1880.] 



Voltaic Currents on Diffusion of Liquids. 



251 



neutral and alkaline ones, &c, also of different degrees of dilution, 

 the influence upon the effects of the kind and strength of the liquids, 

 the strength of the current, and of various other circumstances, was 

 determined. Two different apparatuses were employed in these ex- 

 periments, viz., a single and a double meniscus one ; the former was 

 the most easily managed, and yielded the phenomena in their simplest 

 form. 



It was found that the phenomena are very complex, and consist of 

 a mixture of physical and chemical effects. That they are chiefly due 

 to electrolytic changes, to differences of specific gravity, to ordinary 

 liquid diffusion, to electrolytic transfer and diffusion of liquids, and to 

 heat of conduction-resistance. That also the mutual contact surfaces 

 of the liquids act to some extent as electrodes and osmotic diaphragms. 

 That by the electrolytic action, various phenomena, such as layers of 

 liquid of different specific gravities, &c, lines, strata of colour, altera- 

 tion of form of the surfaces of the meniscuses, liquid currents, &c, 

 were produced. That by the heat evolved, expansion and streaming 

 upwards and downwards of the liquids, evolution of gas and steam, and 

 boiling of the liquids at their contact surfaces took place. Nearly all 

 • the phenomena and their conditions have been examined and discussed, 

 but some of them, however, have not been fully explained. 



The movements of the meniscus and apparent transfer of the bulk 

 of the liquid were examined in a special apparatus in several ways, 

 and a conclusion was drawn from these and the previous experiments, 

 viz., that by the causes already named an apparent, and, also to a small 

 extent, real bodily movement of the liquid did occur. 



Some experiments were also made with two superimposed electro- 

 lytes in capillary tubes for the purpose of comparing the effects pro- 

 duced on passing an electric current through them, with those which 

 occur in a capillary electroscope. The conclusion arrived at was, that 

 the two classes of movement were considerably different, and that the 

 movements obtained with the two electrolytes were more allied to 

 those observed by Armstrong (" Phil. Mag.," xxiii, 1843, pp. 194-202), 

 Quincke ("Pogg. Ann.," vol. cvii, 1859), Jurgensen and Wiede- 

 mann, on passing electric currents through non-conducting liquids in 

 capillary tubes. 



By means of numerous and varied experiments with apparatuses 

 similar to that used by Porrett (Thomson's " Annals of Philosophy," 

 vol. viii, 1816, pp. 74-76) in electric osmose experiments, the influence 

 of electric diffusion of liquids upon the results was examined. From 

 the results of these it was inferred that the apparent movement of the 

 mass of the liquids was considerably different from that of electric 

 osmose ; in the osmose experiments it was found that the amount of 

 liquid transferred was nearly always greater when the current passed 

 from a weak to a strong solution of the same salt or acid than when it 



