1881.] 



caused by Liquid Diffusion and Osmose. 



297 



in glass siphons having turned-np capillary ends. As, in his arrange- 

 ment, acid and alkaline liqnids were in contact with each other, 

 chemical action occurred. Fechner employed a similar apparatus, and 

 'in some of the combinations of liquids employed by him, chemical 

 action took place from a similar cause. In Wild's research, each 

 element consisted of an open cup having two long and vertical wide 

 glass tubes projecting downwards from its lower surface and termi- 

 nated at their lower ends by electrodes. Three kinds of liquid, in 

 four portions, of suitable specific gravities, were carefully placed upon 

 each other in a series of distinct layers, without intervening solid 

 diaphragms, the two portions in contact with the electrodes being pre- 

 cisely similar. Liquids which did not chemically act upon each other 

 were in nearly all cases employed ; but as the electrodes were electro- 

 coppered metallic capsules, and any electric current produced by 

 liquid diffusion has extremely feeble electromotive force, it is probable 

 that chemical action interfered. L. Schmidt (of Halle) employed a 

 perfectly similar apparatus and arrangement of liquids. In Wiillner's 

 arrangement, the liquids were in a wide cylinder formed of several 

 horizontal segments, the liquids being in distinct layers, separated by 

 disks of animal membrane, and the electrodes were circular sheets of 

 brass or copper.* In each and all of these apparatuses and arrange- 

 ments, three distinct liquids were employed, and the phenomena were 

 not reduced to their simplest form by the use of two liquids only. 



In my earliest experiments, I employed several different arrange- 

 ments of apparatus, in which a very limited number of layers of 

 liquid of different degrees of concentration were separated from each 

 other by horizontal diaphragms of parchment paper, the terminal 

 layers having small platinum wire electrodes connected with a 

 Thomson's reflecting galvanometer. No signs of an electric current 

 were however obtained. 



As electromotive force may be of all degrees of strength, from that 

 which only causes two liquids to feebly mix to that by which the most 

 fixed compounds are decomposed; and as the energy of mechanical 

 diffusion and osmose is extremely feeble in comparison with that of 

 chemical union, I concluded that any electric current arising from 

 those causes would probably have very feeble electromotive power, and 

 I therefore made other trials with a larger number of elements. 



In order to try more effectually whether a difference of facility of 

 diffusion, caused by the action of gravity upon two portions of solution 

 of different degrees of concentration and of specific gravity, would 

 produce an electric current, I made the following experiments : — 



Experiment 1. — I took a series of twenty-five small glass tumblers, as 



* See also Kamtz, " Schweigg. Jnl.," 1829, vol. lv, p. 1 ; Lagarve, " Grilb. Ann.." 

 vol. xiv, 1803, p. 230 ; Kohlrausch, " Pogg. Ann.," vol. hurix, 1850, p. 200 ; Worm 

 Miiller, " Pogg. Ann.," vol. cxl. 



z2 



