OF VITAL PHENOMENA 113 



According to Beutner (1913, b) copper ferrocyanide mem- 

 branes are permeable only to the cations of many salts. In order 

 to prevent deterioration of the membrane he set up this series: 

 Calomel electrode | K 4 FeCy e | CuS0 4 +x mol. salt | calomel electrode. 

 With n/40 NaCl in the CuS0 4 solution the emf was 29 milli- 

 volts, and with n/400 NaCl in the CuS0 4 solution the emf was 

 80 mv, or a difference of 51 mv produced by diluting the salt 

 ten times, whereas the difference calculated by Nernst's formula 

 is 58 mv. 



Taking Beutner's results as evidence that these membranes 

 are permeable only to the cations of the salts of alkali metals, we 

 may explain the negative osmose observed by Bernstein on the 

 assumption that the membrane is negative, hence the water posi- 

 tive, the diffusion of the K ions leaving the ferrocyanide solution 

 negative. 



One of the clearest demonstrations of the relation of perme- 

 ability to membrane polarization was shown by Bayliss (1911). 

 He separated two solutions of the sodium salt, congo red, by 

 a membrane of parchment paper. The Na ions can easily pass 

 through the membrane but the congo red anions cannot pass 

 through a dense grade of the paper. Therefore the dilute side 

 is electropositive. Bayliss found the emf, when measured im- 

 mediately after the membrane was wet with the solutions, to 

 approximate that calculated by Nernst's formula, but to fall 

 gradually. He explained the fall by assuming that the osmotic 

 pressure of the concentrated solution draws water through the 

 membrane, thus forming a dilute layer next to the membrane 

 and lowering the difference in concentration between the solu- 

 tions in immediate contact with the two faces of the membrane. 



The author found that even with KC1 solutions, parchment 

 paper gives an emf, but the potential difference falls very rapidly, 

 so that it is impossible to measure the initial potential with the 

 potentiometer method used. The fall in potential is not entirely 

 due to dilution of the more concentrated solution, since after 

 removing and partially drying the membrane and renewing the 

 solutions, the original potential is not obtained. Since the emf 

 becomes zero after a comparatively short time, the fall is prob- 

 ably due, at least in part, to swelling of the membrane and conse- 

 quent enlargement of the pores. In one experiment an emf of 



