1. The facts described in this paper are intended to bear upon the pathogenesis of colitis in man and animals.
2. The toxin of the Shiga dysentery bacillus is liberated from the bacillus through the process of autolysis.
3. In rabbits the toxin is not absorbed directly in an active form by the gastro-intestinal tract; in man, however, absorption of an active poison does take place from the intestine.
4. The toxin is excreted in rabbits, and probably in man as well by the intestine, chiefly probably by the large intestine, which being injured by the act of elimination, reacts by the development of inflammation, etc.
5. In rabbits the characteristic action of the toxin depends upon the integrity of the biliary secretion into the intestine. When the bile is prevented from entering the intestine, either by ligature and section of the duct or by establishment of a biliary fistula, no lesions whatever of the large intestine appear, or they are inconsiderable in extent.
6. The loss of toxin through a biliary fistula does not prevent in rabbits the lethal effects which are caused, apparently, by a nervous poison. The tying off of the bile duct seems to prevent the passage of the toxin, in amounts sufficient to cause the large intestinal lesions, into the blood. The liver, therefore, in this condition tends to hold back the toxin from the general circulation.
7. The peculiar effects of the toxin on the large intestine in rabbits is not produced at once, but would appear to depend upon successive acts of excretion of the poison by the bowel.
8. The establishment of biliary fistula reduces the intensity of action of corrosive sublimate upon the large intestine in rabbits; and the lesions of ricin poisoning in these animals are also modified by this operation.
9. Dysentery toxin is destroyed by peptic digestion, and also, though probably more slowly, by tryptic digestion. The absence of power of the toxin to cause poisoning in rabbits when it is brought directly into the lumen of the intestine, is not explained by the destructive action of trypsin.
10. The character of the histological changes in the cæcum of rabbits caused by the dysentery toxin points to an action upon the substance and not primarily upon the surface of the intestine.