OSMOTIC PHENOMENA 31 



scribed. The diffusion tension of the solvent is greatest in 

 the pure solvent and decreases as the concentration of the 

 solution increases. 



If there were available a membrane permeable to the 

 solute, but impermeable to the solvent, this diffusion tension 

 of the solvent might be directly measured. It would be an 

 osmotic pressure similar to that occasioned by the solute 

 molecules, but of much greater magnitude and in the 

 opposite direction. Though there is no membrane which 

 will make this pressure evident, most of the phenomena of 

 osmotic pressure show that it exists. No membrane can 

 allow the solvent particles to pass absolutely without fric- 

 tion. Thus the question arises: Why is not this pressure of 

 the solvent made evident to a degree equal to the amount of 

 force needed to overcome this friction? The answer is, 

 obviously, that the pressure produced by the solvent on one 

 side of the membrane is practically equaled by that on the 

 other side. In solutions where the principles of osmosis 

 hold true, the dilution of the solvent, due to the presence of 

 the solute, is negligible. 



The following explanation of osmotic pressure has been 

 given by various authors. The quotation is from Davenport : * 

 "Upon the side containing the greater number of molecules 

 of salt [solute] fewer water [solvent] molecules will in a 

 given time strike the membrane than upon the other side; 

 and since the number passing through is proportional to the 

 number striking, relatively fewer molecules of water will 

 consequently pass out, and so there will be a resultant flow 

 of water to that side; and if the mass of water is confined, 

 it will exert great pressure." This explanation is untenable 

 for several reasons. Not nearly all solutions occupy more 

 space than the original mass of pure solvent from which they 

 were prepared. If to exactly a liter of water be added a 



1C. B. DAVENPORT, Experimental Morphology, Vol. I (New York, 1897), p. 71. 



