In the byeways of surgery there can be few conditions more dramatic in their presentation and more terrible in their symptoms than spontaneous perforation of the oesophagus. No case has yet been treated successfully, and diagnosis has only been achieved in a very few before death, and yet there is no fundamental reason why this unsatisfactory position should not be improved in the future. Several things are essential to success: firstly, a knowledge that the accident can and does occur; secondly, a knowledge of the symptomatology; and, thirdly, an early diagnosis. Given these, I am convinced that surgeons will be able to save some of these patients by combining the principles, already well established in the cases of abdominal perforations, with those relevant to thoracotomy.—N. R. Barrett.1 WITH the preceding words, Barrett, in 1946, reported the fact that of 50 patients with spontaneous rupture of the esophagus whose cases