IN A CLOSED GALVANIC CIRCUIT. 



237 



tures, it is very possible that the liquid, by flowing back through 

 parts of the apertures, may annihilate the result of transference 

 through other parts. 



At all events, from the failure of these experiments it must 

 not be concluded that the ph2enomenon of the motion of liquids, 

 by means of a galvanic current, is necessarily dependent upon 

 the presence of a porous partition. On the contrary, the expe- 

 riments yet to be recorded will show that the phaenomenon de- 

 pends solely upon the several constants of the galvanic current, 

 and that the porous partition serves merely to render visible 

 this motion of the liquids, which it effects by destroying causes 

 that would otherwise act against and modify the effect of the 

 galvanic current. 



§3. 



After establishing by the experiments above described, the 

 fact of a motion of liquids from the positive to the negative pole 

 of a closed circuit, an attempt was next made to discover some 

 quantitative relation between this phaenomenon and others already 

 established. 



Of the several instruments w^hich were successively con- 

 structed, for this purpose, the following appeared to be the 

 simplest and most convenient. 



On the top of the porous cylinder a (PL IV. fig. 2.), which was 

 closed underneath, a small bell-shaped glass, c, was cemented, 

 having an aperture into which a perpendicular tube d, with 

 lateral efflux tube e, was fitted. Within the porous one stood 

 a cylinder g of platinum foil. This was connected by means of 

 a wire,/, with one pole of the galvanic battery. This wire,^^ 

 was cemented into an air-tight glass tube let into the upper 

 part of the bell -shaped glass c. Externally, the porous cylinder 

 was surrounded by a second cylinder of platinum foil, which 

 by the wire k was connected with the other pole of the battery. 

 The whole apparatus stood in a wider glass cylinder, n, w^hich, 

 together with the porous cylinder, was filled with water or some 

 other liquid, and protected from dust by means of the glass- 

 plate cover, n. Before the tube e, a small weighed vessel, /, 

 was placed to collect the liquid issuing from e. Previous to 

 experiment, the porous cylinder made use of was boiled for 



