382 Dialysis. 



quiry as to the rapidity with which, liquids of different densities 

 diffused themselves followed almost as a natural sequel to that 

 respecting the gases. Mr. Graham'' s first experiments on the 

 diffusion of liquids were made by means of what he terms filial 

 diffusion, and they were performed as follows : — Solutions of 

 different salts, whose diffusive powers were to be examined, 

 were prepared of equal strength, and phials of exactly the same 

 size and shape were filled ■with these solutions, and then placed 

 separately under the surface of water contained in much larger 

 vessels, the mouths of the phials being left open. Under these 

 circumstances it was found that a certain proportion of the 

 heavy solution contained in the phial rose in opposition to the 

 attraction of gravitation, and mingled with the water by which 

 the phial was surrounded. In the case of coloured solutions, 

 this diffusion was visible to the eye, and in others it was capable 

 of being proved by analysis. It was found, however, that the so- 

 lutions of different bodies diffused themselves with very different 

 degrees of velocity. Thus common salt diffused with twice the 

 rapidity of Epsom salts or sugar. These, again, are double as di- 

 fusive as a solution of gum ; and albumen, or white of egg, in its 

 turn, does not possess one-fourth of the diffusive power of gum, 

 nor scarcely more than one-twentieth of that of common salt. 



These experiments were varied in different modes, by allow- 

 ing the diffusion to take place under slightly varying conditions, 

 but the same general results were obtained. The laws deduced 

 from these phenomena are, that crystalline bodies — such as 

 salt, sugar, nitre, etc. — are much more readily diffusible than 

 those that are amorphous, such as gum, gelatine, albumen, 

 solution of starch, or any substances that enter into combination 

 with water in the same manner that they do. 



Hence, with reference to this subject, Mr. Graham arranges 

 substances into two groups : those crystalline in character 

 and readily diffusible in water he terms crystalloids ; the solu- 

 tion of these is always free from gumminess or viscocity, is 

 sapid, possessing, in a higher or lower degree, the power of 

 affecting the nerves of taste. The other class, whose diffusive 

 power is low, he distinguishes as colloids, because gelatine or 

 glue (colle) may be taken as their type. The solutions of these 

 substances have no disposition to crystallize, and in the solid 

 form they do not possess flat surfaces, such as characterize crys- 

 tals, but exhibit an irregular roundness of outline. Their 

 solutions are always gummy when concentrated, and what is 

 strikingly remarkable, they are all insipid or wholly tasteless. 

 In the moist condition they are liable to undergo great changes, 

 and solutions of them in a state of purity cannot be preserved 

 unaltered for any length of time. 



A solution of a colloid body such as gelatine is found to 



