Diffusion of Liquids. 315 



solution, and then carefully filled with water and placed in a 

 large reservoir likewise containing water. Here it stood for 

 several weeks, the water in the outer vessel being removed 

 from time to time and fresh added. The concentration of the 

 liquid in the cylinder was then determined at different heights, 

 and from these data the amounts of salt diffused to the same 

 heights reckoned. 



It was found that these amounts decreased from below 

 upwards, and indeed in such a manner as to satisfy the require- 

 ments of this previously assumed hypothesis : — " If in a cylin- 

 drical vessel dynamic equilibrium exists, the concentration- 

 differences of any two layers must be related to each other as 

 the distances separating them ; in other words, the densities 

 must decrease from below upward as the ordinates of a right 

 line." 



In another experiment, Fick arranged three tubes of equal 

 diameters but of unequal lengths in the same manner, and 

 allowed the salt to diffuse through them into large vessels of 

 water. He found in this case that the amounts of diffused salt 

 were inversely proportional to the lengths of the tubes, the 

 concentrations in the reservoirs below being the same. 



These results can be expressed in the following equation, 



S =^« ; 



in which S is the amount of salt passing the horizontal section 

 q in the time t ; u x and u 2 , the concentrations of two layers ; and 

 ?, the distance separating them, supposing that a condition of 

 constancy (Beharrungszustanat) has been reached ; k is the 

 so-called constant of diffusion, depending on the nature of the 

 salt in question. 



Beilstein* was the next who published on the subject. He, 

 as Graham, considered the chemical rather than the physical 

 side of the problem; and the method employed by him was in 

 substance as follows. A glass tube about 3 inches long was 

 bent at one end, and the shorter arm ground off until it was 

 not more than a millimetre in length ; the other end was 

 drawn out and fitted with a ground glass stopper. This dif- 

 fusion-cell, resembling somewhat a very short siphon-baro- 

 meter, was filled with salt-solution of known strength and 

 suspended vertically in a large vessel of pure water. Diffusion 

 now took place between the solution in the short arm of the 

 glass and the medium above. After allowing this to continue 

 a few days, the small vessel was removed and the salt still 

 remaining in it determined ; subtracting this from the original 



* Ann. Chem. Pharm. xcix. p. 165, 



2A2 



