326 BACTERIOLOGY. 



minimal fatal dose, and injection of the minimal 

 fatal dose itself is now borne without reaction. 



Behring and Ehrlich made the estimation 

 outside of the animal body by mixing a definite 

 quantity of the toxic solution obtained from 

 cultures with graduated quantities of the pro- 

 tective serum. In this way mixtures of vary- 

 ing grades of action are obtained, first those 

 that manifest unmistakable toxic action, next 

 those by which the course of the disease is 

 retarded, then those in which only local pheno- 

 mena make their appearance, and finally those 

 in which the quantity of serum injected at 

 the same time with the toxin completely in- 

 hibits the action of the latter. A serum o.icc 

 of which renders entirely impotent ten times 

 the minimal lethal dose is termed normal serum 

 One cubic centimeter of this normal serum is 

 called an immunization unit. If o.oicc. of a 

 serum confers protection, the serum would 

 possess ten immunization units. 



Roux's method of calculation starts with the 

 living animal. He computes the body weight 

 of the animal which is protected against the 

 minimal lethal dose by a unit volume of serum. 

 He employs ice. of a toxin solution, o.icc. of 

 which kills a guinea-pig of 500 grams in 24 

 hours ; this starting-point, therefore, is the 

 same as that of Behring and Ehrlich. If ice. 



