764. From all the experiments, combined with these considerations, I conclude that muriatic acid is decomposed by the direct influence of the electric current, and that the quant.i.ties evolved are, and therefore the chemical action is, _definite for a definite quant.i.ty of electricity_. For though I have not collected and measured the chlorine, in its separate state, at the _anode_, there can exist no doubt as to its being proportional to the hydrogen at the _cathode_; and the results are therefore sufficient to establish the general law of _constant electro-chemical action_ in the case of muriatic acid.

765. In the dilute acid (761.), I conclude that a part of the water is electro-chemically decomposed, giving origin to the oxygen, which appears mingled with the chlorine at the _anode_. The oxygen _may_ be viewed as a secondary result; but I incline to believe that it is not so; for, if it were, it might be expected in largest proportion from the stronger acid, whereas the reverse is the fact. This consideration, with others, also leads me to conclude that muriatic acid is more easily decomposed by the electric current than water; since, even when diluted with eight or nine times its quant.i.ty of the latter fluid, it alone gives way, the water remaining unaffected.

766. _Chlorides._--On using solutions of chlorides in water,--for instance, the chlorides of sodium or calcium,--there was evolution of chlorine only at the positive electrode, and of hydrogen, with the oxide of the base, as soda or lime, at the negative electrode. The process of decomposition may be viewed as proceeding in two or three ways, all terminating in the same results. Perhaps the simplest is to consider the chloride as the substance electrolyzed, its chlorine being determined to and evolved at the _anode_, and its metal pa.s.sing to the _cathode_, where, finding no more chlorine, it acts upon the water, producing hydrogen and an oxide as secondary results.

As the discussion would detain me from more important matter, and is not of immediate consequence, I shall defer it for the present. It is, however, of _great consequence_ to state, that, on using the volta-electrometer, the hydrogen in both cases was definite; and if the results do not prove the definite decomposition of chlorides, (which shall be proved elsewhere,--789. 794. 814.,) they are not in the slightest degree opposed to such a conclusion, and do support the _general law_.

767. _Hydriodic acid._--A solution of hydriodic acid was affected exactly in the same manner as muriatic acid. When strong, hydrogen was evolved at the negative electrode, in definite proportion to the quant.i.ty of electricity which had pa.s.sed, i.e. in the same proportion as was evolved by the same current from water; and iodine without any oxygen was evolved at the positive electrode. But when diluted, small quant.i.ties of oxygen appeared with the iodine at the _anode_, the proportion of hydrogen at the _cathode_ remaining undisturbed.

768. I believe the decomposition of the hydriodic acid in this case to be direct, for the reasons already given respecting muriatic acid (763. 764.).

769. _Iodides._--A solution of iodide of pota.s.sium being subjected to the voltaic current, iodine appeared at the positive electrode (without any oxygen), and hydrogen with free alkali at the negative electrode. The same observations as to the mode of decomposition are applicable here as were made in relation to the chlorides when in solution (766.).

770. _Hydro-fluoric acid and fluorides._--Solution of hydrofluoric acid did not appear to be decomposed under the influence of the electric current: it was the water which gave way apparently. The fused fluorides were electrolysed (417.); but having during these actions obtained _fluorine_ in the separate state, I think it better to refer to a future series of these Researches, in which I purpose giving a fuller account of the results than would be consistent with propriety here[A].

[A] I have not obtained fluorine: my expectations, amounting to conviction, pa.s.sed away one by one when subjected to rigorous examination; some very singular results were obtained; and to one of these I refer at 1340.--_Dec. 1838._

771. _Hydro-cyanic acid_ in solution conducts very badly. The definite proportion of hydrogen (equal to that from water) was set free at the _cathode_, whilst at the _anode_ a small quant.i.ty of oxygen was evolved and apparently a solution of cyanogen formed. The action altogether corresponded with that on a dilute muriatic or hydriodic acid. When the hydrocyanic acid was made a better conductor by sulphuric acid, the same results occurred.

_Cyanides._--With a solution of the cyanide of pota.s.sium, the result was precisely the same as with a chloride or iodide. No oxygen was evolved at the positive electrode, but a brown solution formed there. For the reasons given when speaking of the chlorides (766.), and because a fused cyanide of pota.s.sium evolves cyanogen at the positive electrode[A], I incline to believe that the cyanide in solution is _directly_ decomposed.

