240 THE ANTIOEN-ANTIBODY REACTIONS 



have seen, is equally well exemplified by the results obtained when studying the 

 absorption of any antibody by any antigen. 



We have already discussed the Danysz phenomenon (p. 216). It was first 

 observed in mixtures of diphtheria toxin and antitoxin, but a similar phenomenon 

 has been shown to occur in other antigen-antibody reactions, and analogies have 

 been drawn with chemical reactions of other kinds. Moriyama (1937), for instance, 

 noted that the precipitating effect of tannic acid on proteins varies according to 

 the way in which the reagents are mixed. He was able to eUcit Danysz effects 

 in urease-anti-urease and ricin-antiricin systems when tannin replaced the specific 

 antibody in the systems. He attributed the effects to the capacity of the complex 

 proteins concerned to form several quantitatively different compounds according 

 to the proportions present at first mixing — compounds which, being incompletely 

 reversible, do not contribute to a redistribution of two reagents when more of 

 one reagent is added. Pappenheimer (1942), in a recent review, points out that 

 both the discrepancy between Lq and L^ doses of diphtheria toxin and the 

 Danysz effect can be adequately explained on the assumption that toxin (T) and 

 antitoxin (A) form compounds varying in constitution from TjA to TAg (see Pappen- 

 heimer, Lundgren and WilUams 1940). In the neutral mixture, the average 

 composition of the complexes may be TAj. The addition of an excess of T would 

 result in complexes TA and TjA, and only with toxin in marked excess would 

 enough be left free to kill the guinea-pig. Similarly in the Danysz effect ; the 

 first fraction of toxin would form slowly reversible complexes, Like, for example, 

 TA and its polymers, leaving insufficient antitoxin for union with the remainder 

 of the toxin added later. 



The state of toxin and antitoxin in over-neutralized, neutral and under- 

 neutralized mixtures of the two has been studied by Eagle (1937), who precipitated 

 toxin-horse-antitoxin mixtures with an antiserum to horse globulin. Assuming 

 that free toxin was not carried down in the precipitate. Eagle was able to show 

 that antitoxin unites with excess of toxin to form a toxic complex, and toxin with 

 an excess of antitoxin to forpi a complex with antitoxic properties. Only in one 

 proportion was a precisely neutral complex found. Here again we have recorded 

 analogous findings in the reactions of other antigens with other antibodies. This 

 phenomenon is more readily explicable on the adsorption hypothesis than on the 

 theories advanced by EhrUch, or by Arrhenius and Madsen. 



Other Manifestations of. the Antigen-Antibody Reaction. 



The addition of antigens to antibody even in optimal proportions is not always followed 

 by a visible precipitation. Either of the two reagents may be too weakly reactive or 

 too dilute. The classical method of detecting the union of antigen and antibody in these 

 circumstances is by the complement-fixation reaction (see p. 232) ; but the reaction may 

 be made manifest by other less elaborate means. The high sensitivity of the various 

 flocculation tests for syphilis (see Chapter 81) is due to the addition of agents, such as 

 Upins, which shift the precipitating system more to the hydrophobe state. Again, 

 dyes may act both as sensitizers and as indicators in precipitin reactions. For instance 

 Dean (1937), in a study of non-specific precipitation of proteins by the electropositive 

 dye isamine blue, found that appropriate dilutions of the dye added to constant-antibody 

 titration series markedly increased the sensitivity of the titration ; Berger (1943) made 

 use of Janus green and Victoria blue for the same purpose in the flocculation test for 

 syphihs. Cannon and Marshall (1940), taking advantage of the fact that the amount 

 of antibody required for the surface sensitization of large particles is smaUer than that 

 required for the surface sensitization of the same bulk of material in a more finely divided 



