It is a familiar fact that after three injections of antidiphtheritic anatoxin, properly made, a Schick reaction that is positive before vaccination becomes negative in about 95 per cent of the cases. The Schick test, however, which is useful in practice, shows only whether the serum does or does not contain a proportion of specific antitoxin corresponding to one thirtieth of an Ehrlich unit, this proportion being thus far considered as necessary and sufficient for protecting man against a toxic injection determined by the diphtheria bacillus. Interest is not lacking, however, in additional information concerning the exact value of the immunity produced by anatoxin and the fluctuations that this immunity may undergo during the years following vaccination. As soon as one of us (G. R.) suggested anatoxin for the active immunization of man against diphtheria (1923), Martin,1 Darré, Loiseau and A. Laffaille showed, by determinations of the antitoxin present