488 Prof. H. F. Weber's Researches on 



cedure was such that it could not possibly serve for rigorously 

 testing the accuracy of the assumed elementary law. Several 

 times in the last twenty years attempts have been made to 

 employ more exact methods of measuring, and expecially those 

 founded on optical principles, for tracing the course of hydro- 

 diffusion, — thus by Voit* in 1866, and by Johannisjanz t in 

 1877. The success, however, of these supposed more delicate 

 optical methods was no greater than that gained by Fick ; and 

 these new investigations do not at all make it evident that the 

 elementary law assumed as an hypothesis by Fick is in all 

 strictness the natural law that really governs what takes place 

 in diffusion. The more recent of these investigations press 

 upon us the conviction that either all the methods of measuring 

 hitherto made use of for the examination of diffusion are still 

 too rough and imperfect, or diffusion goes on according to an 

 elementary law very different from Fick's. Thus, for example, 

 the values calculated by Johannisjanz for the diffusion-constant 

 of a common-salt solution, on the basis of Fick's law, from the 

 observations of one day each, showed differences amounting to 

 20 per cent, from one day to another of the series of observa- 

 tions lasting several days ; further, he found a mean value for 

 the constant of diffusion of solution of common salt deviating 

 about 45 per cent, from the value previously found by Fick ; 

 and Johannisjanz could perceive no influence of temperature 

 upon the progress of the diffusion, while Fick proved a very 

 marked influence of tempej'ature. 



In the following memoir a new method will be described 

 which affords an extremely sharp test of the accuracy of the 

 elementary law of hydrodiffusion, while it is of a very simple 

 form, and permits the progress of the diffusion to be clearly 

 ascertained almost from moment to moment. 



During the last two years I have been constantly occupied 

 with the following problem: — to discover a unit of electromo- 

 tive force that could be identically reproduced at any time, 

 and to determine its quantity in absolute measure by two me- 

 thods as widely different as possible. On the occasion of those 

 researches I was obliged, in August and September of the 

 year 1877, to investigate thoroughly the influence of the con- 

 centration of aqueous solutions of a zinc- and a copper-salt upon 

 the electromotive force of the galvanic elements constructed 

 out of them. As had already previously been found, the result 

 obtained was, inter alia, that the electromotive force of the 

 Daniell element diminishes as the concentration of the sulphate- 

 of-zinc solution bathing the zinc-electrode increases, while it 

 increases as the degree of concentration of the copper-sulphate 

 * Pogg. Ann. vol. cxxx. t Wiedem. Ann. vol. ii. 



