14 THEORY OF COLLOIDAL BEHAVIOR 



bromides, and nitrates than in water, while in acetates, tartrates, 

 citrates, or sugar it swells less than in water. R. S. Lillie 1 arranges 

 ions according to their depressing effect on the osmotic pressure 

 of gelatin solution in the following way: 



Cl>S0 4 >N0 3 >Br>I>CNS 



These series 2 again betray no relation to the stoichiometrical 

 properties of the ions. As long as these Hofmeister ion series 

 were believed to have a real existence it seemed futile to decide 

 for or against a purely chemical theory of the behavior of colloids 

 since even with a bias in favor of a chemical theory the Hof- 

 meister series remained a riddle. 



The writer believes that he has removed these difficulties by 

 using protein solutions of equal hydrogen ion concentration as 

 the standard of comparison. 



In this way it was found that a number of authors had errone- 

 ously attributed the effects of an alteration of the hydrogen ion 

 concentration upon the physical properties of a protein to a 

 difference in the specific action of the anion or cation added. 

 Thus it was always believed that acetates have almost as great a 

 "dehydrating" action as sulphates, but it was overlooked that 

 acetic acid is a weak acid, and that in the experiments referred to 

 the authors failed to compare the effects of SO 4 and CH 3 COO at 

 the same hydrogen ion concentration. When this error is 

 avoided it can be shown that acetates influence the swelling, 

 osmotic pressure, and viscosity of protein solutions in the same 

 way as chlorides or nitrates, but not in the same way as sulphates; 

 in other words, anions of the same valency act alike. 3 



By taking into consideration the hydrogen ion concentration 

 it was possible to show that the assumption of specific differences 

 in the action of different ions of the same valency and sign of 

 charge was due to a methodical error; and that the Hofmeister 

 rule must be replaced by a simple valency rule, according to 

 which only the valency and sign of charge of an ion influence the 

 colloidal behavior of a protein but that the other properties of 



1 LILLIE, R. S., Am. J. PhysioL, vol. 20, p. 127, 1907-08. 



2 A fuller discussion of these series is found in HOBER, R., " Physikalische 

 Chemie der Zelle und der Gewebe," Leipsic and Berlin, 1914. 



3 LoEB, J., J. Gen. Physiol, vol. 3, p. 391, 1920-21. 



