Summary: The protective effect of nitrogen mustard‐induced neutropenia on the development of the GSR was studied. No coagulation defect was found in neutropenic rabbits and simultaneous administration of Thorotrast and endotoxin did not produce intravascular coagulation or renal cortical necrosis. Infusion of neutropenic rabbits with suspensions of viable granulocytes derived from peritoneal exudates reinduced susceptibility to both intravascular coagulation and renal cortical necrosis. It is concluded that the presence of granulocytes is essential for the production of intravascular coagulation by endotoxin.

Summary: Preparation of rabbits with 6 ml./kg. of Thorotrast, administered intravenously, permits the induction of intravascular clotting and the production of renal cortical necrosis with simultaneous injections of E. coli endotoxin. With this dose of Thorotrast, a dose‐response relationship was observed between the fall in fibrinogen after 4 hours and doses of endotoxin ranging from 5 to 50 μg./kg. body weight. With such preparation, renal cortical necrosis occurred in 46 per cent of animals receiving 100μg./kg. of endotoxin and increased to 80 per cent in animals receiving 100 μg./kg. Thorotrast preparation with 3 ml./kg. permits intravascular clotting to occur with a single injection of endotoxin administered simultaneously. However, no relationship between endotoxin dose and fibrinogen depletion was observed and renal cortical necrosis developed only with relatively high doses of endotoxin. An efficient and predictable experimental model for studying both intravascular clotting and renal cortical necrosis is described by these observations.