THE ACTION OF SALTS 213 



nature of the modification is made clear by the work of Bawden and Kleczkowski (19426) 

 who showed that, under the influence of heat, a combination between the albumins and 

 the antibody globuUn of the antiserum occurred. The resulting complex combined with 

 antigen without precipitation and strongly inhibited the reaction between unmodified 

 antibody and the homologous plant-virus antigen. The pro-zones of unmodified antisera 

 are not necessarily due to preferentially absorbed non-agglutinating antibodies, for Miles 

 (1939) sensitized Brucella suspensions in the absence of salt, with both optimal and 

 16 X optimal amounts of a strongly pro-zoning antiserum, and, after removal of unabsorbed 

 antibody, found that on the addition of saline the rate of flocculation of both suspensions 

 was identical. 



As in precipitin titration, multiple zones of maximal flocculation occur in certain 

 systems. Miles (1939) demonstrated in a BniceUa system two such zones, which appeared 

 to be due to the reactions of separate smooth and rough antigens on the bacterial surface 

 with their respective antibodies. 



The optimal proportions method is not, however, commonly employed in 

 determining the antibody content of an agglutinating serum. In this case, the 

 optimal ratios of any two sera bear the same relation to one another as do the 

 end-points observed when increasing dilutions of the sera are titrated against a 

 constant amount of bacterial suspension ; and the end-point is more easily 

 determined. 



In performing such titrations the experimental conditions, time and temperature 

 of incubation and so on, must be so arranged that the true end-point is attained. 

 If great accuracy is required, a definite degree of agglutination must be selected 

 as marking this end-point, and this is usually taken to be the least degree of floccula- 

 tion that can be easily observed by transmitted light against a dark background, 

 when viewed by the naked eye, or with the aid of a hand lens. When the amount 

 of antiserum in successive tubes is halved, as is commonly the case, it will often 

 happen that no tube shows the standard degree of agglutination. When this is so, 

 a good approximation can sometimes be obtained by using an appropriate inter- 

 polation table based on a sufficient series of preliminary experiments (Dreyer and 

 Inman 1917). 



The Effect of Salts on the Combination of Antigen and Antibody. 



Turning now to the mechanism concerned in the actual flocculation of bacteria, 

 we may consider first the role of electrolytes in this phase of the reaction. The 

 isoelectric point of many bacteria is in the neighbourhood of 3-0 ; at this pH the 

 bacterium is not an uncharged particle, but rather a particle with its positive 

 and negative charges balanced. When suspended in neutral solutions of low 

 salt content, the ionization of those groups responsible for the positive charge 

 is reduced, leaving an excess negative charge. Under these conditions, therefore, 

 the bacteria move towards the anode in an electrophoretic cell (see Putter 1921, 

 Northrop 1922, 1928, Winslow et at. 1923 and many subsequent observers). If 

 the pH is lowered below the isoelectric point, the ionization of the groups responsible 

 for the negative charge is reduced and the bacteria drift towards the kathode. 



Agglutination depends in part upon the electrostatic attraction between opposite 

 charges on different bacteria, but at pH values where all the bacteria carry an 

 excess of one charge, this is impeded by the electrostatic repulsion that like charges 

 have for one another. Many observations have been made on the flocculating 

 effect of salts and acids on bacteria, altogether apart from the action of agglutin- 

 ating antibodies (Bechhold 1904, Michaelis 1911, Beniasch 1911). The range of 