[A] It is a very remarkable thing to see carbon and nitrogen in this case determined powerfully towards the positive surface of the voltaic battery; but it is perfectly in harmony with the theory of electro-chemical decomposition which I have advanced.

772. _Ferro-cyanic acid_ and the _ferro-cyanides_, as also _sulpho-cyanic acid_ and the _sulpho-cyanides_, presented results corresponding with those just described (771.).

773. _Acetic acid._--Glacial acetic acid, when fused (405.), is not decomposed by, nor does it conduct, electricity. On adding a little water to it, still there were no signs of action; on adding more water, it acted slowly and about as pure water would do. Dilute sulphuric acid was added to it in order to make it a better conductor; then the definite proportion of hydrogen was evolved at the _cathode_, and a mixture of oxygen in very deficient quant.i.ty, with carbonic acid, and a little carbonic oxide, at the _anode_. Hence it appears that acetic acid is not electrolyzable, but that a portion of it is decomposed by the oxygen evolved at the _anode_, producing secondary results, varying with the strength of the acid, the intensity of the current, and other circ.u.mstances.

774. _Acetates._--One of these has been referred to already, as affording only secondary results relative to the acetic acid (749.). With many of the metallic acetates the results at both electrodes are secondary (746. 750.).

Acetate of soda fused and anhydrous is directly decomposed, being, as I believe, a true electrolyte, and evolving soda and acetic acid at the _cathode_ and _anode_. These however have no sensible duration, but are immediately resolved into other substances; charcoal, sodiuretted hydrogen, &c., being set free at the former, and, as far as I could judge under the circ.u.mstances, acetic acid mingled with carbonic oxide, carbonic acid, &c.

at the latter.

775. _Tartaric acid._--Pure solution of tartaric acid is almost as bad a conductor as pure water. On adding sulphuric acid, it conducted well, the results at the positive electrode being primary or secondary in different proportions, according to variations in the strength of the acid and the power of the electric current (752.). Alkaline tartrates gave a large proportion of secondary results at the positive electrode. The hydrogen at the negative electrode remained constant unless certain triple metallic salts were used.

776. Solutions, of salts containing other vegetable acids, as the benzoates; of sugar, gum, &c., dissolved in dilute sulphuric acid; of resin, alb.u.men, &c., dissolved in alkalies, were in turn submitted to the electrolytic power of the voltaic current. In all these cases, secondary results to a greater or smaller extent were produced at the positive electrode.

777. In concluding this division of these Researches, it cannot but occur to the mind that the final result of the action of the electric current upon substances, placed between the electrodes, instead of being simple may be very complicated. There are two modes by which these substances may be decomposed, either by the direct force of the electric current, or by the action of bodies which that current may evolve. There are also two modes by which new compounds may be formed, i.e. by combination of the evolving substances whilst in their nascent state (658.), directly with the matter of the electrode; or else their combination with those bodies, which being contained in, or a.s.sociated with, the body suffering decomposition, are necessarily present at the _anode_ and _cathode_. The complexity is rendered still greater by the circ.u.mstance that two or more of these actions may occur simultaneously, and also in variable proportions to each other. But it may in a great measure be resolved by attention to the principles already laid down (747.).

778. When _aqueous_ solutions of bodies are used, secondary results are exceedingly frequent. Even when the water is not present in large quant.i.ty, but is merely that of combination, still secondary results often ensue: for instance, it is very possible that in Sir Humphry Davy"s decomposition of the hydrates of pota.s.sa and soda, a part of the pota.s.sium produced was the result of a secondary action. Hence, also, a frequent cause for the disappearance of the oxygen and hydrogen which would otherwise be evolved: and when hydrogen does _not_ appear at the _cathode_ in an _aqueous solution_, it perhaps always indicates that a secondary action has taken place there. No exception to this rule has as yet occurred to my observation.

779. Secondary actions are _not confined to aqueous solutions_, or cases where water is present. For instance, various chlorides acted upon, when fused (402.), by platina electrodes, have the chlorine determined electrically to the _anode_. In many cases, as with the chlorides of lead, pota.s.sium, barium, &c., the chlorine acts on the platina and forms a compound with it, which dissolves; but when protochloride of tin is used, the chlorine at the _anode_ does not act upon the platina, but upon the chloride already there, forming a perchloride which rises in vapour (790.

