METHODS Experiments in vivo Guinea-pigs were used throughout. The animalwas anaesthetized byan intravenous injection of a mixture of0. 6% chloralose+ 10% urethane (5 ml./kg). One hour after the injection the abdomen was opened and a piece of ileum close to the caecum, 5-7 cm long, was cannulated at both ends. Tyrode solution was passed through the lumen as in the method described by Biilbring & Lin (1958) for isolated loops of intestine. The fluid flowed from a Marriotte bottle, the height of which could be changed, through a narrow tube at a slow rate intothe oral end of the intestine. This was connected to a vertical glass tube acting as a water manometer, the pressure being recorded, by air transmission, with a float recorder. The caudal end of the intestine was connected to a T-piece, one arm of which led to a valve. As a rubber valve constituted a certain resistance we used a floatvalve made of Perspex, as shown in Fig. 1. Fluid entered at A and lifted the conical Perspex float, which was, however, held in a vertical position by the side arm projecting across the tube. The fluid left the tube from the side arm B. The height of the water column in the tube was maintained by the outlet of the syphon tube.