B. In normal serum, hydrophilic protein is adsorbed, forming a protective film around the individual lipoid particles, with a corresponding change in the cataphoretic potential and the isoelectric point towards those of serum protein, the degree of shift depending upon the extent of the adsorbed film. The critical potential, however, is not affected, and the lipoid remains as stable away from its isoelectric point as in the absence of serum. The water-soluble film of unchanged protein is readily removed by washing, and does not prevent the subsequent combination of the underlying lipoid with the specific component of syphilitic serum.

C. When the lipoid antigen is added to syphilitic serum, in addition to this loose adsorption of normal protein it combines more or less irreversibly with a specifically altered fraction of the serum globulin (reagin), demonstrable in the washed precipitate both chemically and by sensitization experiments. Like adsorbed normal serum, it depresses the surface potential and causes a shift in the isoelectric point; but there the similarity ends. The reagin-globulin is rendered water-insoluble by its firm combination with the lipoid, exactly as any antibody is denatured upon combination with its specific antigen (bacteria, red cells, or dissolved protein). The hydrophobic films of reagin have five times as great an affinity for each other as the original lipoid surfaces; accordingly, the critical potential is raised from its original value of 1 to 5 millivolts to 10 to 15 millivolts, that of particles of denatured globulin or of any antigen-antibody complex, and relatively small quantities of electrolytes (at serum pH, cations) suffice to depress the stabilizing potential below this critical level, with resultant aggregation and flocculation. In brief, a specific globulin combines with the colloidal particles of the antigen, conferring upon them the unstable properties of a suspension of denatured protein.

Like the antibody film on bacteria, or red cells, and unlike normal adsorbed protein, the reagin globulin on the lipoid particle can adsorb ("fix") complement. When this protein film is destroyed by heat-coagulation, the complement-fixing property is lost; concomitantly, the specific groups of the lipoid having been freed from the closely adherent reagin, the antigen becomes again active, able to react with more syphilitic serum.

These changes in the properties of reagin globulin upon its combination with the lipoid antigen (denaturation) are in every sense analogous to those effected in any antibody by its specific antigen, and are probably due to the same, as yet unknown, factors. It has been suggested for bacterial and red cell "agglutinins" and protein "precipitins," that the groups of the antibody determining its specificity are also those which endow it with its hydrophilic properties; when these combine with antigen, residual free hydrophobic groups determine the surface properties of the complex. The same tentative hypothesis may be offered for the denaturation of reagin globulin by the lipoid antigen.

The complete analogy between the flocculation reactions for syphilis and the so-called specific reactions (bacterial and red cell agglutination; protein precipitation) suggests that like agglutinins, precipitins, etc., reagin globulin represents an antibody response to products of infection.