Summary and Conclusions A toxin has been produced from active growing cultures of pneumococcus which destroys leucocytes as demonstrated by the Neisser and Wechsberg method. This leucocidin has been demonstrated in aerobic and anaerobic cultures, from growths of both virulent and avirulent strains and from types I, II and III. It is believed that a peroxide which is formed in the aerobic cultures modifies the potency of the toxins through oxidation. The peroxide appears earlier and is of greater concentration in cultures of the virulent strains; this may account for the avirulent cultures apparently containing toxins of as great a potency as the virulent strains, although they may actually be weaker. The leucocidin is readily oxidized; it loses strength in a few days under vaseline seal at 5°C; it is not injured at 70°C. for one hour. Evaporation in vacuum at 5°C. does not deteriorate it; and the toxicity may be enhanced three to four times by the addition of laked red blood cells. The soluble specific substance of pneumococcus type I was found to be destructive for leucocytes: 0.5 mgm. to 15,000,000 leucocytes, but this polysaccharide is not identical with the leucocidin. Rabbit clasmatocytes are more resistant to the pneumococcus leucocidin than are rabbit polymorphonuclears, and leucocytes of the naturally resistant dog are less affected than the leucocytes of the rabbit. Organ tissues, as exemplified by liver and adrenal, are destroyed by this pneumococcus toxin, the adrenal being less resistant than the liver. Hemotoxin was produced from an avirulent strain of pneumococcus type III in cultures from four hours to three weeks old. As it was present only in cultures of twenty-two hours and older when virulent strains were employed, it would seem that the peroxide hindered its production in young growths. Apparently the same conditions in the cultures govern the production and preservation of the hemotoxin and leucocidin, for where the most potent leucocidin was produced the best hemotoxin was found. But the leucocidin is separable from the hemotoxin by its greater thermostability and its appearance in cultures where no hemotoxin can be demonstrated. The leucocidin filtrates are not toxic for mice in 2 cc. amounts injected intraperitoneally, and repeated doses give very slight protection. Ten cubic centimeters of the filtrates injected intrapleurally into rabbits does not produce any toxic symptoms and an avirulent strain of pneumococcus injected simultaneously does not become virulent. Sterile supernatant fluids of pleural exudates of rabbits with empyema while in themselves not toxic, upon repeated doses intravenously, produce an antileucocidin. Normal rabbit serum does not neutralize the pneumococcus leucocidin, but the majority of human sera tested possessed a neutralizing property.