Theory of Solution Pressure. 419 



imagine a partition Z, impermeable to silver ions, but allow- 

 ing the free passage of N0 3 ions through it. A partition of 

 this kind, though only imaginary, introduces no ideas different 

 in principle from those connected with the actually realizable 







A 



2 



B 











C , 







C 2 



V 



) 

























X 







X 





Y 











Y' 





partitions X and Y, and so cannot affect the validity of the 

 cycle ; it would, however, have the effect on any passao-e of 

 N0 3 ions from one side to the other, of making the negative 

 charges carried by these the sole charge transferred across it. 

 When the cell is on open circuit the tendency of N0 3 ions 

 to diffuse through Z from A to B, would therefore produce at 

 Z a potential step balancing it, the amount of which, however 

 may be easily calculated on the ordinary osmotic pressure 

 theory. 



The following cycle may now be carried out : — First, a 

 charge Q is allowed to flow round the circuit, in a direction 

 such that silver is precipitated at A, and goes into solution at 

 B. These amounts of silver are both Q/ne gram ions, ne beino- 

 the charge required to deposit 1 gram ion of the silver. At 

 the same time Q/we gram ions of N0 3 pass from A to B, con- 

 stituting the current through Z. Thus the volume of both 

 the positive and negative ions in A is decreased by VjQ/W 

 Vi( = l/c'i) being the volume of a gram ion of either constituent, 

 and work is done on the cell by the motion of X to X' 

 to the amount 2p^ . Q/«e. Similarly work is obtained by 

 the expansion of the solution in B against Y (which moves 

 to Y') to the extent "2p 2 v 2 • Q/ne. 



The second part consists in cutting off from the rest by a 

 semipermeable partition the increase YY' of the silver nitrate 

 in B, and compressing it till its concentration becomes <\ 



2G2 



