INTRODUCTION Foster and Whipple1have reported a method for the quantitative analysis of blood fibrin which is suitable for clinical investigation because it is relatively accurate, and requires but 5 c.c. of blood on analysis. Their work demonstrated that small variations in the level of blood fibrin occur in normal dogs while disease (experimentally produced) in these animals causes very wide and rapid fluctuations of this blood protein. Injury to any tissue, excepting the liver, was found to cause an increase in the blood fibrin content. This increase was constantly observed, for example, when sterile abscesses were produced by the subcutaneous injection of turpentine, when the animal was exposed to massive doses of deep roentgen rays or when suffering from "distemper," a common acute respiratory infection in dogs. If the liver is severely injured, as by chloroform or phosphorus, a decrease and not an increase in blood fibrin occurs. However,