ATTRACTION OF THE MEMBRANE FOR LIQUIDS. 25 



particles, it is impossible to explain how water and brine can move in the same 

 tube with unequal velocity in opposite directions ; the two liquids being supposed 

 to have a mutual attraction, that is, to be miscible. This attraction must act with- 

 in the tube just as well without ; and we might, therefore, suppose, that when the 

 two liquids have become mixed, the mixture could only move in one direction with 

 a medium velocity. 



Assuming that a mixture is formed in the open orifices of the pores or tubes, or 

 in any part of them, it is difficult to see, why saline particles should not pass from 

 one side to the water, or aqueous particles to the saline ones in the bladder, since 

 the mutual attraction must be regarded as equal on both sides. The chemical 

 affinity of the two liquids does not explain the efflux. 



If we suppose, that in certain pores only brine, in others only pure water 

 moves, the phenomenon ought not to occur when all the pores are filled with 

 water or with brine, or when the tube is tied with a double, treble, or fourfold, 

 bladder. But the properties of bladder are seen in the finest as well as thickest 

 membranes, and one, two, or three layers make no difference in the ultimate 

 result.^) 



The kind of influence which the nature of the partition, or its attraction for the 

 liquids in contact with it, exerts on the phenomenon, is seen by comparing the action 

 of an animal membrane with that of a thin sheet of caoutchouc.* 



In a tube, closed with bladder, which is filled with alcohol, and immersed in 

 pure water, the volume of alcohol is increased ; more water passes to the alcohol 

 than alcohol to the water.t 



If, without making any other change in the experiment, the tube be closed with 

 a thin sheet of caoutchouc, the volume of the alcohol now diminishes while that of 

 the water increases. 



Here, all the circumstances of the mixture of the two liquids have remained the 

 same except the nature of the partition, which makes the difference in the result. 



When we fill with brine a tube, closed with bladder, (Fig. 8,) j^ ff g 



and place it in a vessel of water, so that the bladder and water ' 



only communicate by a single drop, the liquid in the tube 

 increases in bulk, and rises in the tube, as if the bladder had 

 been immersed in the water; but the drop becomes gradually 

 smaller, till after an hour or two, a complete separation takes 

 place, and the drop tears itself away from the water.( 2 ) 



If the cause of the change of volume in this experiment were 

 the unequal resistance which the bladder opposes to the passage 

 of the two liquids with equal attraction (equal force) on both 

 sides, the phenomenon just described would be inexplicable ; for 

 a resistance can no doubt impede, but is not capable of producing 

 motion. But we see, that the water in this experiment is raised 

 to a higher level, and moreover, the tearing asunder of the drop can only be the 

 effect of a powerful attraction, residing in the substance of the bladder. 



(') With respect to the theory, that, when a saline solution is mixed with pure 

 water, if the two liquids are separated by a membrane, particles of salt alone pass 

 through the pores of the bladder to the water, and particles of water alone to the brine, 

 the following experiments may throw some light on the question. For the sake of 

 greater accuracy, the results were determined by weighing. The apparatus, Fig. 3, was 

 used. The tube contained 8'67 grammes of saturated brine, in which were 2'284 gram- 

 mes of salt and 6'38 of water. After 24 hours it had gained 1'79 grammes in weight 

 and it now contained only 0-941 grammes of salt. It had therefore lost 1-343 grammes 

 ot salt, and gained 3'13 of water. According to the above theory, 1 atom of salt and 



[ 2 ] If we pour into a tube, of an inch wide, and closed with bladder as much 

 mercury as covers the surface of the bladder, then fill it with brine, and place it in 

 pure water, the volume of the liquid in the tube increases exactly as if the mercury 

 were not there. J 



* The nature of the membrane has an important influence. 

 T Experiment with bladder, and with caoutchouc 



4 



