ON ELECTEOLTSIS IN ITS PHYSICAL AND CHEMICAL BEARINGS. 311 



of tlie ions is disturbed by collision with other ions (or molecules), a certain de- 

 parture from equal velocities will arise. This departure is, nevertheless, so small 

 that it cannot be regarded as accurately established. _ , ^ ^. 



But if one considers greater concentration, the molecules are jammed together, 



to a degree which may be represented by — x lO-^, where /x is the molecular 



conductivity of the given concentration. Thus arise double ' molecules, treble 

 molecules, and so forth. The electrolysis of a double molecule l^J^ occurs either 

 according to the scheme I + U., or according to the scheme I^J + J ; though indeed 

 the former occurs much the more often, 



To take an example : if all the molecules in a solution are double molecules, 

 then, according to the first alternative (if one assumes all ions to travel at thesame 

 pace, which is nearly correct), analysis will show that I remams quite still (i.e. it. 

 wanders equally in opposite dh'ections), and the whole motion is performed by J.^ 

 According to the second alternative, one is led direct to the opposite result it 

 part is decomposed under scheme 1 and the rest under scheme '2, all intermediate 

 conditions can be represented. This is, I believe, the true cause of the so-called 

 unequal wandering of the ions. . , ^ , . , 



That water has an extremely small conductivity, its chemical behaviour makes 

 clear. Indeed it will probably never be detected by electrolysis. The phenomena 

 accompanying the electrolysis of CuSO, are explicable in the simplest way by 

 supposing that CuSO. is decomposed by water, even though but very little (one 

 faiows that CuSO, has an acid reaction), and naturally the so-formed sulphuric 

 acid takes part in the electrolysis, whence arises the free acid at anode, ihis w-iii 

 after all be scarcely noticeable. With stronger currents II,,SO^ appears at the 

 kathode, and the hydrogen of this will be gradually transported to the anode by 

 electrolysis. . . ., ... . 



The experiment with MgSO, can be explained in a similar way, without 

 supposing a noticeable electrolysis of water. The dilute solution contains much 

 more free II,SO, than the concentrated. So if a current flows according to 

 annexed scheme — 



MgSO^ . . . MgSO, 



H,S04 



H.;S04 

 •V : — • 



dilute sol. 



-^^ 



MgSO, .... MgSO, 



MgSO, 



H-SO, 



■* -^ 



strong sol. 



the left-hand side of the partition will lose two H.,SO, and gam one, so the solu- 

 tion will become alkahne ; whereupon Mg(OH).„ which is only soluble m excess ot 

 acid, will precipitate. This Mg(OH),, we can neglect in electrolysis, for it probably 

 conducts no better than NH3. Besides, CO, in the solution could cause a small 

 precipitate of MgC03. ... , -^ 



I should be very glad to hear your view of the expositions m my essay,| it you 

 will Mndly pass an opinion on it. You will probably think my enunciations too 

 bold, and that I have insufficiently established my conclusions ; meanwhile, I may 

 refer to the work of Ostwald, who has found my statements, m § 15 ot the second 

 part, to fully correspond with experiment ; also, I can refer to the simultaneous 

 work of Bouty (February 1884 ; my work was undertaken June 6, 1883), and 

 the later investigations of Kohlrauscb, which completely prove that with extreme 

 dilution all salts examined conduct about equally well (Part I. page 41, law o). 



It follows moreover from Ostwald's experiments that the law 41 (Part 11. p. 4bj, 

 which I have deduced from purely theoretic considerations, corresponds exactly 

 with experience. ... 



I hope that these circumstances will mitigate your criticism ot the many 

 incompletenesses, since they show that 1 have gone at least partly on right Imes, 



» First part, ' On the Conductivity of very Dilute Solutions ' ; second part, 'On 

 the Chemical Theory of Electrolytes.'— ^cafZ. des Sciences de Suede, June 1883. See 

 below, page 357. 



