The importance of the roentgen examination in the diagnosis of perforations of the hollow viscera has been repeatedly stressed and evaluated. In all cases in which a rupture of the stomach, small bowel, or colon, whether of spontaneous or traumatic origin, is suspected, the diagnosis can be made relatively certain by means of roentgen study. The roentgen findings depend upon the occurrence of a spontaneous pneumoperitoneum, gas from the gastro-intestinal tract escaping into the free peritoneal cavity. In all the literature on this subject it has been stressed that it is necessary to have the patient in the upright position in order that the gas may rise to the upper portion of the abdomen, coming to lie just inferior to the diaphragm. Thus the shadow of the gas can be made out by contrast with the diaphragm above and the denser shadows of the solid viscera below. When the upright position is impossible, it has been suggested that films might be made in the left lateral decubitus position, a postero-anterior exposure being undertaken, so that the gas will rise under the lateral abdominal wall and again manifest itself by contrast with the surrounding tissues. In a series of cases of perforation of the colon from various causes we have observed another roentgen sign of pneumoperitoneum which is particularly valuable because it is manifested in the ordinary scout roentgenograms of the abdomen made with the patient in the supine position. Essentially, this finding consists in the ability to visualize, on the film, the outer as well as the inner wall of the bowel. It is curious that in the extensive literature on artificial pneumoperitoneum there is no specific mention of this finding, although it has undoubtedly been observed repeatedly. It appears to occur only in those cases in which relatively large quantities of gas and some fluid have entered the peritoneal cavity. Under such circumstances there is a sufficient quantity of contrast medium present to separate the individual loops of bowel from each other. The loops themselves are usually greatly distended with gas because of the peritonitis which is present. As a result, it is possible to observe both the contour of the inner wall of the bowel and of the outer, the thickness of the wall itself being quite apparent. In the ordinary case of ileus it is relatively easy to observe the inner wall of the bowel because of the contrast between it and the gas which distends the bowel. Likewise the thickness of the bowel wall can be approximately ascertained because, with two loops of bowel lying in close juxtaposition to each other, the gas filling each loop delineates its wall; thus the thickness between the two layers of gas ordinarily represents the thickness of both bowel walls. In case of effusion into the peritoneal cavity, the diagnosis can often be made because of the increased thickness of the shadow separating the layers of gas, which results from the entrance of fluid between the two loops.