JACQUES LOEB AND ROBERT F. LOEB 189 



Of all the characteristics for the two groups mentioned by Noyes, 

 only one is essential, namely, the difference in the ease with which 

 the two groups are precipitated by electrolytes, while the two other 

 differences mentioned by him are only accidental. Thus a solution 

 of crystalline egg albumin requires at ordinary temperature a high 

 concentration of electrolyte for precipitation, although the viscosity 

 of such a solution is low and although the solution has no tendency to 

 gelatinize; while at a sufficiently high temperature, low concentrations 

 of electrolytes will precipitate crystalline egg albumin though the 

 viscosity of the solution is now high and though the solution has 

 now a tendency to gelatinize. Hence only one of the differences 

 mentioned in Noyes' definition remains; namely, the difference in 

 the relative concentration required to precipitate colloids from their 

 solution or their suspension. 



The discrimination between the two types of precipitation accord- 

 ing to whether high or low concentrations of electrolytes are required, 

 is, however, essential and it becomes our first problem to account for 

 this difference. We shall see that it is accompanied by another differ- 

 ence. When high concentrations of electrolytes are required for the 

 precipitation of proteins the sign of charge of the protein particles 

 is of little significance. Sulfates are much more effective for the 

 salting out of gelatin or crystalline egg albumin from their watery 

 solution than chlorides, regardless of whether the solution of the 

 protein is at the isoelectric point or on the acid or alkaline side of it. 



When low concentrations suffice for precipitation, the sign of 

 charge of the protein ions becomes of paramount importance. On 

 the acid side of the isoelectric point the active ion of the precipitating 

 salt is the anion, while on the alkaline side it is the cation. It happens 

 that this is also true for the Donnan effect and that in the Donnan 

 effect also, low concentrations of electrolytes suffice for the depressing 

 effect of a salt. This suggests the possibility that precipitation of 

 the second group of colloids, i.e., where low concentrations are required, 

 is in some way connected with the Donnan equilibrium; and that 

 Hardy's rule is only the consequence of this fact; while where high 

 concentrations of electrolytes are required for precipitation the 

 forces determining the process have no connection with the Donnan 

 equihbrium. 



