-14 REFUTATION OF THE SALT RADICAL THEORY, 
without uniting with a cathion. In this explanation I am 
supported by the author's account of a well known experi- 
ment by Faraday, in which a solution of magnesia and 
water was made to act as electrodes at their surfaces respec- 
tively. 
82. There can, I think, be no better proof that no reli- 
ance should be placed on the experiments with membranes, 
in this and other cases where the existence of compound 
radicals in acids is to be tested, than the error into which 
an investigator, so sagacious as my friend Professor Daniell, 
has been led, in explaining the complicated results. 
83. The association of two electrolytes, and the chemical 
reaction between the potassa and acid, which is admitted to 
have evolved the hydrated oxide, seem rather to have 
created difficulties than to have removed them. 
84. In this view of the subject, I am supported by the 
opinion of Faraday, as expressed in the following lan- 
guage:— 
u When other metallic solutions are used, containing, for instance, 
peroxides, as that of copper combined with this or any decomposable 
acid, still more complicated results will be obtained, which, viewed as 
the direct results of electrochemical action, will, in their proportions, 
present nothing" but confusion; but will appear perfectly harmonious 
and simple, if they be considered as secondary results, and will accord 
in their proportions with the oxygen and hydrogen evolved from water 
by the action of a definite quantity of electricity." 
85. I cannot conceive, that in any point of view the com- 
plicated and "confused" results of the experiment of 
Daniell with electrolytes separated by membranes, are ren- 
dered more intelligible by supposing the existence of salt 
radicals. I cannot perceive that the idea that the anion in 
the sulphate is oxysulphion, makes the explanation more 
satisfactory than if we suppose it to be oxygen. Were a 
solution of copper subjected to electrolysis alone, if the 
oxide of copper were the primary object of the current, the 
result would be analogous to the case of sodium, excepting 
