1. Far more information about the effect of serum or other substances upon the germicidal action of soaps can be obtained by determining the germicidal titers over a wide range of pH than by determining the titer at a single pH. In this way a characteristic curve for each test substance is obtained.

2. The curve for a particular concentration of serum bears a definite relationship to the curve for salt solution (buffer) alone. Wherever the titer in salt solution is high, very small amounts of serum greatly diminish that titer. Wherever the titer in salt solution is low, small amounts of serum leave the titer unchanged. Thus small additions of serum flatten the curves and make them more nearly horizontal. If further large amounts of serum are added, a further reduction in titer takes place at all reactions.

3. The calcium of serum has only a very slight effect upon the soap titer.

4. The protein of serum is probably inhibitory to soaps; but the curve for partially defatted serum, and the curves for other protein substances tested, do not run parallel to the serum-soap curves.

5. The various lipoids that are known to be present in serum are inhibitory to the action of soaps, both as emulsions and as clear solutions.

6. The action of serum upon soaps may be regarded as a complex reaction, in which lipoids, protein, and, to a lesser extent, calcium salts take part. Their effect is due to the fact that these substances, by combining with the soaps, remove them from the field of germicidal action.