ELECTROLYSIS 113 Radicles as NH 4 and S0 4 behave as elements, and have each a definite position in the series. Theory of Electrolysis. p . Any hypothesis which seeks to account for the phenomena of 4s of electrolysis has mainly to deal with the two points (1) that the lt j. ions appear only at the electrodes, and (2) that the electricity at the 3i same time is conducted between the electrodes. From the behaviour of electrolytes in contact on the passage of the current we conclude that if we had a series of cells consisting alternately of K.C1 and XaCl, the result of electrolysis might be represented thus : Before electrolysis, after it, KCl,NaCl,KCI,NaCl; -KXaCl,KCl,NaCl,Clf. Now, we may suppose similar effects to occur if the cells were all identical, and farther we may consider the collection of molecules in any electrolyte as such a series of cells in contact, and argue the electrolytic process to be a series of decompositions and recombina tions along a line of molecules resulting finally in the decomposition of molecules at the electrodes alone. The decomposition of any oxygen salt would be similar, with the exception that the one ion is complex. Thuo might represent the decomposition of zinc sulphate. This idea of alternate decompositions and recombinations was originally sug gested by Grotthuss in 1805, who, however, attributed the separation to attractions, due to the electrodes, varying inversely as the squaie of the distance. Faraday (Esp. Tic.?., 481 -563, series iv.) discusses the theory, and, while denying the attractions of the electrodes defends the idea of decomposition and reformation, chiefly against Dela Rive and Kiil ault and Chom pro, and considers thattlie etlect of the passage of the current is due to a change of the chemical affinities of (he components of the electrolyte, and he points out (1343 sqq., ser. viii.) that the decomposition is probably preceded by a polarized state of the particles, as explained by him in his theory of electro static induction. This is confirmed by experiments of Tribe (Proc. Hoy. Soc., 1875-6), who inserted 198 small strips of silver in rows, parallel to the line joining the electrodes in dilute CuS0 4 , and observed that copper was deposited on the ends facing the anode, while gas was given off from the other ends; by comparing the amounts of the deposit he explored the electric field, showing that it was roughly similar to the magnetic field due to a north and a south polo. itorf s Many investigators have suggested additions to Grotlhuss s jory. hypothesis (see Viedemaun, i. 421, <.), and in particular Hittorf (I u j 1 ]. A ii ii., Ixxxix. ) has expanded it to explain the migration of the ions in salt solutions investigated by him. He supposes that the molecules are equal distances opart, and that the ions when separated travel with different velocities to the points of recombination, and consequently those points are not the middle points between pairs of adjacent molecules. He thus considers that the cation travels th of the distance between molecules while the n auiou travels ( 1 - - )th of the distance. If then we suppose the ions separated at the electrodes to be remoyed, we may imagine the positions of the particles in the medium of solution before and after electrolysis to be represented by a. respectively, thus (where?i-4) : the effect will be the same as if (1) the particle nearest the negative electrode were removed, and all the other particles in the line moved towards that electrode thof the distance between the particles, or. as if (2) the particle at the positive electrode were removed, and all the rest shifted ( 1 - Jth part of the distance between the particles towards it. If we suppose the solvent separated by a porous wall into two portions, we shall have after electrolysis in the portion containing the cathode (en the first supposition) corresponding to deposition of one equivalent of hycuogen in the voltameter (1) a gain of one equivalent of cation deposited; (2) a loss of one equivalent of salt since the decomposed mole cules are supposed taken from there ; (B) a gain of equivalent of salt due to translation. Hence the whole increase in the amount of the cation, free and combined, due to electrolysis is equivalent. In the portion containing the anode we shall have (1) a gain of oue equivalent of anion set free, and (2) a loss of equivalent of electro lyte due to the translation ; and hence the whole increase in the amount of auion free and combined round the anode is ( 1 I V n / equivalent. The same results are obtained if the second supposition be made. Hence the n here used is identical with the n used above in the account of the phenomena