THE MICROBIAL TOXINS. 75 



Nevertheless antidiphtheria serum must not be 

 considered as an antidote; and in pathological 

 diphtheria, the more serum is required the later 

 it is used.* In certain cases, if employed too late, 

 it may prove ineffective. 



The preventive action of the serum is remarkable. 

 In 10 ooo inoculated cases Behring and Ehrlich 

 have had but 10 cases of diphtheria, and these 

 were, moreover, of a benign character. The dura- 

 tion of the immunizing action appears to be from 

 three weeks to two months. 



This diphtheria antitoxin was first prepared 

 by Guerin and Mace f by adding to the anti- 

 diphtheria serum a large volume of alcohol, wash- 

 ing the precipitate, and drying it in a vacuum. 

 It is soluble in water, and loses its activity when 

 heated to 65 C. Wassermann | has proposed 

 to extract it from the milk of immunized animals, 

 by first coagulating the milk by rennet in the 

 presence of sodium chloride, filtering, and removing 

 the fat from the clear liquid by means of chloroform. 

 After decanting, the clear solution obtained is 

 precipitated by adding to it 30 to 33 per cent, of 

 ammonium sulphate. The precipitate is dried in 

 a vacuum on a polished porcelain slab after having 



*BAYEUX: These de Doctoral, Paris, 1899. 

 t Compt. rend, de /' Acad. des Sc., Apr. 5, 1895. 

 Zeitschr. fur Hygiene, xvm, p. 235. 



