214 THE ANTIGEN -ANTIBODY REACTIONS 



pH over whicli acid agglutination occurs may vary considerably from one bacterial 

 species to another, and has sometimes been employed as a differential criterion, 

 but its usefulness in this respect would seem to be very limited (see Sgalitzer 1914, 

 Buchanan 1919). 



The relation of salt agglutination to agglutination by specific antisera has been 

 studied in some detail by Northrop and de Kruif (1922a, h). Working with Sahn. 

 typhi, they showed that, with those kations that caused a great reduction in 

 surface potential in low concentrations, flocculation of either sensitized or un- 

 sensitized bacteria tends to occur when the surface potential is reduced below 

 about 15 mvt. With kations that reduce the surface potential only in higher 

 concentrations unsensitized bacteria are not agglutinated, even when the surface 

 potential is reduced to zero, while sensitized bacteria flocculate when the potential 

 falls below 15 mvt. The effects produced on sensitized and unsensitized Salm. 

 typhi by the trivalent kation Al'", the bivalent kation Ba", and the monovalent 

 kation Na", are shown in Fig. 34 (Northrop and de Kruif 1922a). 



In an attempt to determine why unsensitized bacteria fail to agglutinate at relatively 

 high salt concentrations, even when their surface potential is reduced to zero, Northrop 

 and de Kruif (1922a) studied the effect of salt concentration on the " cohesive force " of 

 bacterial cells, by measuring the forces required to separate two cover-slips coated with 

 bacteria. They found that the mutual attraction of unsensitized bacterial ceUs was reduced 

 with increasing salt concentration, whereas this effect was not produced when the bacteria 

 had previously been covered with serum protein. It would seem then that one of the ways 

 in which serum protein, or a specific antibody, exerts its effect is by so altering the bacterial 

 surface that it reacts to a reduction of surface potential in relatively high salt concentrations 

 in the same way that unsensitized bacteria react to an equal reduction of potential in 

 very low salt concentrations. 



Before leaving this question of the influence of electrolytes on agglutination, or on other 

 antigen-antibody reactions, we may note briefly the effect of salt concentrations higher 

 than those that we have considered in the previous paragraphs. High concentrations 

 of sodium chloride (2N or above) inhibit both precipitation and agglutination. These 

 reactions may be delayed in concentrations of 0-2 — 0-5 N sodium chloride. (See Streng 

 1909, Friedberger and Goldschmidt 1910, Landsteiner and Welecki 1911, Eagle 1932.) 

 Schmidt (1930) noted that the reaction was delayed in high concentrations of various salts, 

 and that the order of activity of the anions was approximately that of the Hofmeister series, 



namely (CIO4' > SCN' > CIO3' > NO3' > Br' > IO3' > SO472 > CI' > NOg' > F') 

 and Marrack and Smith (1930) found that toxin-antitoxin floccides were dispersed by 

 strong salt solutions, the order of activity being I' > SCN' > Br' > NO3'. (See also p. 230.) 

 It is probable that the effect is due to a direct action of the salt on the serum proteins. 



The effect of specific antibody itseff on the surface charge of bacteria, at the pH (7-8) 

 and the salt concentration (0-9 per cent. = 0-15 N NaCl) at which agglutination reactions 

 are usually carried out, may be negligible (Shibley 1926, Marrack 1934) ; though some 

 workers (see Brown and Broom 1929) have ascribed great importance to this effect. 



There seems little doubt that the essential effect of sensitization is not its direct 

 action on the surface charge, but the fact that sensitized bacteria react as hydro- 

 phobe colloids, even in moderately high salt concentrations, while unsensitized 

 bacteria do not. These hydrophobe particles remain dispersed only when their 

 surface potential is maintained at a level greater than about 13-15 mvt. When 

 the charge is reduced below this level by the action of electrolytes the bacteria 

 flocculate. It may be noted (Streng 1909, Northrop and de Kruif 19226) that when 

 the amount of antibody combined with the bacteria is very small, agglutination 

 may occur only over a very narrow range of salt concentration. 



