Influence of Voltaic Currents on the Diffusion of Liquids. 83 



a less extent at the surface of contact of the liquids with the meniscus 

 tube ; but exactly in what manner the current operates at the contact 

 surfaces is a difficult question. 



The effects of the current are both physical and chemicaL A 

 uniform cause or homogeneous force (such as an electric current is 

 commonly supposed to be), whilst acting under uniform conditions is 

 usually considered to produce uniform immediate effects. In the 

 present case, however, it appears as if one portion of the electric 

 current produces direct chemical decomposition of the solution ; a 

 second immediately disassociates water from saline matter by diffusion, 

 and a third produces direct electric convection, or simple mechanical 

 movement of the liquid ; or the diffusion may be viewed as a result of 

 electric convection. Either then the total current of uniform property 

 acts not under uniform conditions, but in several different ways, pro- 

 ducing effects which require different degrees of electromotive force to 

 produce them, or the current consists of several portions of electricity 

 of different electromotive force. The supposition that these different 

 effects require currents of different electromotive force to produce 

 them, is in harmony with the generalised idea, that instances of 

 mechanical union, or mere solution of solids in liquids, merges into 

 those of definite chemical union by insensible degrees, and the two 

 classes thus form an unbroken series, ranging from those of liquids 

 which mix in every degree with each other, to those of the most 

 powerful chemical combinations, in which the ingredients unite only 

 in rigidly definite proportions. 



With regard to the purely physical effects, viz., electric osmose and 

 transfer of mass, we may reasonably refer them to electric convection 

 attending electric discharge between surfaces in opposite electrical 

 states, the charged condition of those surfaces being dependent upon 

 the special local resistance to electric transfer which always exists in 

 greater or less degree at the surfaces of contact of bodies of different 

 physical and chemical structure. In capillary tubes the charged sur- 

 face of the glass becomes also an important circumstance. The 

 unequal transfer of acids and bases must also affect the results. Acids 

 always travel in greater proportion towards the anode than the metals 

 to which they have been united move towards the cathode, provided 

 the solutions are not kept quite neutral. 



And with regard to the electro- chemical effects, ions are probably 

 liberated at every surface of junction of liquids of sufficiently different 

 composition through which the current passes. A sufficient difference 

 of property of the two bodies (in addition to liquidity and a sufficient 

 electromotive power of the current) appears to be a necessary condi- 

 tion of electrolysis ; with lesser differences probably only electric 

 osmose is produced. (I have not examined the effects of electric 

 currents passing in oblique directions through a surface of contact of 



g 2 



