510 LIVING TOKPEDOS IN BERLIN. 



This indeed is still impossible supposing- R = o, for then it re- 

 duces itself to I\ > I k . But if R has a great value, like that with 

 which we are dealing, it approximates to the actually existing 

 inequality / > k ; and accordingly it may be considered proved that, 

 other things being equal, the irreciprocity increases with the 

 length of the tract of organ traversed. 



9. The question, whether the appearance of irreciprocal 

 conduction in the organ depends upon polarisation or con- 

 duction, cannot be answered by the introduction of additional 

 resistance. 



In the previous paragraph we set out with the hypothesis, that 

 irreciprocal conduction may be expressed by an additive term in the 

 denominator of Ohm's formula for the heterodromous current 

 strength. This procedure, which has already proved available for 

 our purpose, will be justified later ; for the present, we must not 

 forget that the question whether irreciprocal conduction depends 

 upon unequal resistance in both directions, or upon polarisation, is 

 still open. At first sight, the surest way to answer this question 

 seems to be the mode of experiment mentioned already in the 

 previous communication (p. 453), viz. to introduce into the primary 

 circuit, a resistance in addition to that of the organ preparation so 

 great, that the latter shall vanish. In this case, if the difference of 

 the two currents should disappear, that of the polarisations per- 

 sisting, the origin of the former difference would be shown to be 

 the difference of resistance. In an experiment of this nature, in 

 which I used a long narrow tube filled with physiological solution 

 of rock-salt for the additional resistance, the difference of the 

 current strengths in both directions did in fact disappear ; but 

 since, on account of lowered functional activity, the difference of the 

 polarisation currents was also absent, no certain conclusion could be 

 drawn from it. In the experiments of the previous paragraph, the 

 apparent irreciprocity of the conduction was persistent, although 

 the resistance of the preparation was but small in comparison with 

 that of the rest of the circuit. It did not however disappear, and 

 the polarisation was not observed, so that we cannot conclude from 

 these observations that irreciprocity depends upon polarisation. 



I have now carried out several experiments according to the 

 method under consideration, with apparently good result. As liquid 

 resistances are difficult to graduate and to estimate, I replaced the 

 tube containing rock-salt solution by two plug rheostats of Siemens 



