[B] Elements of Chemical Philosophy, pp. 144. 161.

[C] It is remarkable that up to 1804 it was the received opinion that the metals were reduced by the nascent hydrogen.  At that date the general opinion was reversed by Hisinger and Berzelius (Annales de Chimie, 1804, tom. li. p. 174,), who stated that the metals were evolved directly by the electricity:  in which opinion it appears, from that time, Davy coincided (Philosophical Transactions, 1826, p. 388).

747.  But when we take to our assistance the law of constant electro-chemical action already proved with regard to water (732.), and which I hope to extend satisfactorily to all bodies (821.), and consider the quantities as well as the nature of the substances set free, a generally accurate judgement of the primary or secondary character of the results may be formed:  and this important point, so essential to the theory of electrolyzation, since it decides what are the particles directly under the influence of the current, (distinguishing them from such as are not affected,) and what are the results to be expected, may be established with such degree of certainty as to remove innumerable ambiguities and doubtful considerations from this branch of the science.

748.  Let us apply these principles to the case of ammonia, and the supposed determination of nitrogen to one or the other electrode (554. 555,).  A pure strong solution of ammonia is as bad a conductor, and therefore as little liable to electrolyzation, as pure water; but when sulphate of ammonia is dissolved in it, the whole becomes a conductor; nitrogen almost and occasionally quite pure is evolved at the anode, and hydrogen at the cathode; the ratio of the volume of the former to that of the latter varying, but being as 1 to about 3 or 4.  This result would seem at first to imply that the electric current had decomposed ammonia, and that the nitrogen had been determined towards the positive electrode.  But when the electricity used was measured out by the volta-electrometer (707. 736.), it was found that the hydrogen obtained was exactly in the proportion which would have been supplied by decomposed water, whilst the nitrogen had no certain or constant relation whatever.  When, upon multiplying experiments, it was found that, by using a stronger or weaker solution, or a more or less powerful battery, the gas evolved at the anode was a mixture of oxygen and nitrogen, varying both in proportion and absolute quantity, whilst the hydrogen at the cathode remained constant, no doubt could be entertained that the nitrogen at the anode was a secondary result, depending upon the chemical action of the nascent oxygen, determined to that surface by the electric current, upon the ammonia in solution.  It was the water, therefore, which was electrolyzed, not the ammonia.  Further, the experiment gives no real indication of the tendency of the element nitrogen to either one electrode or the other; nor do I know of any experiment with nitric acid, or other compounds of nitrogen, which shows the tendency of this element, under the influence of the electric current, to pass in either direction along its course.