Thus, in these experiments, we did not demonstrate the phenomenon of adaptive cytoprotection in the canine gastric mucosa. The extent of mucosal barrier disruption, as evidenced by a fall in transmucosal electrical potential difference and an increase in net hydrogen ion flux, was similar in all experiments, both with and without prior exposure to bile. These findings conflict with those cited previously in which adaptative cytoprotection was demonstrated. Perhaps, this is an example of a species specific phenomenon, since prior studies have primarily involved the rat gastric mucosa; but species difference would not explain the discrepancy between our results and the preliminary report by Scheurer and colleagues (10). It is possible that a cholecystogastrostomy is, indeed, cytoprotective, for reasons that may, or may not, be attributable to mild exposure to bile. Many variables, including bile, lecithin and cholesterol exposure as well as possible neutralization of acid, are operative in such a model. In our experiments in which bile salts alone were used, no cytoprotection was observed. To be sure of the accuracy of our observations, we developed three separate experimental designs. Each time we failed to demonstrate adaptive cytoprotection, we initiated another experimental design. After failing to observe adaptive cytoprotection in three different sets of experiments, we are confident that the phenomenon cannot be induced in the dog with bile salt injury.