PROPERTIES OF ANTITOXINS 227 



killing the animal and triturating the conjunctiva with a fatal dose of 

 abrin, an injection of the emulsion of the right or immunized conjunctiva 

 was without effect, whereas the emulsion from the left proved fatal. - 

 Thus it will clearly be seen that the cells that had absorbed the abrin 

 had developed and contained antiabrin in sufficient amounts to neu- 

 tralize the poison. 



While leukocytes, such as Metchnikoff's macrophages, are likewise 

 active in the formation of antitoxins, it is certain that they are not the 

 only cells involved. Metchnikoff claims that antitoxins are merely 

 toxins altered by leukocytic activity, rather than constituents of tissue- 

 cells; this explanation is, however, inadequate, and it has been shown 

 experimentally that the quantity of antitoxin produced is so far in ex- 

 cess of the amount of toxin injected as to render this view untenable. 



Structure of Antitoxins. According to the side-chain theory, anti- 

 toxins are the simplest of antibodies, being composed of a single arm or 

 haptophore group for union with the toxin, and called receptors of the 

 first order. While illustrations of this theoretic structure will convey 

 .the impression of mere physical contact or union with toxin, it is to be 

 remembered that experimental data indicate that the union and con- 

 sequent neutralization of the toxin are chemical processes. 



Properties of Antitoxins. While chemical analyses to determine 

 the nature of antitoxin serums were made as early as 1897, little is 

 known regarding it because it is impossible to secure the antitoxic ele- 

 ment free from serum and serum constituents. Belfanti and Carbone 

 found that most of the antitoxin in a serum is precipitated with the 

 globulins by saturation with magnesium sulphate. This work, which 

 has been verified by Atkinson and Pick, shows that the antitoxin is 

 carried down with the globulin precipitates, but does not necessarily 

 prove that it is itself a globulin. Later Gibson and Banzhaf showed that 

 the portions of the globulin precipitate soluble in saturated sodium 

 chlorid solution carried most of the antitoxin, and with this discovery 

 a practical method of eliminating much of the non-antitoxic portion of 

 the serum was perfected. 



The relation of antitoxins to proteids has also been studied, digestive 

 ferments being permitted to act on antitoxic serum. It has been shown 

 that antitoxin resists tryptic digestion to a well-marked degree; in this 

 respect it resembles the serum globulin. All the evidence obtained in- 

 dicates that a closer relation of antitoxins to proteids exists than has 

 been shown for the toxins, although all attempts to separate antitoxins 

 from proteids have thus far failed. 



