218 INFECTION AND IMMUNITY 



clot and stored in the refrigerator. In order to diminish the chances of 

 contamination, five-tenths per cent of carbolic acid or four-tenths per 

 cent of tri-cresol may be added. 



Antitoxin is fairly stable and if kept in a cool, dark place, may re- 

 main active, with but slight deterioration, for as long as a year. Kept 

 in a dry state, in vacuo, over anhydrous phosphoric acid, by the method 

 of Ehrlich, it retains its strength indefinitely. 



Standardization. — Since antitoxin units are measured in terms of 

 toxin, it is obvious that uniformity of measurement necessitates the 

 possession by the various laboratories of a uniform toxin. Antitoxin 

 being more stable than toxin, uniformity of toxin is obtained by means 

 of a standard antitoxin distributed from a central laboratory. This was 

 first done by Ehrlich in Germany, and is now done for the United States 

 by the Pubhc Health and Marine Hospital Service laboratories. Bottles 

 of the distributed antitoxin are marked with the number of units con- 

 tained in each cubic centimeter. Dilutions of this antitoxin are mixed 

 with varying quantities of the toxin to be tested, the mixtures are al- 

 lowed to stand for fifteen minutes to permit union of the two elements, 

 and injections into guinea-pigs of 250 grams weight are made. In this 

 way, the L+ dose of the toxin is determined. (The L+ dose, as we have 

 seen in a previous section [p. 208], is the quantity of poison not only suf- 

 ficient to neutralize one antitoxin unit,* but to contain an excess beyond 

 this sufficient to kill a guinea-pig of 250 grams in four to five days. 

 L+ is chosen rather than L,,, the simple neutralizing dose, because of 

 the difference between toxins in their contents of toxoid and toxon.^) 



The L+ dose of the toxin having thus been determined, this quantity 

 is mixed with varying dilutions of the unknown antitoxin.^ Thus, 

 given an antitoxin in which 300 to 400 units to the cubic centimeter 

 are suspected, dilutions of 1 :200, 1 :250, 1 :300, etc., are made. One 

 cubic centimeter of each of these is mixed with the L+ dose of the toxin, 

 and the mixtures are injected into guinea-pigs of about 250 grams. If 

 the guinea-pig receiving L+ plus the 1 : 250 dilution lives and the one re- 

 ceiving L^ plus the 1 : 300 dilution dies in the given time, we know that 

 the unit sought must lie between these two values, and further similar 

 experiments will easily limit it more exactly. The possibility of error in 



' A unit of diphtheria antitoxin is a quantity of antitoxin sufficient to protect a 

 guinea-pig of 250 grams against 100 times the fatal dose of diphtheria toxin. 



^Madsen, in Kraus u. Levaditi, " Handbuch," etc., 1907. 



' Donitz, " Die Werthbem. der Heilsera," in Kolle u. Wafisermann, " Hand- 

 buch," etc. 



