LIVING TORPEDOS IN BERLIN. 513 



The two polarisations are here both taken to be proportional to 

 the current density as in the first communication (pp. 455). It is 

 self-evident that they are also proportional to the length of the 

 preparation. 



The question whether we can decide between the two possibilities 

 in this way, is answered in the negative by the very statement of 

 the above formulas. More than fifty years ago, when treating 

 of the similar question as to the existence of a transition resistance 

 besides the polarisation of the electrodes, G. S. Ohm directed atten- 

 tion to the fact that in such expressions as (d] and (e) the way in 

 which polarisation and resistances severally enter is only apparent, 

 that polarisation therefore stands in the same relation to an ad- 

 ditional resistance as the constant resistances do, and that it cannot 

 be distinguished from these by the introduction of an additional 

 resistance. 



We find in fact 



nG T nG 



1-- j- - , 1, - - -7- - , 



and if for sake of brevity we put 



l_ 

 <1 



7= * G 



_ nGlP 

 - 2 >- 



This expression is finite and positive, so long as qA n >IP-, qA n =lP 

 would make / infinite, and %A n < IP would give the upper hand to 

 the heterodromous current. By introducing the additional resist- 

 ance and exchanging n for N, (f) becomes 



I-I, = NGIP 



When W is great, this expression approaches to the value 

 J I- NGIP > 



'~ qW- ' 

 Ll 



