Putting together the results of such experiments as those outlined above, we can set down the following definite facts.
1. The substance of the bacterial cell can be roughly divided into two antigenic entities. One of these is the so called "nucleoprotein" substance, the other the residue substance or soluble material of Dochez and Avery and Avery and Heidelberger, both of which have been repeatedly characterized in preceding papers.
2. Immunization with the nucleoprotein, if such nucleoprotein is rendered free of bacterial bodies or fragments of bacterial bodies by Berkefeld filtration, incites the production only of antinucleoprotein antibodies which, with slight group overlapping, are species-specific but, as determined by the previous studies of Avery and, subsequently, those of Lancefield, are not type-specific to the same degree as the residue antibodies.
3. Immunization with dissolved residue alone leads to no antibody formation whatever. This residue, as indicated in several of our own previous studies, represents the haptophore group upon which specificity depends and which, in the simple process of solution, is disrupted from another substance together with which it represented a complete antigen in the antibody-forming sense.
4. The formation of specific antiresidue antibodies is apparently dependent upon the injection of morphologically formed elements, at least as far as experiment can determine at the present time; for, as in the pneumococcus experiments, the most available process of solution and the injection of all the materials so obtained from the whole bacteria fails to yield antiresidue antibodies, as though in the mere process of dissolving the residue haptophore group were dissociated from its association with the larger molecule to which, in the whole bacteria, it lends specificity.
5. While antiresidue antibodies are only formed when such undisrupted bacterial cell substances are present in the immunizing substance, immunization with whole bacteria, even when attempts are made to preserve them from solution by formalin, leads to the formation of both antiresidue and antinucleoprotein antibodies, probably because a certain amount of solution inevitably takes place after injection within the animal body.