[B] Ibid. pp. 144, 145.

473.  This opinion has, I think, been shown by other philosophers not to be accurate, though I do not know where to refer for a contradiction of it.  Sir Humphry Davy himself said in 1801[A], that dry nitre, caustic potash and soda are conductors of galvanism when rendered fluid by a high degree of heat, but he must have considered them, or the nitre at least, as not suffering decomposition, for the statements above were made by him eleven years subsequently.  In 1826 he also pointed out, that bodies not containing water, as fused litharge and chlorate of potassa, were sufficient to form, with platina and zinc, powerful electromotive circles[B]; but he is here speaking of the production of electricity in the pile, and not of its effects when evolved; nor do his words at all imply that any correction of his former distinct statements relative to decomposition was required.

  [A] Journal of the Royal Institution, 1802, p. 53.

  [B] Philosophical Transactions, 1826, p. 406.

474.  I may refer to the last series of these Experimental Researches (380. 402.) as setting the matter at rest, by proving that there are hundreds of bodies equally influential with water in this respect; that amongst binary compounds, oxides, chlorides, iodides, and even sulphurets (402.) were effective; and that amongst more complicated compounds, cyanides and salts, of equal efficacy, occurred in great numbers (402.).

475.  Water, therefore, is in this respect merely one of a very numerous class of substances, instead of being the only one and essential; and it is of that class one of the worst as to its capability of facilitating conduction and suffering decomposition.  The reasons why it obtained for a time an exclusive character which it so little deserved are evident, and consist, in the general necessity of a fluid condition (394.); in its being the only one of this class of bodies existing in the fluid state at common temperatures; its abundant supply as the great natural solvent; and its constant use in that character in philosophical investigations, because of its having a smaller interfering, injurious, or complicating action upon the bodies, either dissolved or evolved, than any other substance.

476.  The analogy of the decomposing or experimental cell to the other cells of the voltaic battery renders it nearly certain that any of those substances which are decomposable when fluid, as described in my last paper (402.), would, if they could be introduced between the metallic plates of the pile, be equally effectual with water, if not more so.  Sir Humphry Davy found that litharge and chlorate of potassa were thus effectual[A].  I have constructed various voltaic arrangements, and found the above conclusion to hold good.  When any of the following substances in a fused state were interposed between copper and platina,