394 PATHOGENIC BACTERIA. possess this property, and unless trypsin is present it must be dependent upon the proteolytic activity of other bac- teria for a suitable form of proteid food. Perhaps these bacteria form an albuminate molecule, which like leucin and tyrosin cannot be broken up into indol, and thus there might be caused an important modification of the metabolism of the colon bacillus, which might have either an immediate or remote influence upon its acquisi- tion of disease-producing properties, for our own experi- ments indicate that the power to form indol, and the actual forming of it, are to some extent an indication of the possession of pathogenesis.'' To the laboratory animals the colon bacillus is patho- genic in varying degree. Intraperitoneal injections into mice cause their death in from one to eight days if the culture is virulent. Guinea-pigs and rabbits also suc- cumb to intraperitoneal and intravenous injection. Sub- cutaneous injections are of less effect, and in rabbits seem to produce abscesses only. When the bacilli are injected into the abdominal cavity a sero-fibrinous or purulent peritonitis occurs, the bacilli being very numerous in the abdominal fluids. The pathogeny of the colon bacillus is clue to irritating, chemotactic substances in its protoplasm. The experi- ments of Pfeiffer and Kolle and Loffler and Abel have proved very conclusively that the poisonous principle is in, and cannot by any means be separated from the bodies of the bacteria. Frequent transplantation lessens the virulence, passage through animals increases it. Numerous observers have found that cultures of the bacillus obtained from cholera, cholera nostras, and other intestinal diseases are much more pathogenic than those obtained from normal feces or from pus. Cumston,1 from a careful study of thirteen cases of sum- mer infantile diarrheas, comes to the following conclu- sions: 1 International Medical Magazine, Feb., 1897.