328.  This experiment is neither more nor less than a repetition, in a refined manner, of that made by Dr. Pearson in 1797[A], and previously by MM.  Paets Van Troostwyk and Deiman in 1789 or earlier.  That the experiment should never be quoted as proving true electro-chemical decomposition, is sufficiently evident from the circumstance, that the law which regulates the transference and final place of the evolved bodies (278. 309.) has no influence here.  The water is decomposed at both poles independently of each other, and the oxygen and hydrogen evolved at the wires are the elements of the water existing the instant before in those places.  That the poles, or rather points, have no mutual decomposing dependence, may be shown by substituting a wire, or the finger, for one of them, a change which does not at all interfere with the other, though it stops all action at the changed pole.  This fact may be observed by turning the machine for some time; for though bubbles will rise from the point left unaltered, in quantity sufficient to cover entirely the wire used for the other communication, if they could be applied to it, yet not a single bubble will appear on that wire.

  [A] Nicholson’s Journal, 4to. vol.  I. pp. 311, 299. 349.

329.  When electro-chemical decomposition takes place, there is great reason to believe that the quantity of matter decomposed is not proportionate to the intensity, but to the quantity of electricity passed (320.).  Of this I shall be able to offer some proofs in a future part of this paper (375. 377.).  But in the experiment under consideration, this is not the case.  If, with a constant pair of points, the electricity be passed from the machine in sparks, a certain proportion of gas is evolved; but if the sparks be rendered shorter, less gas is evolved; and if no sparks be passed, there is scarcely a sensible portion of gases set free.  On substituting solution of sulphate of soda for water, scarcely a sensible quantity of gas could be procured even with powerful sparks, and nearly none with the mere current; yet the quantity of electricity in a given time was the same in all these cases.

330.  I do not intend to deny that with such an apparatus common electricity can decompose water in a manner analogous to that of the voltaic pile; I believe at present that it can.  But when what I consider the true effect only was obtained, the quantity of gas given off was so small that I could not ascertain whether it was, as it ought to be, oxygen at one wire and hydrogen at the other.  Of the two streams one seemed more copious than the other, and on turning the apparatus round, still the same side in relation to the machine; gave the largest stream.  On substituting solution of sulphate of soda for pure water (329.), these minute streams were still observed.  But the quantities were so small, that on working the machine for half an hour I could not obtain at either pole a bubble of gas larger than a small grain of sand.  If the conclusion which I have drawn (377.) relating to the amount of chemical action be correct, this ought to be the case.