upon the metals.  They must, however, be chosen from the metals themselves, for there are no bodies of this kind except those substances and charcoal.  To decide the matter by experiment, I made the following arrangement.  Melted tin was put into a glass tube bent into the form of the letter V, fig. 78, so as to fill the half of each limb, and two pieces of thick platina wire, p, w, inserted, so as to have their ends immersed some depth in the tin:  the whole was then allowed to cool, and the ends p and w connected with a delicate galvanometer.  The part of the tube at x was now reheated, whilst the portion y was retained cool.  The galvanometer was immediately influenced by the thermo-electric current produced.  The heat was steadily increased at x, until at last the tin and platina combined there; an effect which is known to take place with strong chemical action and high ignition; but not the slightest additional effect occurred at the galvanometer.  No other deflection than that due to the thermo-electric current was observable the whole time.  Hence, though a conductor, and one capable of exerting chemical action on the tin, was used, yet, not being an electrolyte, not the slightest effect of an electrical current could be observed (947.).

923.  From this it seems apparent that the peculiar character and condition of an electrolyte is essential in one part of the voltaic circuit; and its nature being considered, good reasons appear why it and it alone should be effectual.  An electrolyte is always a compound body:  it can conduct, but only whilst decomposing.  Its conduction depends upon its decomposition and the transmission of its particles in directions parallel to the current; and so intimate is this connexion, that if their transition be stopped, the current is stopped also; if their course be changed, its course and direction change with them; if they proceed in one direction, it has no power to proceed in any other than a direction invariably dependent on them.  The particles of an electrolytic body are all so mutually connected, are in such relation with each other through their whole extent in the direction of the current, that if the last is not disposed of, the first is not at liberty to take up its place in the new combination which the powerful affinity of the most active metal tends to produce; and then the current itself is stopped; for the dependencies of the current and the decomposition are so mutual, that whichsoever be originally determined, i.e. the motion of the particles or the motion of the current, the other is invariable in its concomitant production and its relation to it.