Recent developments in the use of streptotoxins (scarlatina, erysipelas) and bacillotoxins (diphtheria) for diagnostic and immunizing purposes in both naturally immunizing and nonimmunizing diseases have opened up a wide field of speculation as to the relative value of all exotoxins in regard to an induced immunity. Although the virulence of staphylococci is ordinarily low, clinical experience indicates that natural immunity to these organisms is not common. In fact, pyodermas are fairly common, and although they are in many instances only acute transient phenomena, probably because of the patient's ability to manufacture antibodies rapidly, there are sufficient chronic persistent and recurrent instances which, in such persons, would indicate the lack of spontaneous antibody production. In such conditions as sycosis vulgaris and recurrent furunculosis, the development of any agent with specific immunizing properties would be decidedly welcome. The problem is all the more interesting because such conditions as scarlet fever, diphtheria and