XI. THE PLANT 223 



by which a substance, dissolved in a solvent, moves from the more 

 concentrated to the less concentrated portions of the liquid. This 

 movement is an indication of a motion which is constantly occurring 

 in the particles of a dissolved substance, but which is only readily ap- 

 parent when more of the substance moves in one direction or into a 

 given space than moves in the other direction or out of it. 



Graham in 1850, investigated the rates of diffusion of various com- 

 pounds in aqueous solution and found great differences with different 

 substances. Many crystallisable substances moved comparatively 

 rapidly, while colloidal bodies, e.g., gum, albumin, tannin, moved with 

 extreme slowness. He found, moreover, that colloidal bodies, either 

 in solution or when thoroughly wetted with water, allowed crystalloids 

 to diffuse freely, but practically stopped all diffusion of dissolved 

 colloids. 



The practical application of this phenomenon to the separation of 

 crystalloids from colloids is known as dialysis and is usually ac- 

 complished by means of a vessel, the bottom of which is composed of 

 animal or vegetable parchment. The mixture of the two dissolved 

 substances is placed within this vessel, which is then floated upon 

 pure water ; the crystalloid passes through the membrane, but the 

 colloid does not. The diffusion of the crystalloid goes on with dimin- 

 ishing speed until the concentration of the solution within and without 

 the dialyser becomes equal. Even then, it is to be clearly understood, 

 the action probably does not cease ; but, since just as much leaves 

 the inner vessel in a given time as enters it, the process does not 

 readily show itself. 



By repeatedly renewing the water in the outer vessel the whole of 

 the crystalloid may be removed from the liquid within the dialyser, 

 while the amount of colloid is not appreciably diminished. 



Many animal and vegetable membranes, consisting as they do of 

 colloidal matter, capable of swelling in water, allow of ready diffusion 

 of water through them. 



Most of these membranes, too, offer little resistance to the diffusion 

 of dissolved crystalloids, but stop that of dissolved colloids. Some, 

 however, while permitting water to diffuse through readily, almost 

 entirely stop dissolved matter of all kinds even crystalloids. 



In all cases of diffusion it is obvious that any dissolved substances 

 move from the stronger solution to the weaker one, so that equality in 

 concentration is approached. In many cases the attainment of this 

 equality is facilitated by the movement of a greater quantity of the 

 solvent from the weak to the strong solution. Certain substances, 

 when arranged as a partition between a solution and the solvent, will 

 permit of the passage of the solvent only, but not the dissolved sub- 

 stance. Perfect semi-permeable membranes, as they are termed, are 

 not known, but a near approach to them can be made artificially by 

 produsing a precipitate of copper ferrocyanide in the walls of a porous 

 earthf nware cell. If such a cell be filled with a solution, closed, and 

 its int irior be connected with a manometer, it will, when immersed in 

 the sclvent, show the production of considerable pressure, in some 

 cases amounting to several atmospheres. The pressure attained in 



