It has been shown elsewhere (1) that carbon dioxide is superior to air as a contrast medium for roentgenologic examination of the colon, having the following advantages: 1. Carbon dioxide is absorbed more rapidly than air. This reduces the duration of discomfort for the patient. 2. If the gas should accidentally enter a patulous bleeding vessel, there is no danger of embolus, since carbon dioxide combines with hemoglobin and goes into solution in the plasma immediately upon entering the blood stream. 3. In case of accidental perforation of the bowel, carbon dioxide is quickly absorbed from the abdomen or planes of dissection. The following simple experiment was performed to confirm facts already established and to obtain graphic proof concerning the rates of absorption of carbon dioxide and of air from the colon. A normal, healthy young woman was selected from a group of volunteers. The colon was first prepared for the examination by the administration of a laxative. A control film prior to the introduction of gas into the colon showed the latter to be virtually empty (Fig. 1). The flowmeter on a tank of compressed carbon dioxide was set for a flow of one liter a minute and the gas was allowed to enter the colon under fluoroscopic observation. When the colon appeared to be properly distended, the gas supply was shut off. The elapsed time, as determined by a stopwatch, was thirty-six seconds. Thus, approximately 600 c.c. of carbon dioxide were introduced into the colon. A film of the distended colon was made immediately and subsequent films were obtained at fifteen-minute intervals for one hour, and an additional film at two hours (Figs. 2–7). The subject was requested to retain the gas, since any discharge by rectum would vitiate the accuracy of the study. While the young woman was thoroughly co-operative and dependable, the writer imposed a further aesthetic control by remaining present throughout the period of observation. The entire procedure was surprisingly free from subjective discomfort. Since the gas was supplied in a smooth, continuous flow from a tank, there was no undue stretching of the bowel and no spasm. Almost all of the gas was absorbed from the colon in forty-five minutes, and, in accordance with the known physiologic activity of the colon, the gas in the proximal portion was absorbed first. A small amount of gas remained in the distal colon at two hours. This is believed to be due entirely to a voluntary effort to prevent its escape by rectum; in routine roentgenologic practice, it would have been expelled. The amount of gas remaining, however, does not appear to be much greater than the normal residual amount in the colon before the introduction of carbon dioxide (Figs. 1, A and 7, A). The experiment was repeated on the same subject three weeks later, with air instead of carbon dioxide. The bowel was prepared as for the first examination.