiments with dogs. Producing lobar pneumonia by the Robertson method, they noted a mortality of 16 per cent in the group treated by pneumothorax and a mortality of 72 per cent in the control group.