Chap. XXII.] 



THE ENDOSMOMETER. 



253 



First, it is to be noticed that the two liquids on 

 opposite sides of the membrane must be miscible. 

 Liquids that cannot mix when in direct contact with 

 one another can mix still less through 

 a septum. They, or at least one of them, 

 must be capable also of permeating the 

 membrane. Secondly, there are two ele- 

 ments in the process, that of the mixture 

 of the dissimilar liquids and that of the 

 increased volume of one of them. 



The increase of volume does not always 

 take place on the side of the fluid of 

 greater density. If salt solution be on one 

 side and water on the other, the increased 

 volume will be on the side of greater 

 density, that of the salt solution ; but osuiometer. 

 when alcohol is on one side of the animal 

 membrane, and water on the other, the increased 

 volume is on the side of the alcohol, the side of 

 less density. The character of the membrane has 

 to do with the change of volume; for while, as just 

 noted, when water and alcohol are separated by an 

 animal membrane more water passes to the alcohol 

 through the membrane than alcohol to the water, 

 when water and alcohol are separated by a septum of 

 caoutchouc the alcohol passes in greater abundance 

 through the membrane, and the volume on the side of 

 the water is increased. Further, the mixture of the dis- 

 similar liquids can still be carried on when the change 

 of volume is forcibly prevented. This is proved 

 by an experiment described by Liebig. A short 

 wide tube is connected to a long narrow tube, the 

 narrow tube being vertical and the wide tube standing 

 out from it. The wide tube is filled with brine, and 

 closed with a piece of bladder. Down the vertical 

 narrow tube mercury is now poured, whose pressure 

 is exerted against the brine in the wide tube, and 



