CALVIN B. COULTER 519 



t»f the pH value in consequence of hemolysis. All agree, however, 

 in showing an increased dissociation with increased acidity .^^ 



The presence of electrolyte as NaCl greatly increases the proportion 

 of sensitizer combined with cells at all reactions except those near 

 pH 5.3. At this point the combination of sensitizer with cells is 

 independent of the presence of electrolyte. This recalls the obser- 

 vation, to which reference has already been made, that the agglutina- 

 tion of sensitized cells is independent of electrolyte at pH 5.3; it occurs 

 as readily in the presence as in the absence of salt. 



From a comparison of the two curves (Figs. 1 and 2) it is seen that 

 they are almost identical. Since the volumes in the experiments 

 are practically constant it is evident that, under the conditions 

 both of combination de novo and of dissociation from combination, 

 an equilibrium is established in a given volume between the amount 

 of sensitizer free and that combined with cells, for any given hydrogen 

 ion concentration. 



The isoelectric point of serum globulin in which fraction the immune 

 bodies are believed to be carried has been given by Rona and 

 Michaelis^'^ as about pH 5.4, and the isoelectric point of t3rphoid 

 agglutinin has been found by Michaelis and Davidsohn^^ to lie near 

 pH 5.2. It is probable that the hemolytic sensitizer used here has 

 the same value. The point of maximal combination of sensitizer 

 and cells coincides therefore with the isoelectric point of the sensitizer. 



The amphoteric electrolytes, with which the immune bodies must 

 be classed on the basis of their behavior in the electric field (Michaelis 

 and Davidsohn;^' Landsteiner and Pauli^^), owe their electrical charge 



^^ It was found that no destruction or irreversible modification of the sensitizer 

 is brought about by the degrees of acidity or alkalinity reached in the experi- 

 ments. A series of tubes, each containing 0.1 cc. of sensitizer in a volume of 

 10 cc. of saccharose, was brought to various reactions corresponding to those in 

 the experiments with cells, and kept at 38°C. for 35 minutes. After centrifuga- 

 tion the supernatant fluids were adjusted in reaction and titrated for antibody 

 content. No significant differences were found between any of the tubes; the 

 differences were within the experimental error. 



^^ Rona, P., and Michaelis, L., Biochem. Z., 1910, xxviii, 193. 



^^Landsteiner, K., and Pauh, W., 25th Kong. Inn. Med., cited by Landsteiner, 

 K., Kolloide und Lipoide in der Immunitatslehre, in KoUe, W., and Wassermann, 

 A., Handbuch der Pathogenen Mikroorganismen, Jena, 2nd edition, 1913, ii, 

 1241. 



