Chemistry and Physics. 377 



chiefly due to nitrogen and oxygen gases separately. — Verh. d. 

 deatsch. phys. Gesellsch., No. 21, p. 953, 1911. h. s. u. 



7. The Mechanism of the Semi-permeable Membrane, and a 

 New Method of Determining Osmotic Pressure. — Heretofore, the 

 semi-permeable membranes used in measuring osmotic pressures 

 had to fulfil the condition of rigidity, either directly or by being 

 deposited in a rigid support. A very ingenious scheme for abol- 

 ishing this difficult requirement has been devised and tested by 

 F. T. Trouton. Although the principles involved are of a gen- 

 eral nature, it will be conducive to clearness to restrict the follow- 

 ing explanation to a typical case. 



Suppose the problem is to find the value of the osmotic pressure 

 of an aqueous solution of pure sugar of a given concentration. 

 For theoretical purposes, we may imagine a rectangular glass 

 vessel, of the type often used in stationary storage cells, separated 

 into two compartments by a vertical, transverse, impervious 

 diaphragm, e. g., a sheet of glass. This diaphragm must not 

 extend to the level of the top of the vessel. One compartment is 

 nearly filled with water, and the other with sugar solution. 

 These liquids are then placed in hydrostatic communication by 

 having superposed upon their upper surfaces a layer of liquid 

 ether of sufficient depth to completely submerge the upper edge 

 of the partition. Since sugar is insoluble in ether, while ether 

 dissolves a small percentage of water, it follows that the layer of 

 ether will take the role of the usual effectively rigid, semi-per- 

 meable membrane. In fact, this is the key-note of Trouton's 

 innovation, namely, to substitute a liquid semipermeable partition 

 for a rigid one. To be sure, both water and sugar solution take 

 up some ether, but this complication is not serious since it can be 

 relegated to the sphere of determinate corrections. Ether dis- 

 solves about 1*05 per cent of water when placed in contact with 

 the same, but ether absorbs less than this from a sugar solution, 

 the amount depending upon the concentration of the solution. 

 For equilibrium at the water-ether surface the ether must, there- 

 fore, contain 1*05 per cent of water, while at the solution-ether 

 interface a smaller quantity is necessary to establish equilibrium. 

 Diffusion through the ether prevents this equilibrium from being 

 established, consequently, water will pass across from the water 

 side to the solution side of the partition. If the ether could rig- 

 idly maintain its position so as to prevent any increase in the 

 volume of the sugar solution, the hydrostatic pressure of this 

 solution would increase, due to the accession of water. Under 

 these ideal circumstances the process would come to an end when, 

 owing to the increase of pressure, the percentage of water 

 absorbed by the ether from the sugar solution attained the same 

 value as the fraction of pure water taken up by the ether at 

 atmospheric pressure. The pressure competent to effect this 

 state of equilibrium in the ether would be the equivalent of the 

 osmotic pressure of the sugar solution. 



Am. Jour. Sci.— Fourth Series, Vol. XXXIII, No. 196.— April, 1912. 

 25 