804.). These are, therefore, instances of secondary actions of both kinds, produced in bodies containing no water.

780. The production of boron from fused borax (402. 417.) is also a case of secondary action; for boracic acid is not decomposable by electricity (408.), and it was the sodium evolved at the _cathode_ which, re-acting on the boracic acid around it, took oxygen from it and set boron free in the experiments formerly described.

781. Secondary actions have already, in the hands of M. Becquerel, produced many interesting results in the formation of compounds; some of them new, others imitations of those occurring naturally[A]. It is probable they may prove equally interesting in an opposite direction, i.e. as affording cases of a.n.a.lytic decomposition. Much information regarding the composition, and perhaps even the arrangement, of the particles of such bodies as the vegetable acids and alkalies, and organic compounds generally, will probably be obtained by submitting them to the action of nascent oxygen, hydrogen, chlorine, &c. at the electrodes; and the action seems the more promising, because of the thorough command which we possess over attendant circ.u.mstances, such as the strength of the current, the size of the electrodes, the nature of the decomposing conductor, its strength, &c., all of which may be expected to have their corresponding influence upon the final result.

782. It is to me a great satisfaction that the extreme variety of secondary results has presented nothing opposed to the doctrine of a constant and definite electro-chemical action, to the particular consideration of which I shall now proceed.

-- vii. _On the definite nature and extent of Electro-chemical Decomposition._

783. In the third series of these Researches, after proving the ident.i.ty of electricities derived from different sources, and showing, by actual measurement, the extraordinary quant.i.ty of electricity evolved by a very feeble voltaic arrangement (371. 376.), I announced a law, derived from experiment, which seemed to me of the utmost importance to the science of electricity in general, and that branch of it denominated electro-chemistry in particular. The law was expressed thus: _The chemical power of a current of electricity is in direct proportion to the absolute quant.i.ty of electricity which pa.s.ses_ (377.).

[A] Annales de Chimie, tom, x.x.xv. p. 113.

784. In the further progress of the successive investigations, I have had frequent occasion to refer to the same law, sometimes in circ.u.mstances offering powerful corroboration of its truth (456. 504. 505.); and the present series already supplies numerous new cases in which it holds good (704. 722. 726. 732.). It is now my object to consider this great principle more closely, and to develope some of the consequences to which it leads.

That the evidence for it may be the more distinct and applicable, I shall quote cases of decomposition subject to as few interferences from secondary results as possible, effected upon bodies very simple, yet very definite in their nature.

785. In the first place, I consider the law as so fully established with respect to the decomposition of _water_, and under so many circ.u.mstances which might be supposed, if anything could, to exert an influence over it, that I may be excused entering into further detail respecting that substance, or even summing up the results here (732.). I refer, therefore, to the whole of the subdivision of this series of Researches which contains the account of the _volta-electrometer_ (704. &c.).

786. In the next place, I also consider the law as established with respect to _muriatic acid_ by the experiments and reasoning already advanced, when speaking of that substance, in the subdivision respecting primary and secondary results (758. &c.).

787. I consider the law as established also with regard to _hydriodic acid_ by the experiments and considerations already advanced in the preceding division of this series of Researches (767. 768.).

788. Without speaking with the same confidence, yet from the experiments described, and many others not described, relating to hydro-fluoric, hydro-cyanic, ferro-cyanic, and sulpho-cyanic acids (770. 771. 772.), and from the close a.n.a.logy which holds between these bodies and the hydracids of chlorine, iodine, bromine, &c., I consider these also as coming under subjection to the law, and a.s.sisting to prove its truth.

789. In the preceding cases, except the first, the water is believed to be inactive; but to avoid any ambiguity arising from its presence, I sought for substances from which it should be absent altogether; and, taking advantage of the law of conduction already developed (380. &c.), I soon found abundance, amongst which _protochloride of tin_ was first subjected to decomposition in the following manner. A piece of platina wire had one extremity coiled up into a small k.n.o.b, and, having been carefully weighed, was sealed hermetically into a piece of bottle-gla.s.s tube, so that the k.n.o.b should be at the bottom of the tube within (fig. 68.). The tube was suspended by a piece of platina wire, so that the heat of a spirit-lamp could be applied to it. Recently fused protochloride of tin was introduced in sufficient quant.i.ty to occupy, when melted, about one-half of the tube; the wire of the tube was connected with a volta-electrometer (711.), which was itself connected with the negative end of a voltaic battery; and a platina wire connected with the positive end of the same battery was dipped into the fused chloride in the tube; being however so bent, that it could not by any shake of the hand or apparatus touch the negative electrode at the bottom of the vessel. The whole arrangement is delineated in fig. 69.

