The permeability of the gastric mucosa of dog.

1. Tritiated water, [(14)C]urea, [(14)C]thiourea, [(14)C]sucrose and [(59)Fe]-haemoglobin were used to study the permeability of a semi-isolated piece of the great curvature of the canine stomach.2. The osmotic pressure of the solutions placed in contact with the secretory surface of the epithelium was changed by means of dextrose or urea. The mucosa behaved as a semi-permeable membrane, meaning that water flowed under gradients of osmotic pressure. Regardless of the solute used, about 45 x 10(-6) ml. of water flowed/cm(2)/min under a gradient of one atmosphere.3. The permeability constants of the probing molecules were determined under zero net volume flow obtained by placing isosmotic dextrose or isosmotic urea in the chamber. The constants decreased as the molecular volume of the probing molecules increased.4. The transport of all the non-electrolytes across the epithelium decreased significantly when the chamber contained isosmotic dextrose. Basically, this effect seems to be a result of the reduction of the area available for diffusion caused by the high molecular volume of dextrose.5. The increased hindrance to diffusion of the probing molecules caused by the added solutes is considered as good evidence that the probing molecules diffuse by way of pores filled with water.6. The equation derived by Renkin (1954) fits the results obtained if we assume that the equivalent membrane has pores of at least two different radii. The calculated radii vary somewhat with the solute placed in the chamber, though about 88% of the area available for diffusion consists of pores with radii smaller than 2.5 A.7. The equivalent pore radius, calculated from Kedem & Katchalsky's (1961) formula for pores of one single radius, contradicts some experimental findings. Once again, the results obtained would be reproduced more accurately by an equivalent membrane pierced by parallel pores of at least two different diameters.8. A procedure is suggested for calculating the proportion of pores of different radii. It seems likely that the pore radii vary in a continuous distribution from the large pores which allow the diffusion of haemoglobin, to pores hardly permitting the passage of water. The wide pores would form a small fraction of the total area available for the diffusion of water.