JOHN H. NORTHROP AND PAUL H. DE KRUIF 



649 



is given, the pH was not regulated and the results are due in part to 

 changes in the hydrogen ion concentration. 



Inspection of the charts shows that in all experiments there is 

 complete agglutination as soon as the potential is reduced below a 

 value of about 15 millivolts (either positive or negative) provided the 

 salt concentration is below 0.001 n. Below this salt concentration, 

 therefore, the agglutination is seen to depend solely on the potential. 

 Any substance which reduces the potential below about 15 milli- 

 volts will cause agglutination. There is another range of salt concen- 

 tration above 0.10 n in which no agglutination occurs, although there 

 is no measurable potential. Between these two ranges of salt con- 



+4 B +18 



o- .g 



S -2 ^ -9 



I 



-4£ -18 



10-4- iO-3 10-2 0.10 1.0 



Salt concentration, ecjuivalents per liter 



Fig. 8. Effect of salts on the potential and agglutination of Type D at pH 2.0. 



centration there is a zone in which agglutination occurs at various 

 potential levels. This is evidently the result that we would expect if 

 the salt acted in low concentration primarily on the potential, and in 

 high concentration on the cohesive force. There would be an inter- 

 mediate zone in which the agglutination could not be predicted from 

 either measurement alone. This explanation is borne out by the 

 measurements of the cohesive force shown in Fig. 9. These show that 

 the cohesive force is markedly decreased in concentrations of more 

 than 0.01 n; i.e., the range in which the critical potential begins to 

 decrease. The figure shows that the efifect on the cohesion is not 

 connected with the valency nor with the electrical effects of the ions. 

 LaCla is far more effective than NaCl in reducing the potential, but 