790. Under these circ.u.mstances the chloride of tin was decomposed: the chlorine evolved at the positive electrode formed bichloride of tin (779.), which pa.s.sed away in fumes, and the tin evolved at the negative electrode combined with the platina, forming an alloy, fusible at the temperature to which the tube was subjected, and therefore never occasioning metallic communication through the decomposing chloride. When the experiment had been continued so long as to yield a reasonable quant.i.ty of gas in the volta-electrometer, the battery connexion was broken, the positive electrode removed, and the tube and remaining chloride allowed to cool.

When cold, the tube was broken open, the rest of the chloride and the gla.s.s being easily separable from the platina wire and its b.u.t.ton of alloy. The latter when washed was then reweighed, and the increase gave the weight of the tin reduced.

791. I will give the particular results of one experiment, in ill.u.s.tration of the mode adopted in this and others, the results of which I shall have occasion to quote. The negative electrode weighed at first 20 grains; after the experiment, it, with its b.u.t.ton of alloy, weighed 23.2 grains. The tin evolved by the electric current at the _cathode_: weighed therefore 3.2 grains. The quant.i.ty of oxygen and hydrogen collected in the volta-electrometer = 3.85 cubic inches. As 100 cubic inches of oxygen and hydrogen, in the proportions to form water, may be considered as weighing 12.92 grains, the 3.85 cubic inches would weigh 0.49742 of a grain; that being, therefore, the weight of water decomposed by the same electric current as was able to decompose such weight of protochloride of tin as could yield 3.2 grains of metal. Now 0.49742 : 3.2 :: 9 the equivalent of water is to 57.9, which should therefore be the equivalent of tin, if the experiment had been made without error, and if the electro-chemical decomposition _is in this case also definite_. In some chemical works 58 is given as the chemical equivalent of tin, in others 57.9. Both are so near to the result of the experiment, and the experiment itself is so subject to slight causes of variation (as from the absorption of gas in the volta-electrometer (716.), &c.), that the numbers leave little doubt of the applicability of the _law of definite action_ in this and all similar cases of electro-decomposition.

792. It is not often I have obtained an accordance in numbers so near as that I have just quoted. Four experiments were made on the protochloride of tin, the quant.i.ties of gas evolved in the volta-electrometer being from 2.05 to 10.29 cubic inches. The average of the four experiments gave 58.53 as the electro-chemical equivalent for tin.

793. The chloride remaining after the experiment was pure protochloride of tin; and no one can doubt for a moment that the equivalent of chlorine had been evolved at the _anode_, and, having formed bichloride of tin as a secondary result, had pa.s.sed away.

794. _Chloride of lead_ was experimented upon in a manner exactly similar, except that a change was made in the nature of the positive electrode; for as the chlorine evolved at the _anode_ forms no perchloride of lead, but acts directly upon the platina, it produces, if that metal be used, a solution of chloride of platina in the chloride of lead; in consequence of which a portion of platina can pa.s.s to the _cathode_, and would then produce a vitiated result. I therefore sought for, and found in plumbago, another substance, which could be used safely as the positive electrode in such bodies as chlorides, iodides, &c.

The chlorine or iodine does not act upon it, but is evolved in the free state; and the plumbago has no re-action, under the circ.u.mstances, upon the fused chloride or iodide in which it is plunged. Even if a few particles of plumbago should separate by the heat or the mechanical action of the evolved gas, they can do no harm in the chloride.

795. The mean of three experiments gave the number of 100.85 as the equivalent for lead. The chemical equivalent is 103.5. The deficiency in my experiments I attribute to the solution of part of the gas (716.) in the volta-electrometer; but the results leave no doubt on my mind that both the lead and the chlorine are, in this case, evolved in _definite quant.i.ties_ by the action of a given quant.i.ty of electricity (814. &c.).

796. _Chloride of antimony._--It was in endeavouring to obtain the electro-chemical equivalent of antimony from the chloride, that I found reasons for the statement I have made respecting the presence of water in it in an earlier part of these Researches (690. 693. &c.).

