Electromotive Force of a Constant Voltaic Cell. 411 



think that we are justified in refusing to accept such a result 

 until not only the E.M.F. of the given voltaic combination 

 has been accurately measured, but it has also been clearly 

 shown to what chemical changes this electromotive force is 

 due. 



To show this it is necessary to draw a considerable current 

 for some time from the cell in order to find by analysis what 

 chemical changes have taken place ; and, further, it is neces- 

 sary to measure the E.M.F. during the passage of this current. 

 Unfortunately, as soon as we attempt to draw a considerable 

 current per square centimetre of plate surface from a so-called 

 constant cell the E.M.F. immediately falls off considerably. 

 To give an example : Dr. Alder Wright finds that a current of 

 30 or 40 microamperes per square centimetre is quite sufficient 

 to cause an appreciable fall in the E.M.F. of the cell experi- 

 mented on. Now a current of 30 microamperes running for 

 an hour from a cadmium cell would dissolve about *00006 

 grm. of metallic cadmium, an amount hardly sufficient for 

 analytical purposes. 



Now this fall in E.M.F., on taking an appreciable current 

 per square centimetre from a constant voltaic cell, is usually 

 ascribed to " polarization," "transition resistance," &c. On 

 considering it more carefully, it seems to be largely due, pos- 

 sibly entirely due, to alterations in the layers of liquid next 

 the plate. 



To keep to the cadmium cell, an increase of cadmium iodide 

 in the layer of liquid next the cadmium plate, or a diminution 

 of the free iodine in the layer of liquid against the platinum 

 plate, would cause a considerable diminution of E.M.F. 



If, then, we suppose the plates immersed in a sufficient 

 amount of solution to remain of practically constant strength 

 though a considerable current be running from the cell for 

 some time, the amount of current per square centimetre which 

 we can draw from the cell without a fall in E.M.F., will be 

 largely, perhaps entirely, a question of the rate at which 

 diffusion of the cadmium iodide and free iodine takes place. 



It follows, then, obviously that if we wish to draw a con- 

 siderable current from a voltaic cell under conditions similar 

 to those during the measurement of E.M.F. in the usual 

 way, the liquid contents of the cell must be kept well stirred. 



I determined therefore to make some experiments on a 

 cadmium-iodine cell with moving plates, drawing from the 

 cell a considerable current per square centimetre ; a sufficient 

 current, in fact, to enable me to estimate quantitatively at the 

 end of about two hours the nature and amount of the salts 

 formed in the cell, 



