JACQUES LOEB 665 



smaller than that of either Ce or La. The drop in the curves when 

 the concentration of the CeCls or AICI3 solution exceeds m/1,024 is 

 due to the concentration of the anion, as is shown by the fact that 

 the drop is greater in the case of Al2(S04)3 than in the case of AICI3. 

 This latter phenomenon has been discussed in previous papers. The 

 pH of the Al2(S04)3 solutions almost coincided with that of the AICI3 

 solutions of the same concentration. 



IV. 



It is therefore certain that the trivalent ions bring about a positive 

 electrification in the collodion membrane in contact with water when 

 the membrane has been treated with a protein, but that they cannot 

 reverse the sign of the charge of collodion membranes not so treated. 

 This indicates that the reversal is at least partly due to an action of 

 the trivalent cations on the protein. A chemical combination can only 

 occur between Ce or Al and gelatin on the alkaline side of the isoelec- 

 tric point where the gelatin is capable of combining with metals. In 

 the experiments described thus far the pH of the CeCl3 solutions was 

 with one exception always > 4.7. The question arises, how will a 

 CeCls solution act on the acid side of the isoelectric point of the 

 protein forming the inner lining of the collodion bag where the mem- 

 brane is already rendered positive by the acid? 



m/256 LaCls solutions were prepared at different pH, from 7.0 

 to 2.6, by adding KOH or HNO3 to the distilled water used for 

 the solutions as required for the pH. These solutions were put into 

 collodion bags and the latter were dipped into H2O of exactly the 

 same pH as that of the solutions inside the bag. The initial rise of 

 water (in the first 20 minutes) inside the bags was then observed. 

 Similar experiments were made with CeCl3. The collodion mem- 

 branes had previously been treated with gelatin. Fig. 5 gives the 

 results. It is plain that the curves consist of two distinct parts and 

 that the dividing line seems to lie near the isoelectric point of gelatin. 

 On the alkaline side of the point, i.e. for pH 4.7 or above, the initial 

 rate of diffusion is near 140 mm.; and it varies very little with a change 

 of pH. As soon, however, as the pH falls to 4.7 or below, the rate 

 of diffusion rises steeply to 300 and 340 mm. for the two salts. ... 



