The problem of the reaction of the body fluids in pneumococcus infections has claimed our interest for some time. Together with the grouping of the pneumococci which was begun in our laboratory in the fall of 1916, the following studies were made to determine if possible, the extent and the way in which the body fluids reacted to the pneumococci and to antipneumococcic sera. The materials examined were obtained from patients admitted to the Allegheny General Hospital. Sputum from persons presenting clinical manifestations of pneumonia were sent to the laboratory at the earliest opportunity and the grouping made according to the method of Dochez and Gillespie (1). When the type of pneumococcus to which the infection belonged had been determined, tests were begun for the purpose of demonstrating any defensive antibody formations that might be present in the body fluids. For this purpose, examinations were made of the blood serum, spinal fluids, urine, pleural fluid and, in those instances where autopsies were held, pericardial fluids.