Cadeac even advocates the theory that the same germ possessed of greater or lesser virulence, is always present in ordinary stables and horses, and habitually causes in exposed or debilitated animals an ordinary fibrinous pneumonia with no perceptible tendency to transmission by contagion; that, in other cases when a considerable number of horses have their defensive powers impaired, it gains a wide extension; and, that in some such cases, the germ that has been living as a comparatively harmless saprophyte, suddenly acquires an unwonted potency, and breaking down the barrier of partial immunity, attacks exposed animals on a large scale and irrespective of weather perturbations, or debilitated conditions. He quotes from Trasbot instances that seem to support this hypothesis, which is not at all in disaccord with the habits of bacterial life, yet we require a solid basis in bacteriological experiment to make it unassailable. _Bacteriology._ Siedamgrotzky (1882) found in the hæmorrhagic centres in the affected lung and in the pleural exudate _micrococci_. Dieckerhoff (1882) and Mendelsohn (1883) found in the pleural exudate streptococci. Chain cocci were also found by Peterlein (1884), Perroncito (1885), Delamotte and Chantemesse (1888), and Mosselman and Lienaux (1893). Schütz (1887) found a diplococcus which he studied very fully and this is corroborated by Lustig’s ovoid bacterium, by Cadeac’s micrococcus and diplococcus. In Dr. V. A. Moore’s cultures at the N. Y. S. Veterinary College cocci were found constantly in pure culture, sometimes as a diplococcus, but under slightly altered conditions the streptococcus form predominated. As the difference between two, and three or more cocci in chain form is merely a question of early or late separation of cocci which multiply in line, the apparent discrepancies in the above observations do not imply any real difference in the microbe. Inoculated in pure cultures the Schütz diplococcus killed mice in 24 to 48 hours with enhanced virulence of the germ. In the _rabbit_, _subcutem_, it usually killed in 24 to 48 hours, but some survived; _intravenously_ or _intratrachealy_ it killed more certainly and speedily and in either case with pleural, pericardial and even peritoneal lesions. In the _Guinea-pig_, _subcutem_, it caused extensive effusion, and death in two to six days with chest lesions. In the _dog_ it caused hyperthermia, but no marked lesion and no mortality. In the _horse_ there were no infectious resultant lesions. A pure culture thrown into the lung tissue of an old horse at the N. Y. S. V. College, determined an extended pleuritic adhesion and lung hepatization. The age of this subject was opposed to any marked susceptibility. The apparent immunity of the horse in Schutze’s cases might depend on the insusceptibility of the animals selected during or after an epizoötic, or on the absence of the predisposing causes so strongly insisted on by Cadeac. Rats, chickens and pigs proved immune.