of migration of the ions. D Almeida considered that the phenomena were due to the fact that round the positive electrode an envelope of free acid was formed by electrolysis, and that this became a second electrolyte in contact with the salt solution. It is also evident that by supposing the salt to be electro lysed in a hydrated state, i.e. combined with a number of molecules of water which may travel with either the anion or cation, an ex planation of the phenomena may be arrived at (Burgoin, Bull. Soc. Chun. [2] xvii. 244). Hittorf explained the lemaikable cases of the iodides of zinc and cadmium by a somewhat similar assump tion (v. supra, p. 110). F. Kohliausch (Nacht. i. d. K. Gcs. d. JFiss., Gottingen, ]7 Mai Exten- 1876, 4 April 1877) has recently pointed out a most remaikable sion by relation between the migration constants and the conductivities of Kohl- extremely dilute solutions containing electrochemical ly equivalent rausch. amounts of haloid or oxygen salts. Thus if ? ? a be the conduc tivities of such solutions of two salts MK, MR " containing one component M the same in both, and if M,, n 2 be the corresponding migration constants of the ion M, then Kohliausch shows that the equation,- - - holds with remarkable accuracy for many salts. The quantities 7 1} 7., are called the "specific molecular conductivi- k ties 1 of the solutions, and are defined by the equation 1= , where ,-777 is the specific conductivity of a very dilute solution at 18 C. referred to mercury, and u is the ratio of the number of grammes of salt per unit of volume of solution to its electrochemical equiva lent in hydrogen units. The results are in accordance with the hypothesis that the conductivity of an electrolyte is proportional to the sum of the oppositely directed velocities of the anion and cation ; the velocity of any ion is supposed to depend on the friction of the surrounding fluid, and is accordingly constant for the same ion in diiierent solutions if these are extremely dilute. A table of relative velocities can be formed from the migration constants of Hittotf, Wiedemann, and Weinke; a multiple of these velocities gives numbers such that the sum of those corres ponding to two ions gives a value for the molecular conductivity of a solution of the compound of the two ions agreeing very closely with the experimental determinations of Kohlrausch and Grotnan. In order to explain the conduction of electricity during electiolysis Hypo- several hypotheses have been suggested, which involve tfie idea thesis of of continuity of electricity, even in the molecules. On this view conduc- the elementary atoms when lecombining carry with them a certain tiou. quantity of electricity, which, indeed, by Faraday s law, must be the same for every group of atoms constituting a chemical equiva lent. Berzehus, for example, considers that when two atoms, c g., II and L l, unite to form HC1, electric distribution takes phice similar to that of magnetism in a bar magnet, H being the posi tive pole, Cl the negative. The hypothetical positive pole is then attracted by the negative electrode, and the attraction is so great at the electrode itself as to overcome the chemical affinity of the H for Cl, and separation is the lesult, while the electricity of the electrode and of the II combine and are neutralized. The liberated chlorine atom then behaves in the same way towards its next neighbour, and so the current of electricity is set up. For an account of the allied theories see Wiedemann, I.e., and see Hypo- also Clerk Maxwell s icmaiks upon the subject (Eke. Mag., vol. i. thesis of 259 sqq.). These hypotheses neatly all involve the idea, raore Clausing, or less defined, of a statical molecule, i.e., a molecule at lestielatively to other molecules, and consisting of relatively fixed atoms; but while we legnrd heat as the energy of molecular motion, this notion of a molecule cannot be sustained, and accordingly the above hypo thesis can serve, as Maxwell suggests, merely to give precision to our ideas. ClausiustAx/gr. ^H.,ci. 338), however, has applied the kinetic hypothesis of the constitution of bodies to electrolysis, and from his suggestions we can form some conception of the method of proceed ing in electrolytic action. He supposes the molecules in the ordi nary state to be in a state of agitation, and the atoms composing the molecules to be also in motion, sometimes separating, some times recombining with other separated atoms, so that decomposi tion and recombinations are continually going on, but in no definite direction. The mean result is an apparent state of equilibrium. When, however, an electromotive force acts upon the electrolyte,
vi n. --15