ANALYSIS OF COMPLEX MATTER 123 



observed ; one of the copper plates will become heavier by 

 deposition upon it of copper. The other plate, however, will 

 be found to be correspondingly lighter. We are thus led to 

 investigate the other plate used in the case of silver nitrate. 

 This will be found lighter, but not correspondingly lighter. 



In order to obtain accurate results the quantities measured 

 should be as large as possible. This may be effected by in- 

 creasing the number of cells, or allowing the change to proceed 

 for a longer time. 



It will be noticed that the change in question appears to 

 take place at the places of contact between the solution and 

 the rest of the circuit. It is, however, difficult to understand 

 how these regions can be alone affected, for we have no reason 

 to suppose any one portion of the liquid between the plates to 

 differ from another with regard to its electric condition. There 

 is apparently, therefore, an internal change set up through- 

 out the liquid which is in the circuit, but this change is only 

 manifested at the places where it starts and ends. The result 

 is as if the particles of the solid in solution had been decom- 

 posed, and the products directed or arranged in a special 

 manner. That the copper added to the heavier plate is derived 

 evidently from the solution, and not, as might appear, from the 

 plate which has lost matter, may be proved by using platinum 

 or silver plates instead of copper. The deposition of copper 

 proceeds in much the same manner as before. 



81. The Quantity of a given kind of Matter, liberated under 

 the same Electric Conditions, is independent of the kind of 

 Matter associated with it ; while the Quantities of Different 

 kinds of Matter, liberated under the same Electric Condi- 

 tions, have a fixed Invariable Ratio to one another. Solu- 

 tions of copper sulphate, copper nitrate, and silver nitrate are 

 placed in an electric circuit. Clean copper plates are used as 

 before for making connections, except with the silver nitrate 

 solution, into which silver plates dip. Each of the plates is 

 carefully weighed and marked. It is convenient to use in 

 these experiments a key, by which the circuit can be closed 

 or opened at will. While the necessary connections are 

 being made it remains open. We may then submit the three 



