SOME CONSEQUENCES OF GRAHAM'S WORK 607 



entering gas gives rise within the vessel may be maintained. 

 Thus if a heated palladium vessel containing nitrogen, at a 

 pressure say of 200 millimetres, be surrounded with hydrogen 

 at atmospheric pressure, the hydrogen enters but no nitrogen 

 passes out ; ultimately, when the vessel is full of hydrogen, in 

 the sense that the molecules of this gas pass out as fast as they 

 pass in, the pressure within the vessel rises to that of the 

 hydrogen without plus that originally exerted by the nitrogen ; 

 the excess of pressure within the vessel is obviously that pro- 

 duced by the impact of the nitrogen molecules beyond that 

 produced by the impact of the hydrogen molecules. The 

 palladium therefore acts as a differential septum in separating 

 hydrogen from nitrogen. 



In like manner, differential septa may be prepared through 

 which water will pass but not dissolved substances such as 

 sugar and salt. If a closed vessel impermeable to sugar, pro- 

 vided with a fine gauge and filled with a solution of sugar, be 

 placed in water, the pressure gradually rises to a maximum 

 corresponding to the strength of the solution, the rise being due 

 to the passage of water inwards. For reasons which are in no 

 way clear to us at present, in dilute solutions the pressure 

 developed in the presence of substances such as sugar is the 

 same as the pressure the substance would exert if present in 

 the gaseous state in the volume occupied by the solution. 



The pressure shown by the gauge is unfortunately spoken 

 of as the osmotic pressure of the solution and has been attributed 

 to the dissolved substance directly, as though it were acting as 

 a gas. But the two cases are very different. In the one, there 

 is room for the entry of the gas and all that happens is that the 

 pressure rises within the vessel; moreover, the gas which cannot 

 pass out exercises no attractive power upon the entering gas. 

 In the case of the solution, as liquids are but very slightly com- 

 pressible, there is practically no room for the entering water and 

 expansion takes place into the gauge. There can be no doubt 

 that osmotic pressure, so-called, is not a pressure in the ordinary 

 sense but a negative pressure or attraction ; solutions in water, 

 in fact, are all more or less attractive of water. 



When diffusion takes place into water from a solution in 

 water, obviously as the substance diffuses out from the solution 

 water must diffuse into the solution ; diffusion must be a process 

 of exchange. It is conceivable that the attraction which solutions 



