454 LIVING TOKPEDOS IN BERLIN. 



determination of this question, and I have not yet "been able to 

 repeat the experiment. 



I took an opportunity of approaching- the desired end in another 

 way. Instead of observing- the polarisation after opening the 

 primary current, I tried to form some notion of it during its 

 persistence, by short circuiting the preparation through the leading- 

 off clay points and the galvanometer circuit. The battery con- 

 sisted of 20 Groves ; the galvanometer coil (S) of 5000 turns had to 

 be placed at a great distance from the mirror in order to keep the 

 deflections within due limits. The closing time amounted at first 

 to o"o764, afterwards to i // -O24. The result was astonishing ; for 

 while the homodromous current through the galvanometer (P) 

 appeared almost twice as strong as the hefcerodromous, the branch 

 of the homodromous current led off through the galvanometer (S) 

 appeared very much weaker than that from the heterodromous 

 (Series 25). When I replaced the preparation by a model of 

 physiological clay (a much worse conductor, by the way) the 

 difference had disappeared a proof that irreciprocal resistance did 

 not exist anywhere in the circuit. 



It thus seems settled that the difference depends upon irreci- 

 procal resistance. For as it is not conceivable that with an equal 

 resistance in the preparation the led-off branch of the stronger current 

 should turn out to be weaker, one is led to suppose that this depends 

 exclusively on the better conductivity of the preparation for the 

 homodromous current. More precise considerations however, aided 

 by a schematic calculation, teach that this view is not sufficient, 

 and that at all events under the assumption taken as the basis of 

 the calculation, the result must be different. 



For the sake of simplicity we assume that both currents produce 

 relatively negative polarisation of equal strength, but that rela- 

 tively positive polarisation is produced only by the homodromous. 

 We assume the electromotive force of both polarisations to be pro- 

 portional to the density of the battery current in the preparation, 

 taking the transverse section and the length as each severally =i. 

 We will first enquire into what happens when the specific resistance 

 of the preparation is unequal, as we thus avoid the complication 

 which would arise from the assumption that the difference of 

 resistance in the two directions may be a function of the current 

 density. 'This assumption, though well worthy of consideration, is 

 very difficult to express mathematically. Also we look upon the 

 preparation as a linear conductor, or as if both the dorsal and 



