132 Messrs. Gee and Rolden on 



which gives equal kicks in the two directions, and its ordinate 

 equal to these kicks. The conditions for the occurrence of 

 reciprocity are then investigated. He finds (1) that lessen- 

 ing the size of the point electrode causes reciprocity to occur 

 with lower E.M.F.'s, but that fairly large changes in the size 

 of the plate electrode or in the distance of the electrodes have 

 not much influence; (2) that decreasing the temperature 

 causes reciprocity to occur with lower E.M.F.'s; and (3) that 

 the degree of concentration of the solution is of influence. 

 A discussion of the possible causes of irreciprocal conduction 

 follows. Polarization, he says, cannot be the cause of irre- 

 ciprocity, because he has obtained the latter with non-polari- 

 zable electrodes. Further, some measurements of the polari- 

 zation by a wippe method show, as far as such a method is 

 satisfactory *, that polarization is not the main cause of irre- 

 ciprocity. His arguments against polarization, based on the 

 measurements of e x (see p. 130) are, as before explained, of 

 no value. His main argument against a chemical transition 

 resistance is the following experiment : — A solution of potas- 

 sium iodide is taken, and a certain number of currents sent 

 through until the E.M.F. is found for which reciprocity occurs; 

 the electrodes are then taken out and cleaned, and fresh solu- 

 tion used. The same E.M.F. is now again applied, and it is 

 found that reciprocity still occurs; or, in other words, changes 

 in the electrolyte caused by the passage of previous currents 

 do not affect the value of the E.M.F., for which the momentary 

 currents are the same in the two directions. The direction 

 in which the E.M.F. is first applied is also quite indifferent. 

 He remarks, however, that the reciprocal point in other 

 electrolytes is, to some extent, affected by the passage of 

 previous currents. 



Christiani to a small extent makes use of permanent cur- 

 rents, but only in the case of the so-called non-polarizable 

 electrodes. The following specimen of his results, with 

 amalgamated zinc electrodes in concentrated zinc sulphate, 

 will indicate their general features : — 



* The values of the polarization obtained by this method are of course 

 open to the objection that they depend on the rate of fall of polarization 

 after the removal of the charging battery. 



