In many bacterial diseases, agglutinins for the particular organism are present in the blood serum of the affected animal. Consequently if the blood serum of the animal be mixed with a suspension of the organism supposed to be the cause of the disease and the latter be agglutinated, one is justified in considering it the causative agent, provided certain necessary conditions are fulfilled. In the first place it must be remembered that the blood of normal animals frequently contains agglutinins (“normal agglutinins”) for many different bacteria when mixed with them in full strength. Hence the serum must always be diluted with physiological salt solution (0.85 per cent.). Further, closely related bacteria may be agglutinated to some extent by the same serum. It is evident that if they are closely related, their protoplasm must contain some substances of the same kind to account for this relationship. Since some of these substances may be agglutinogens, their introduction into the animal body will give rise to agglutinins for the related cells, as well as for the cell introduced. The agglutinins for the cell introduced “chief agglutinins,” will be formed in larger quantity, since a given bacterial cell must contain more of its own agglutinogen than that of any other cell. By _diluting the blood serum_ from the animal to be tested the agglutinins for the related organisms (so-called “coagglutinins” or “partial agglutinins”) will become so much diminished as to show no action, while the agglutinin for the specific organism is still present in an amount sufficient to cause its clumping. _Agglutinins are specific for their particular agglutinogens_, but since a given blood serum may contain many agglutinins, the _serum’s specificity for a given bacterium_ can be determined only by diluting it until this bacterium alone is agglutinated. Hence the necessity of diluting the unknown serum in varying amounts when testing against several known bacteria to determine for which it is specific, _i.e._, which is the cause of the disease in the animal. The agglutinins in the serum may be removed from it by treating it with a suspension of the cells for which agglutinins are present. If the “chief” cell is used all the agglutinins will be absorbed. If related cells are used, only the agglutinins for this particular kind are removed. These “absorption tests” furnish another means of determining specificity of serum, or rather of determining the “chief agglutinin” present.