206 INFECTION AND IMMUNITY 



from differences which may have existed in the depth of injection or 

 rapidity of absorption. 



In order, however, that any such method of standardization of an- 

 titoxin may be practically applicable, it is necessary to produce either 

 a stable toxin or an unchangeable antitoxin. This Ehrlich achieved for 

 antitoxin by drying antitoxic serum in vacuo and preserving it in the 

 dark, at a low temperature and in the presence of anhydrous phosphoric 

 acid. By the use of such a stable antitoxin, various toxins may be 

 measured and other antitoxic sera estimated against these. 



Given thus a constant antitoxin, the standardization of toxins would 

 be a comparatively simple matter were the poison obtainable in a per- 

 fectly pure state. Unfortunately for thp ease of measurement, how- 

 ever, this is not the case. The problem is rendered difficult by a number 

 of complicating factors, many of which have been brought to light by 

 Ehrlich 1 in his laborious researches into the quantitative relationship 

 between the two reacting bodies. 



As previously stated, it had been noted by Ehrlich and others that 

 toxin solutions would deteriorate with time; that is, a toxin-bouillon 



Body Cell 



FIG. 55. TOXIN AND BODY CELL. 



which was found soon after production to contain, say, eighty toxin 

 units in each cubic centimeter, would, after four or five months, be found 

 to contain but forty units in the same gross quantity. It had lost, there- 

 fore, in this case, just one-half of its toxic power. In spite of this loss, 

 however, Ehrlich found that such bouillon had retained its full original 

 power of neutralizing antitoxin. If the reaction was purely one of 

 chemical neutralization, there seemed to,be but one explanation of this. 

 The toxin molecule must contain two separate atom groups. One of 

 these must possess the power of binding antitoxin and be stable; this 



i Ehrlich, Klin. Jahrbuch, vi, 1897; Deut. med. Woch., 1898. 



