4o8 PROTOPLASMIC ACTION AND NERVOUS ACTION 



P.D. across the membrane, no effect is obtained in half 

 an hour or more. The effective gradient of about lo 

 volts, across a partition 50 ju thick, is equivalent to a 

 fall of about 2,000 volts per centimeter. 



From these experiments we may conclude that an 

 arrangement of electrolyte solutions containing oxidizable 

 materials and partitioned by thin films of water-insoluble 

 material having high electrical resistance will be the 

 seat of chemical change (oxidations and reductions) 

 occurring at the surface of the partitions when the system 

 is traversed by an electric current of sufficient intensity. 

 It seems probable that this kind of structural arrange- 

 ment is the one upon which the electrical sensitivity of 

 Hving matter depends. We have already seen that this 

 type of structure is characteristic of living protoplasm. 

 The potentials in protoplasm are much smaller than 

 those used in the experiment above; apparently 50 to 

 100 milUvolts is the usual order of the variations of 

 potential in the bioelectric processes; but the proto- 

 plasmic films are much thinner than those used in our 

 model; and since the steepness of the gradients rather 

 than the absolute values of the potentials is the essential 

 factor to be considered, the conclusion seems justified 

 that conditions similar to the above exist in living 

 protoplasm when the system is traversed by a current. 



The parallel between the passive iron system and an 

 irritable protoplasmic system such as a nerve axone may 

 be described in general terms as follows. In both cases 

 there are two electrically conducting phases separated 

 by a thin impermeable film of chemically alterable 

 material. In protoplasm both of the phases are electro- 

 lytic conductors. In the passive iron system only one 