797. I endeavoured to experiment upon the _oxide of lead_ obtained by fusion and ignition of the nitrate in a platina crucible, but found great difficulty, from the high temperature required for perfect fusion, and the powerful fluxing qualities of the substance. Green-gla.s.s tubes repeatedly failed. I at last fused the oxide in a small porcelain crucible, heated fully in a charcoal fire; and, as it is was essential that the evolution of the lead at the _cathode_ should take place beneath the surface, the negative electrode was guarded by a green-gla.s.s tube, fused around it in such a _manner as to expose only the k.n.o.b of platina_ at the lower end (fig. 70.), so that it could be plunged beneath the surface, and thus exclude contact of air or oxygen with the lead reduced there. A platina wire was employed for the positive electrode, that metal not being subject to any action from the oxygen evolved against it. The arrangement is given in fig. 71.

798. In an experiment of this kind the equivalent for the lead came out 93.17, which is very much too small. This, I believe, was because of the small interval between the positive and negative electrodes in the oxide of lead; so that it was not unlikely that some of the froth and bubbles formed by the oxygen at the _anode_ should occasionally even touch the lead reduced at the _cathode_, and re-oxidize it. When I endeavoured to correct this by having more litharge, the greater heat required to keep it all fluid caused a quicker action on the crucible, which was soon eaten through, and the experiment stopped.

799. In one experiment of this kind I used borate of lead (408. 673.). It evolves lead, under the influence of the electric current, at the _anode_, and oxygen at the _cathode_; and as the boracic acid is not either directly (408.) or incidentally decomposed during the operation, I expected a result dependent on the oxide of lead. The borate is not so violent a flux as the oxide, but it requires a higher temperature to make it quite liquid; and if not very hot, the bubbles of oxygen cling to the positive electrode, and r.e.t.a.r.d the transfer of electricity. The number for lead came out 101.29, which is so near to 103.5 as to show that the action of the current had been definite.

800. _Oxide of bis.m.u.th._--I found this substance required too high a temperature, and acted too powerfully as a flux, to allow of any experiment being made on it, without the application of more time and care than I could give at present.

801. The ordinary _protoxide of antimony_, which consists of one proportional of metal and one and a half of oxygen, was subjected to the action of the electric current in a green-gla.s.s tube (789.), surrounded by a jacket of platina foil, and heated in a charcoal fire. The decomposition began and proceeded very well at first, apparently indicating, according to the general law (679. 697.), that this substance was one containing such elements and in such proportions as made it amenable to the power of the electric current. This effect I have already given reasons for supposing may be due to the presence of a true protoxide, consisting of single proportionals (696. 693.). The action soon diminished, and finally ceased, because of the formation of a higher oxide of the metal at the positive electrode. This compound, which was probably the peroxide, being infusible and insoluble in the protoxide, formed a crystalline crust around the positive electrode; and thus insulating it, prevented the transmission of the electricity. Whether, if it had been fusible and still immiscible, it would have decomposed, is doubtful, because of its departure from the required composition (697.). It was a very natural secondary product at the positive electrode (779.). On opening the tube it was found that a little antimony had been separated at the negative electrode; but the quant.i.ty was too small to allow of any quant.i.tative result being obtained[A].

[A] This paragraph is subject to the corrective note now appended to paragraph 696.--_Dec. 1838._

802. _Iodide of lead._--This substance can be experimented with in tubes heated by a spirit-lamp (789.); but I obtained no good results from it, whether I used positive electrodes of platina or plumbago. In two experiments the numbers for the lead came out only 75.46 and 73.45, instead of 103.5. This I attribute to the formation of a periodide at the positive electrode, which, dissolving in the ma.s.s of liquid iodide, came in contact with the lead evolved at the negative electrode, and dissolved part of it, becoming itself again protiodide. Such a periodide does exist; and it is very rarely that the iodide of lead formed by precipitation, and well-washed, can be fused without evolving much iodine, from the presence of this percompound; nor does crystallization from its hot aqueous solution free it from this substance. Even when a little of the protiodide and iodine are merely rubbed together in a mortar, a portion of the periodide is formed. And though it is decomposed by being fused and heated to dull redness for a few minutes, and the whole reduced to protiodide, yet that is not at all opposed to the possibility, that a little of that which is formed in great excess of iodine at the _anode_, should be carried by the rapid currents in the liquid into contact with the _cathode_.