2. It is shown that, even in blood, where, with safe concentrations of dye, there seems to be a significant excess of protein over dye, the stoichiometric excess of dye required by the above mechanism is available.

3. Experiments are presented indicating that the time required for such a bacteriostat to act is very short, being probably conditioned largely by speed of mixing or of penetration.

4. Apparent disappearance of dye from blood stream need not mean that the blood has lost bacteriostatic value.

5. Data are presented indicating that the behavior of dyes in causing flocculation of organisms is affected by the same factors and in the same way as in inhibiting growth.

6. Direct evidence of ionic combination between dye ion and protein ion is presented by noting conductivity decrease when the two ions are mixed, and also noting the displacement of H-ion from unionized protein by dye cation, or of OH-ion by dye anion.