500 AMPHOTERIC COLLOIDS. IV 



basis of these properties. The viscosity rises for some time after the 

 melting of gelatin and only approximately constant values can be 

 obtained by measuring the time of outflow immediately; i.e., within 5 

 minutes after the melting. The swelling is influenced by the resist- 

 ance of the gelatin to the expanding forces leaving aside other sources 

 of error. The osmotic pressure in connection with measurements on 

 conductivity seems by far the most valuable property for the analy- 

 sis of the molecular character of the influence of electrolytes upon the 

 physical behavior of colloids. 



The approximate parallelism between the curves for viscosity, swell- 

 ing, and osmotic pressure excludes the idea that sweUing and viscosity 

 are determined by a hydratation due to the degree of ionization of 

 proteins as Pauli and many other colloid chemists assume. These 

 hypotheses were developed without measurements of conductivity. 



V. Action of Trivalent Cations. 



The facts of the preceding chapters furnish a suggestion for the 

 understanding of the action of trivalent metals. If the same tenta- 

 tive assumption may be made, namely that the gelatin ions held by 

 the polyvalent metal ion dissociate in one aggregate, thus causing a 

 diminution in the number of particles present in the solution without 

 a diminution in the number of charges, it follows that this anionic 

 aggregate must be the greater the greater the valency of the metal 

 ion — leaving aside the atomic volume. Hence trivalent metals like 

 Ce or Al should, if in combination with gelatin, lead to anionic aggre- 

 gates consisting of a simple multiple of three. We cannot test this 

 inference but we can show that gelatin compounds with a trivalent 

 metal like cerium and aluminium are practically insoluble, and this 

 insolubility might be the consequence of the increase in the number 

 of gelatin anions forming an aggregate with the valency of cations. 



1 gm. of finely powdered Cooper's gelatin (which is essentially cal- 

 cium gelatinate of pH = 7.0) was put for 1 hour into each of a series 

 of beakers containing 100 cc. of a solution of Ce2Cl6 varying in concen- 

 tration from m/4 to m/ 16,384, then drained on a filter and washed 

 four times with distilled water to remove the excess of salt solution. 

 The conductivity osmotic pressure, alcohol number, and total swell- 



