INFLUENCE OF MEMBRANES UPON OSMOTIC CURRENTS. 481 



tubes of capillary diameter, that this force has been called 

 capillary attraction. 



There is no essential difference between the phenomena 

 of imbibition by porous substances and the elevation of 

 liquids in capillary tubes. In both, it is necessary that the 

 liquid should be capable of wetting the surfaces to which it 

 is exposed. The imbibition of water by a sponge is depend- 

 ent upon capillary attraction. If the liquid thus taken up 

 remain and fill the interstices, no current takes place ; but 

 if it be constantly removed by evaporation, or otherwise, 

 a constant now is produced. It is in this way that an 

 alcohol-lamp is emptied of fluid, which is taken up by the 

 wick and lost by evaporation. This single current is pro- 

 duced by the diffusion of the liquid in the atmosphere in the 

 form of vapor but if an apparatus be constructed in which 

 the liquid absorbed is gradually diffused in another liquid, 

 the result will be the same. Suppose, for example, that an 

 endosmometer be constructed in which the septum consists of 

 a porous substance, such as unglazed earthenware. Let this 

 apparatus be so filled with a saline solution, and the septum 

 immersed in pure water. The porous septum will then 

 gradually take up the water and bring it in contact with the 

 saline solution ; when, in obedience to the law of diffusion 

 of liquids, the water passes in a diffusive current to the saline 

 solution, and a steady current wUl be established through 

 the septum. As diffusion can only take place between liquids 

 which are miscible with each other, this condition is indis- 

 pensable to the production of a current. If we now suppose 

 that both liquids are capable of wetting the septum, in such 

 an apparatus as we have described, they will meet in the 

 substance of the septum. Once in contact with each other, 

 whether within or without the septum, the liquids will 

 diffuse ; the water flowing toward the saline solution, which 

 in its turn flows toward the water. In this way, two currents 

 are produced. It has been found in experiments of this kind 

 that the two currents are generally unequal, the more power- 

 31 



