ANTITOXINS AND RELATED ANTIBODIES 139 



the least amount of antitoxin which, when mixed with 100 M. L. 

 D. of the toxin, would prevent a 250-gm. guinea-pig injected with 

 the mixture from dying within four days. This implied the keep- 

 ing of the toxin of a certain strength for the standardization of 

 antitoxin. The toxin was found to be unstable, and the same 

 toxin would at different times yield different results. Furthermore, 

 antitoxins are neutralized by toxoids as well as toxins. The toxin 

 cannot be, preserved for long periods without deterioration, as 

 would be necessitated if used as a government standard. Ehrlich, 

 therefore, made use of a toxin which he had studied in his labora- 

 tory to standardize a large quantity of serum. This serum he 

 dried in a current of warm air in a partial vacuum. As a result, 



Fig. 68. A trocar inserted into the jugular vein of the horse. The com- 

 pressor causes a noticeable engorgement of the vessel (Kretz in Kraus and 

 Levaditi). 



he secured a considerable amount of serum in the form of dried 

 scales. The number of immunity units (U. I.) per gram of this 

 dried material was very accurately determined. Exactly equal 

 amounts by weight of this standard serum were placed in each 

 of a large number of special tubes. The serum was placed in 

 one arm and phosphorus pentoxid (P 2 O 5 ), an active dehydrating 

 agent, in the other. The air was then exhausted as completely 

 as possible by an air-pump and the tubes sealed. They then were 

 placed in a dark refrigerator, where a constant temperature was 

 maintained. The antitoxin under these conditions was found 

 to retain its potency undiminished for a long period. Ehrlich 



