IN A CLOSED GALVANIC CIRCUIT. 235 



glass margins at b are ground on to each other, and held firmly 

 together by means of the board e e, and the three screws fifnf^- 

 The glass margin, b, of one of the vessels is ground internally, so 

 that in the hollow thereby produced, plates of different substances 

 can be cemented, and thus introduced between the two glass 

 vessels. In the necks d d,, two vertical glass tubes, g g^, are fixed 

 by corks. The necks, c Cp are also closed with corks, through 

 which two wires, h h,, pass into the interior of the glass vessels. 

 To the interior extremities of these wires, circular metallic plates, 

 i ip are screwed ; these at first were of platinum foil. After 

 cementing a porous plate of clay between the vessels at b, the 

 latter were filled with water. When the wire h was connected 

 with the positive, the wire h^ with the negative pole of a galvanic 

 battery, the water ascended in the tube g, and sunk in the tube 

 g ; at the same time oxygen was liberated at «, hydrogen at i^ ; 

 both of which escaped through the two tubes g, g^. 



The quantity of gas thus liberated is so small in proportion 

 to the quantity of water raised, that we are by no means justified 

 in ascribing this ascent to the liberation of gas alone. [In one 

 case the total volume of liberated gas in the apparatus amounted 

 only to about one-third of the volume of water elevated into 

 the tube ^y.] 



Another experiment showed this more plainly. Instead of 

 the platinum plate, a plate of copper was substituted; and in 

 place of water, a solution of sulphate of copper was used. Here 

 no liberation of gas took place ; but the solution nevertheless 

 ascended in the tube g^ and sank in the tube g. 



This phaenomenon must therefore have a different cause, and 

 can only be explained by assuming that a transference of the 

 liquid particles takes place from the metallic plate i in con- 

 nexion with the positive pole, through the porous diaphragm to 

 the plate i^ in connexion with the negative pole. A peculiar 

 moving action of the galvanic current is the only cause to which 

 we can attribute this transference. Before endosmose could 

 cause it, the liquid on both sides of the porous clay must have 

 become heterogeneous; hence, when pure water alone was 

 employed, the phaenomenon should not at all have shown itself. 

 Further, the motion of the liquids by the galvanic current is 

 much more active than that which endosmose could produce ; 



t) 2 



