At the present time, toxin-antitoxin is undoubtedly the most popular agent in America for producing an active immunity against diphtheria. The original preparation, first recommended by von Behring 1 in 1913, was an underneutralized mixture of toxin and antitoxin, so combined that tests made on guinea-pigs showed a slight excess of toxin. Von Behring worked on the hypothesis that the slight excess of toxin had an immunizing action. This theory, however, was disproved by Park,2 who showed the immunizing effect of fully neutralized mixtures of toxin-antitoxin, and also by Loewenstein and Bousson,3 who used over\x=req-\ neutralized mixtures. The latter investigators showed, further, that immunity produced by the use of toxin-antitoxin mixtures develops later than the immunity induced by the use of free toxin. They concluded that a dissociation of the combination of toxin and antitoxin took place in the body, and that the toxin which was gradually liberated stimulated the production of immune bodies.Hundreds of thousands of children have been treated with mixtures of toxin-antitoxin, especially in America, where a high percentage of negative control Schick tests have been reported-from 68.7 to 98 per cent by Park and Zingher.4 An occasional death from the use of toxinantitoxin has been reported, though the number is insignificant as compared with the thousands of children who have been successfully immunized without deleterious effects.