In a recent critical review of the literature on the antigenic properties of the ultraviruses, the writer (1928) called attention to the fact that these agents of disease, as far as they have been studied, seem to exhibit antigenic properties which suggest placing them in the general group of toxins and ferments rather than with such antigens as bacteria and other more highly organized proteinaceous bodies. Since the various facts which have suggested such a probability have been fully treated in the paper referred to, it will be unnecessary to discuss them at length here. The writer will, therefore, confine himself in this paper to a more general outline of the theoretical and other considerations which have prompted the studies that are to appear under the general title indicated. According to Zinsser (1915) anl~igens may be classified into two general classes.